deviation from ideal mental health case study

Psychopathology Revision Notes

Saul Mcleod, PhD

Editor-in-Chief for Simply Psychology

BSc (Hons) Psychology, MRes, PhD, University of Manchester

Saul Mcleod, PhD., is a qualified psychology teacher with over 18 years of experience in further and higher education. He has been published in peer-reviewed journals, including the Journal of Clinical Psychology.

Learn about our Editorial Process

Olivia Guy-Evans, MSc

Associate Editor for Simply Psychology

BSc (Hons) Psychology, MSc Psychology of Education

Olivia Guy-Evans is a writer and associate editor for Simply Psychology. She has previously worked in healthcare and educational sectors.

On This Page:

Psychopathology refers to the study of mental disorders in terms of their causes, development, course, classification, and treatment.

Psychopathology describes a wide array of mental health conditions, including but not limited to depression, anxiety disorders, bipolar disorder, schizophrenia, and various personality disorders.

Understanding and studying psychopathology is a crucial part of diagnosing and treating these mental health disorders.

Definitions of Abnormality

Statistical infrequency.

Under this definition of abnormality , a person’s trait, thinking, or behavior is classified as abnormal if it is rare or statistically unusual.  With this definition, it is necessary to be clear about how rare a trait or behavior needs to be before we class it as abnormal.

For instance, one may say that an individual with an IQ below or above the average level of IQ in society is abnormal.

However, this definition obviously has limitations, it fails to recognize the desirability of the particular behavior.

Going back to the example, someone who has an IQ level above the normal average wouldn’t necessarily be seen as abnormal. Rather, on the contrary, they would be highly regarded for their intelligence.

This definition also implies that abnormal behavior in people should be rare or statistically unusual, which is not the case.  Instead, any specific abnormal behavior may be unusual, but it is not unusual for people to exhibit some form of prolonged abnormal behavior at some point in their lives.

This definition can provide an objective way, based on data, to define abnormality if an agreed cut-off point can be identified.

No value judgments are made –Homosexuality was defined as a mental disorder under early versions of the diagnostic criteria used by psychiatrists and would not be seen as ‘wrong’ but merely as less frequent than heterosexuality.

Limitations

However, this definition fails to distinguish between desirable and undesirable behavior. Statistically speaking, many very gifted individuals could be classified as ‘abnormal’ using this definition. The use of the term ‘abnormal’ in this context would not be appropriate.

Many rare behaviors or characteristics (e.g., left-handedness) have no bearing on normality or abnormality.  Some characteristics are regarded as abnormal even though they are quite frequent.  Depression may affect 27% of elderly people (NIMH, 2001).  This would make it common, but that does not mean it isn’t a problem

Deviation from Social Norms

Deviation from ideal mental health suggests that we define mental illness by looking at the absence of signs of physical health (Jahoda).

A person’s thinking or behavior is classified as abnormal if it violates the (unwritten) rules about what is expected or acceptable behavior in a particular social group. Their behavior may be incomprehensible to others or make others feel threatened or uncomfortable. Social behavior varies markedly when different cultures are compared.

For example, it is common in Southern Europe to stand much closer to strangers than in the UK. Voice pitch and volume, touching, the direction of gaze, and acceptable subjects for discussion have all been found to vary between cultures.

With this definition, it is necessary to consider: (i) The degree to which a norm is violated, the importance of that norm, and the value attached by the social group to different sorts of violations. (ii), E.g., is the violation rude, eccentric, abnormal, or criminal?

Comprehensive – Covers a broad range of criteria, most of which is why someone would seek help from mental health services or be referred for help – Makes it a good tool for thinking about mental health.

This definition gives a social dimension to the idea of abnormality, which offers an alternative to the “sick in the head” individual.

Social norms can vary from culture to culture. This means that what is considered normal in one culture may be considered abnormal in another. This definition of abnormality is an example of cultural relativism .

Unrealistic – Most people do not meet all the ideals because few people experience personal growth all the time – The criteria may be ideals rather than actualities.

One limitation of the deviation of social norms definition is that norms can vary over time. This means that behavior that would have been defined as abnormal in one era is no longer defined as abnormal in another.

For example, drink driving was once considered acceptable but is now seen as socially unacceptable, whereas homosexuality has gone the other way. Until 1980 homosexuality was considered a psychological disorder by the World Health Organization (WHO), but today is considered acceptable.

Failure to Function Adequately

Failure to function adequately (FFA) refers to an abnormality that prevents the person from carrying out the range of behaviors that society would expect, such as getting out of bed each day, holding down a job, and conducting successful relationships, etc.

Rosenhan & Seligman suggested seven criteria that are typical of FFA. These include personal distress (e.g., anxiety or depression), unpredictability (displaying unexpected behaviors and loss of control), and irrationality, among others. The more features of personal dysfunction a person has, the more they are considered abnormal.

To assess how well individuals cope with everyday life, clinicians use the Global Assessment of Functioning Scale (GAF), which rates their level of social, occupational, and psychological functioning.

The definition provides a practical checklist of seven criteria individuals can use to check their level of abnormality.

It matches the sufferers’ perceptions. As most people seeking clinical help believe that they are suffering from psychological problems that interfere with the ability to function properly, it supports the definition.

FFA might not be linked to abnormality but to other factors. Failure to keep a job may be due to the economic situation, not to psychopathology.

Cultural relativism is one limitation; what may be seen as functioning adequately in one culture may not be adequate in another. This is likely to result in different diagnoses in different cultures.

FFA is context dependent; not eating can be seen as failing to function adequately, but prisoners on hunger strikes making a protest can be seen in a different light.

Deviation from Ideal Mental Health

Jahoda suggested six criteria necessary for ideal mental health. An absence of any of these characteristics indicates individuals as being abnormal, in other words displaying deviation from ideal mental health.

• Resistance to stress: Having effective coping strategies and being able to cope with everyday anxiety-provoking situations.
• Growth, development, or self-actualization: Experiencing personal growth and becoming everything one is capable of becoming.
• High self-esteem and a strong sense of identity: Having self-respect and a positive self-concept.
• Autonomy: Being independent, self-reliant, and able to make personal decisions.
• Accurate perception of reality: Having an objective and realistic view of the world.

The difficulty of meeting all criteria, very few people would be able to do so, and this suggests that very few people are psychologically healthy.

Cultural relativism: these ideas are culture-bound, based on a Western idea of ideal mental health, and should not be used to judge other cultures.

AO2 Scenario Question

Diane is a 30-year-old businesswoman, and if she does not get her own way, she sometimes has a temper tantrum. Recently, she attended her grandmother’s funeral and laughed during the prayers. When she talks to people, she often stands very close to them, making them feel uncomfortable.

Identify one definition of abnormality that could describe Diane’s behavior. Explain your choice.

“Diana’s behavior could be defined as deviating from social norms.

Although she is 30 she still has childish temper tantrums, she acted in a socially abnormal way at her grandmother’s funeral and she disobeys social norms about how close it is appropriate to stand to people.

She is deviating from what is regarded as socially normal, thus according to this definition she would be defined as psychologically abnormal.”

The following article appeared in a magazine:

‘Hoarding disorder – A ‘new’ mental illness

Most of us are able to throw away the things we don’t need on a daily basis. Approximately 1 in 1000 people, however, suffer from hoarding disorder, defined as ‘a difficulty parting with items and possessions, which leads to severe anxiety and extreme clutter that affects living or work spaces.’

Apart from ‘deviation from ideal mental health,’ outline three definitions of abnormality. Refer to the article above in your answer. (6 marks)

The Biological Approach to OCD

Characteristics of ocd.

Obsessive Compulsive Disorder (OCD) is an anxiety disorder characterized by intrusive and uncontrollable thoughts (i.e., obsessions) coupled with a need to perform specific acts repeatedly (i.e., compulsions).

Common clinical obsessions are fear of contamination (esp., being infected by germs), repetitive thoughts of violence (killing or harming someone), sexual obsessions, and obsessive doubt. Compulsions are the behavioral responses intended to neutralize these obsessions.

The most common compulsions are cleaning, washing , checking , counting , and touching. To the compulsive, these behaviors often seem to have magical qualities. If they are not performed exactly, “something bad” will happen.

Some O.C.D. sufferers will meticulously perform their rituals hundreds of times and experience extreme anxiety if prevented from carrying them out. Cleaning/washing rituals are more common in women; checking rituals are more common in men.

Cognitive (What do you THINK?): Obsessions dominate ones thinking and are persistent and recurrent thoughts, images, or beliefs entering the mind uninvited and which cannot be removed. At some point during the course of the disorder, the person has recognized that the obsessions or compulsions are excessive or unreasonable.

Emotional (How do you FEEL?): Obsessive thoughts often lead to anxiety, worry, and distress.

Behavioral (How do you BEHAVE?): Compulsions are the repetitive behavioral responses intended to neutralize these obsessions, often involving rigidly applied rules. Most OCD sufferers recognize their compulsions as unreasonable but believe something bad will happen if they don’t perform that behavior.

A02 Exam Style Question

Steven describes how he feels when he is in a public place.

I always have to look out for people who might be ill. If I come into contact with people who look ill, I think I might catch it and die. If someone starts to cough or sneeze, then I have to get away and clean myself quickly.

Outline one cognitive characteristic of OCD and one behavioral characteristic of OCD that can be identified from the description provided by Steven. (2 marks)

The approach can also be criticized for ignoring environmental influences. For example, people are not born with OCD. They might learn it from their environment through the process of classical and operant conditioning.

Strengths of this approach include its testability via neuroscience research  and evidence for genetic and neurotransmitte r involvement in conditions such as schizophrenia. For example, the dopamine hypothesis argues that elevated levels of dopamine  are related to symptoms of schizophrenia.

Biological explanations are reductionist as they focus on only one factor, and at present, our understanding of biochemistry is oversimplified. This means other psychological factors, such as cognitions, are ignored.

The biological explanations are also deterministic because they ignore the individual’s ability to control their own behavior, which in turn may affect their biochemistry levels.

Genetic Explanations

Genetics is the study of genes and inheritance. OCD seems to be a polygenic condition, where several genes are involved in its development. Family and twin studies suggest the involvement of genetic factors. The prevalence of OCD in the random population (about 2–3%) is the baseline against which the concordance rates can be compared.

The SERT gene (Serotonin Transporter) appears mutated in individuals with OCD. The mutation causes an increase in transporter proteins at a neuron’s membrane. This leads to an increase in the reuptake of serotonin in the neuron, which decreases the level of serotonin in the synapse.

The COMT gene is a gene that regulates the function of dopamine. It appears that this gene is also mutated in individuals with OCD. However, this mutation causes the opposite effect as the SERT mutation discussed above. The mutated variation of the COMT gene found in OCD individuals causes a decrease in COMT activity and, therefore, a higher level of dopamine.

Carey and Gottesman (1981) found that identical twins showed a concordance rate of 87% for obsessive symptoms and features compared to 47% in fraternal twins. This difference suggests that genetic factors are moderately important.

The higher concordance rate found for identical twins may be due to nurture, as identical twins are likely to experience a more similar environment than fraternal twins since they tend to be treated the same.

Genes alone do not determine who will develop OCD—they only create vulnerability. Thus, they are not a direct cause, as other factors must trigger the disorder.

Evidence for this is that the concordance rates are not 100%, which shows that OCD is due to an interaction of genetic and other factors.

OCD may be culturally rather than genetically transmitted as the family members may observe and imitate each other’s behavior, as predicted by social learning theory. Alternatively, family members might be more vulnerable to OCD because of the stressful environment rather than genetic factors.

Neural Explanations

Neural mechanisms refer to brain regions , structures such as neurons , and the neurotransmitters involved in sending messages through the nervous system.

One region of the brain, the prefrontal cortex (PFC), is involved in decision-making and the regulation of primitive aspects of our behavior. An overactive PFC causes an exaggerated control of primal impulses.

For example, after a visit to the bathroom, your primal instinct to survive by avoiding germs is brought to your attention. You may make the decision to wash your hands to remove any harmful germs you may have encountered.

Once you have performed the appropriate behavior, the PFC reduces in activation, and you stop washing your hands and go about your day. It has been suggested that if you have OCD, your PFC is over-activated. This means the obsessions and compulsions continue, leading you to wash your hands again and again.

Abnormalities, or an imbalance in the neurotransmitter serotonin, could also be related to OCD. Reduced serotonin and excessive dopamine may cause OCD.

Serotonin is the chemical thought to be involved in regulating mood. OCD patients have low levels of serotonin.

Additionally, Dopamine is abnormally high in individuals with OCD. High levels of dopamine have been thought to influence concentration. This may explain why OCD individuals experience an inability to stop focusing on obsessive thoughts and repetitive behaviors.

The brains of OCD patients are structured and function differently from those of other people. Brain scans of OCD patients reliably show increased activity in the PFC (Salloway & Duffy, 2002).

Whether low serotonin causes OCD is unknown. All that’s known is that low serotonin and OCD are related. It is difficult to establish whether the low levels of neurotransmitters cause OCD, are an effect of having the disorder, or are merely associated. Causation cannot be inferred as only associations(i.e., correlations) have been identified.

We do not know whether high levels of dopamine cause OCD or whether OCD is caused by something else and the effect is high levels of dopamine.

The biochemistry hypothesis does not account for individual differences because the research does not explain why one individual develops OCD and another develops a different mental disorder because low serotonin levels are also found in other mental disorders. Thus, these biochemical abnormalities are not specific to OCD and may be true of mental distress.

Psychological therapy (CBT) can be a very successful treatment, and this is difficult to account for in the serotonin hypothesis.

Biological Treatment – Drugs

Two classes of drugs have proved effective in the treatment of obsessive-compulsive disorder: serotonin reuptake inhibitors (SRIs) and selective serotonin reuptake inhibitors (SSRIs). Both classes of drug increase serotonin levels and so support the neural explanation / biochemical hypothesis.

Drugs that mainly affect neurotransmitters other than serotonin are of little or no value in treating obsessive-compulsive disorder.

Studies using drugs have shown a reduction in dopamine levels is positively correlated with a reduction in OCD symptoms.

Experiments that inject animals with drugs that increase levels of dopamine have caused the animals to demonstrate OCD-type behaviors.

Drugs that increase serotonin (antidepressants such as SSRIs) have been shown to reduce OCD symptoms. Soomro et al. found that SSRIs were significantly better than placebos in reducing symptoms in 17 different clinical trials

But research results relating to serotonin are varied – sometimes symptoms have been made worse. There is a great deal of contradictory research.

Drugs show only partial alleviation of the symptoms, so the process is not fully understood. The exact function of neurotransmitters in the development of OCD is far from understood.

Most SSRIs have side effects that can be unpleasant, e.g., dry mouth, a slight tremor, fast heartbeat, constipation, sleepiness, and weight gain.

The success of antidepressant drugs as a treatment does not necessarily mean the biochemicals are the cause of OCD in the first place. This is known as the treatment etiology fallacy, and, using headaches as an example, aspirin works well as a treatment, but this doesn’t mean the headache was due to an absence of aspirin.

Cognitive Approach to Depression

Characteristics of depression.

Depression is a mood or affective disorder. This mental Illness is a collection of physical, emotional, mental, and behavioral experiences that are severe, prolonged, and damaging to everyday functioning.

The criteria for depression to be diagnosed using the DSM-IV-TR is that at least five or more symptoms of depression should be apparent. The possible symptoms include:

Behavioral (How do you BEHAVE when you’re depressed?): Neglect of personal appearance, loss of appetite, disturbed sleep patterns (insomnia), loss of energy (tiredness), withdrawal from others.

Emotional (How do you FEEL when you’re depressed?): Intense sadness, irritability, apathy (loss of interest or enjoyment), feelings of worthlessness, and anger.

Cognitive (How do you THINK when you’re depressed?): Negative thoughts, lack of concentration, low self-esteem, poor memory, recurrent thoughts of death, and low confidence.

The cognitive approach believes depression stems from faulty cognitions about others, our world, and us. This faulty thinking may be through cognitive deficiencies (lack of planning) or cognitive distortions (processing information inaccurately). These cognitions cause distortions in the way we see things and cause behavior such as depression.

Ellis suggested depression occurs through irrational thinking, while Beck proposed the cognitive triad.

Ben recently moved away from home to go to university. He loved his new life of going out, meeting new friends, and his new university course. However, after a while, he struggled to get out of bed and started to become very tired.

His eating patterns changed, and he lost a lot of weight. He noticed that he got angry at little things and snapped at his friends. When he sat in lectures, he found it hard to concentrate for long periods of time.

Identify the behavioral, emotional, and cognitive aspects of Ben’s state. (3 marks)

Beck’s Negative Triad

The cognitive triad is three forms of negative (i.e., helpless and critical) thinking that are typical of individuals with depression: namely, negative thoughts about the self, the world, and the future. These thoughts tended to be automatic in depressed people as they occurred spontaneously.

For example, depressed individuals tend to view themselves as helpless, worthless, and inadequate. They interpret events in the world in an unrealistically negative and defeatist way, and they see the world as posing obstacles that can’t be handled.

Finally, they see the future as totally hopeless because their worthlessness will prevent their situation from improving.

The negative triad interacts with negative schemas and cognitive biases to produce depressive thinking.

Cognitive biases are distortions of thought processes. Individuals with depression are prone to making logical errors in their thinking, and they tend to focus selectively on certain negative aspects of a situation while ignoring equally relevant positive information.

In addition to cognitive biases, the negative triad is also influenced by schemas . In essence, schemas can be seen as deeply held beliefs that have their origins primarily in childhood. Beck believed that depression-prone individuals develop a negative self-schema. They possess a set of beliefs and expectations about themselves that are essentially negative and pessimistic.

Beck claimed that negative schemas might be acquired in childhood as a result of a traumatic event (e.g., parental or peer rejection). Schemas influence how a person interprets events and experiences in their life. Beck predicted that in depression, ‘latent’ (i.e., dormant) negative schemas that have been formed in childhood become activated by life events or ongoing stressors.

Negative schemas and cognitive biases maintain the negative triad, a pessimistic view of the self, the world (not being able to cope with the demands of the environment), and the future.

It may be that negative thinking generally is also an effect rather than a cause of depression. Perhaps individuals only start experiencing negative thoughts after having developed depression. However, evidence that negative thinking can be involved in the development of depression was obtained by Lewinsohn et al. (2001).

They measured negative thinking in non-depressed adolescents. One year later, the life events of participants over the previous 12 months were assessed, as also whether they were suffering from depression.

The results showed those who had experienced many negative life events had an increased likelihood of developing depression only if they were initially high in negative attitudes. This study supports the theory that negative beliefs are a risk factor for developing depression when exposed to stressful life events.

The cognitive approach to depression is limited in that genetic factors are ignored.

Little attention is paid to the role of social factors relating to life events and gender in the cognitive explanation of depression.

Ellis’ ABC Model

Albert Ellis (1957, 1962) proposes that each of us holds a unique set of assumptions/beliefs about ourselves and our world that serve to guide us through life and determine our reactions to the various situations we encounter.

Unfortunately, some people’s assumptions are largely irrational, guiding them to act and react in ways that are inappropriate and that prejudice their chances of happiness and success. Albert Ellis calls these basic irrational assumptions.

According to Ellis, depression does not occur as a direct result of a negative event but rather is produced by irrational thoughts (i.e., beliefs) triggered by negative events.

Ellis believes that it is not the activating event (A) that causes depression (C) but rather that a person interprets these events unrealistically and therefore has an irrational belief system (B) that helps cause the consequences (C) of depressive behavior.

For example, some people irrationally assume that they are failures if they are not loved by everyone they know (B) – they constantly seek approval and repeatedly feel rejected (C). All their social interactions (A) are affected by this assumption, so a great party can leave them dissatisfied because they don’t get enough compliments.

The precise role of cognitive processes is yet to be determined. It is not clear whether faulty cognitions are a cause of psychopathology or a consequence of it.

Sometimes these negative cognitions are, in fact, a more accurate view of the world: depressive realism.

Cognitive theories lend themselves to testing. When experimental subjects are manipulated into adopting unpleasant assumptions or thoughts, they become more anxious and depressed (Rimm & Litvak, 1969).

Treatment – CBT

How would you use the therapy.

Cognitive behavioral therapy aims to change the way a client thinks by challenging irrational and maladaptive thought processes, and this will lead to a change in behavior as a response to new thinking patterns. Specifically, our thoughts determine our feelings and our behavior.

Therefore, negative – and unrealistic – thoughts can cause us distress and result in problems. When a person suffers from psychological distress, the way in which they interpret situations becomes skewed, which in turn, has a negative impact on the actions they take.

Cognitive therapists help clients to recognize the negative thoughts and errors in logic that cause them to be depressed. The therapist also guides clients to question and challenge their dysfunctional thoughts, try out new interpretations, and ultimately apply alternative ways of thinking in their daily lives.

The clients learn to discriminate between their own thoughts and reality. They learn the influence that cognition has on their feelings, and they are taught to recognize, observe and monitor their own thoughts.

The behavior part of the therapy involves setting homework for the client to do (e.g., keeping a diary of thoughts). The therapist gives the client tasks that will help them challenge their own irrational beliefs.

The idea is that the client identifies their own unhelpful beliefs and then proves them wrong. As a result, their beliefs begin to change. For example, someone who is anxious in social situations may set a homework assignment to meet a friend at the pub for a drink.

CBT would be used when a person’s faulty thinking is affecting their life in a negative way.

A strength of this therapy is that it has shown to be very effective in treating depression; in fact, it has been shown to produce longer-lasting recovery than antidepressants.

Sometimes these negative cognitions are in fact a more accurate view of the world: depressive realism.

An important advantage of CBT is that it tends to be short (compared to psychoanalysis), taking three to six months for most emotional problems. Patients attend a session a week, each session lasting either 50 minutes or an hour.

Another strength is that it can reduce ethical issues – the way this therapy works is that the client is actively involved and in control. They feel empowered as they help themselves.

Jack suffers from depression. His symptoms include loss of concentration, lack of sleep, and struggles to sleep at night. He finds himself having absolutist thoughts that everything is negative and bad all the time.

How might a cognitive behavior therapist tackle Jack’s depression? (4 marks)

Behavioral Approach to Phobias

Characteristics of phobias.

Phobias are a type of anxiety disorder. Phobias are characterized by a marked and persistent fear that is excessive or unreasonable, cued by the presence or anticipation of a specific object or situation (e.g., flying, heights, seeing blood).

The symptoms of phobias can be placed into one of three categories:

Behavioral (How do you BEHAVE when you see your feared object?): The phobic stimulus is either avoided or responded to with great anxiety. For example, someone with a phobia of dogs may cross the road every time they see a dog, therefore receiving negative reinforcement , which will maintain the phobia. This avoidance could interfere with the individual’s normal daily routine.

Emotional (How do you FEEL when you see your feared object?): Exposure to a phobic stimulus nearly always produces a rapid anxiety response.

Cognitive (What do you THINK about your feared object?): A person would recognize that the fear is excessive or unreasonable. The person is consciously aware that the anxiety levels they experience in relation to their feared object or situation are overstated.

The DSM defines three categories of phobias: agoraphobia, social phobia, and specific phobias. Agoraphobia is a fear of open spaces but is better characterized as a fear of being away from home.

Social phobias involve intense fear of social situations or having to interact with other people. Specific phobias relate to a fear of a specific object, such as a spider, or a situation, such as an enclosed space (claustrophobia).

The Two-Process Model

The behavioral approach explains the development and maintenance of phobia, mainly using the theories of classical conditioning and operant conditioning. These were first combined as a single explanation for phobia by Mowrer in the two-process model of phobia.

According to behaviorists, phobias are the result of a classically conditioned association between an anxiety-provoking  unconditioned stimulus (UCS) and a previously neutral stimulus.

For example, a child with no previous fear of dogs gets bitten by a dog and, from this moment onwards, associates the dog with fear and pain. Due to the process of generalization, the child is not just afraid of the dog who bit them but shows a fear of all dogs.

Operant conditioning can help to explain how the phobia is maintained. The conditioned (i.e., learned) stimulus evokes fears, and avoidance of the feared object or situation lessens this feeling, which is rewarding. The reward (negative reinforcement) strengthens the avoidance behavior, and the phobia is maintained.

A02 Questions

Kirsty is in her twenties and has had a phobia of balloons since one burst near her face when she was a little girl. Loud noises such as ‘banging’ and ‘popping’ cause Kirsty extreme anxiety, and she avoids situations such as birthday parties and weddings, where there might be balloons.

Suggest how the behavioral approach might be used to explain Kirsty’s phobia of balloons. (4 marks)

There is empirical support to show how classical conditioning leads to the development of phobias. Watson and Rayner (1920) used classical conditioning to create a phobia in an infant called Little Albert . Albert developed a phobia of a white rat when he learned to associate the rat with a loud noise.

The behaviorist approach adopts a limited in the origins of a phobia, as it overlooks the role of cognition. Ignoring the role of cognition is problematic, as irrational thinking appears to be a key feature of phobias.

Tomarken et al. (1989) presented a series of slides of snake and neutral images (e.g., trees) to phobic and non-phobic participants. The phobics tended to overestimate the number of snake images presented.

In theory, anyone could develop a phobia of a potentially harmful object, although this does not always happen. Despite the fact that most adults have either experienced, witnessed or heard about car accidents where another person is injured, the phobia of cars is virtually non-existent.

Seligman (1970) suggests that humans have a biological preparednes s to develop certain phobias rather than others because they were adaptive (i.e., helpful) in our evolutionary past. For example, individuals that avoided snakes and high places would be more likely to survive long enough and pass on their genes than those who did not.

The idea of biological preparedness is further supported by Ost and Hugdahl (1981), who claims that nearly half of all people with phobias have never had an anxious experience with the object of their fear, and some have had no experience at all. For example, some snake phobics have never encountered a snake.

The cognitive approach criticizes the behavioral model as it does not take mental processes into account. They argue that the thinking processes that occur between a stimulus and a response are responsible for the feeling component of the response.

Treatment – Systematic Desensitization

Systematic desensitization is a type of behavioral therapy based on the principle of classical conditioning. This therapy aims to remove the fear response of a phobia and substitute a relaxation response to the conditional stimulus gradually using counter-conditioning. This will lead to the extinction of the fear response. There are three phases to the treatment:

First , the patient is taught a deep muscle relaxation technique and breathing exercises. E.g., control over breathing, muscle detensioning, or meditation. This step is very important because of reciprocal inhibition, where one response is inhibited because it is incompatible with another. In the case of phobias, fears involve tension, and tension is incompatible with relaxation.

Second , the patient creates a fear hierarchy starting with stimuli that create the least anxiety (fear) and building up in stages to the most fear-provoking images. The list is crucial as it provides a structure for the therapy.

Third , the patient works their way up the fear hierarchy, starting at the least unpleasant stimuli and practicing their relaxation technique as they go. When they feel comfortable with this (they are no longer afraid), they move on to the next stage in the hierarchy. If the client becomes upset, they can return to an earlier stage and regain their relaxed state.

The number of sessions required depends on the severity of the phobia. Usually, 4-6 sessions, up to 12, for a severe phobia. The therapy is complete once the agreed therapeutic goals are met (not necessarily when the person’s fears have been completely removed).

Exposure can be done in two ways:

· In vitro – the client imagines exposure to the phobic stimulus.

· In vivo – the client is actually exposed to the phobic stimulus.

Research has found that in vivo techniques are more successful than in vitro (Menzies and Clarke 1993). However, there may be practical reasons why in vitro may be used.

Practical Issues

One weakness of in vitro systematic desensitization is that it relies on the client’s ability to be able to imagine the fearful situation. Some people cannot create a vivid image, and thus, systematic desensitization is not always effective (there are individual differences).

Systematic desensitization is a slow process, taking, on average, 6-8 sessions. Although, research suggests that the longer the technique takes, the more effective it is.

Theoretical Issues

Systematic desensitization is highly effective where the problem is learned anxiety about specific objects/situations (e.g., phobias). However, SD is not effective in treating serious mental disorders like depression and schizophrenia.

Studies have shown that neither relaxation nor hierarchies are necessary and that the important factor is just exposure to the feared object or situation. Therefore, therapies like flooding may be more effective.

Social phobias and agoraphobia do not seem to show as much improvement. Could it be that there are other causes for phobias than classical conditioning?

For example, if a fear of public speaking originates with poor social skills, then phobic reduction is more likely to occur in a treatment that includes learning effective social skills than systematic desensitization alone.

Empirical Evidence

Rothbaum used SD with participants who were afraid of flying. Following treatment, 93% agreed to take a trial flight. It was found that anxiety levels were lower than those of a control group who had not received SD, and this improvement was maintained when they were followed up six months later.

Ethical Issues

SD creates high levels of anxiety when patients are initially exposed, which raises ethical issues and so questions of appropriateness. It should be noted that virtual reality therapy does help resolve these issues.

Treatment – Flooding

Flooding (also known as implosion therapy) works by exposing the patient directly to their worst fears. (S)he is thrown in at the deep end. For example, a claustrophobic will be locked in a closet for 4 hours, or an individual with a fear of flying will be sent up in a light aircraft.

What flooding aims to do is expose the sufferer to the phobic object or situation for an extended period of time in a safe and controlled environment. Unlike systematic desensitization, which might use in vitro or virtual exposure, flooding generally involves vivo exposure.

Fear is a time-limited response. At first, the person is in a state of extreme anxiety, perhaps even panic, but eventually, exhaustion sets in, and the anxiety level begins to go down. Of course, normally, the person would do everything they can to avoid such a situation.

Now they have no choice but to confront their fears, and when the panic subsides, they find they have come to no harm. The fear (which, to a large degree, was anticipatory) is extinguished.

Prolonged intense exposure eventually creates a new association between the feared object and something positive (e.g., a sense of calm and lack of anxiety). It also prevents the reinforcement of phobia through escape or avoidance behaviors.

Flooding is rarely used, and if you are not careful, it can be dangerous. It is not an appropriate treatment for every phobia. It should be used with caution as some people can actually increase their fear after therapy, and it is not possible to predict when this will occur.

Wolpe (1969) reported the case of a client whose anxiety intensified to such as degree that flooding therapy resulted in her being hospitalized.

Also, some people will not be able to tolerate the high levels of anxiety induced by the therapy and are, therefore, at risk of exiting the therapy before they are calm and relaxed. This is a problem, as an existing treatment before completion is likely to strengthen rather than weaken the phobia.

However, one application is for people who have a fear of water (they are forced to swim out of their depth). It is also sometimes used with agoraphobia. In general, flooding produces results as effective (sometimes even more so) as systematic desensitization.

The method’s success confirms the hypothesis that phobias are so persistent because the object is avoided in real life and is therefore not extinguished by the discovery that it is harmless.

For example, Wolpe (1960) forced an adolescent girl with a fear of cars into the back of a car and drove her around continuously for four hours: her fear reached hysterical heights but then receded and, by the end of the journey, had completely disappeared.

What do the examiners look for?

• Accurate and detailed knowledge
• Clear, coherent, and focused answers
• Effective use of terminology (use the “technical terms”)

In application questions, examiners look for “effective application to the scenario,” which means that you need to describe the theory and explain the scenario using the theory making the links between the two very clear. If there is more than one individual in the scenario you must mention all of the characters to get to the top band.

Difference between AS and A level answers

The descriptions follow the same criteria; however, you have to use the issues and debates effectively in your answers. “Effectively” means that it needs to be clearly linked and explained in the context of the answer.

Read the model answers to get a clearer idea of what is needed.

A-Level Psychology Revision Notes

A-Level Psychology Attachment

Psychology Memory Revision Notes

Social Influence Revision Notes

Psychology Approaches Revision for A-level

Research Methods: Definition, Types, & Examples

Issues and Debates in Psychology (A-Level Revision)

What is the main focus of psychopathology?

The main focus of psychopathology is to study and understand mental illnesses, including their symptoms, causes, and effects on individuals. This can be broken down into several key areas:

Understanding Mental Disorders : This involves studying the specific symptoms and signs of various mental disorders, their course (how they change over time), and their impact on a person’s life.

Determining Causes : Psychopathology seeks to understand the root causes of mental disorders. These causes can be biological (like genetics or brain chemistry), psychological (like personality traits or thought patterns), and environmental/social (like traumatic experiences or societal pressures).

Classification : Psychopathology is also concerned with the development and refinement of classification systems for mental disorders, like the Diagnostic and Statistical Manual of Mental Disorders (DSM) and the International Classification of Diseases (ICD). These systems aim to provide a standardized way to diagnose and treat mental health conditions.

Treatment : An understanding of psychopathology helps in the development of effective treatments for mental disorders. These treatments may include psychotherapy, medication, or a combination of both.

Prevention : By understanding the causes and course of mental illnesses, professionals can devise strategies to prevent the onset of disorders.

So, while psychopathology covers a broad range of topics, the central focus is on understanding mental illness to diagnose better, treat, and prevent mental health disorders.

What is the difference between psychopathology and mental illness?

Psychopathology and mental illness are related but distinct concepts.

Psychopathology is the scientific study of mental disorders, encompassing their symptoms, causes, classifications, and treatments. It involves understanding abnormal thoughts, behaviors, and emotions that may indicate a mental disorder.

On the other hand, mental illness, also known as a mental disorder, refers to the conditions themselves, such as depression, anxiety, schizophrenia, and so on. These are diagnosed conditions characterized by alterations in thinking, mood, or behavior associated with distress or impaired functioning.

Who is the father of psychopathology?

The title “father of psychopathology” is often given to Emil Kraepelin, a German psychiatrist. In the late 19th and early 20th centuries, Kraepelin made significant contributions to the understanding and classification of mental disorders. His work emphasized the importance of long-term patterns and the course of mental disorders, laying the groundwork for our modern diagnostic system.

Definitions Of Abnormality, Including Deviation From Social Norms, Failure To Function Adequately And Deviation From Ideal Mental Health And Statistical Infrequency.

March 5, 2021 - paper 1 introductory topics in psychology | psychopathology.

• Back to Paper 1 - Psychopathology

Description, AO1:

Definition 1: Deviation from Social Norms (DSN)

Norms are commonly expected standards of behaving in a society according to the majority.

Sometimes these are written ( explicit ) and form laws that govern behaviour. Sometimes though, these norms of behaviour are unwritten ( implicit ) but generally accepted (e.g. not standing too close to people at a cash machine).

Whether explicit or implicit, social norms allow for the regulation of normal social behaviour. All societies have standards, or norms, for appropriate behaviours and beliefs (expectations about how people should behave and what they should think).

This definition therefore, suggests that  abnormality can be seen as someone who breaks these ‘rules of society’  (i.e. someone who doesn’t think or behave in a way that is accepted by that society).

Whenever behaviour breaks a “social norm” this is an indication that there may be some form of abnormality or may be a sign of a mental disorder.

Evaluation, AO3 of the DSN Definition:

Weaknesses:

(1) POINT:  The Social norms definition can be questioned, as social norms change over time and therefore a behaviour that broke social norms and was considered abnormal in 1950 may not be viewed as abnormal today.  EVIDENCE:  Being an unmarried mother in the 1940’s and 1950’s would be breaking social norms and therefore this behaviour was classed as abnormal. Many of these women were sectioned as ‘moral imbeciles’ and society demanded that they “give up” their babies.  EVALUATION:  This is a limitation because as this example shows, it is not the individual who has changed but the classification of the behaviour by society. This brings into question the validity of using this definition alone to define abnormality.

(2) POINT:  Social norms differ between cultures and this may be problematic.  EVIDENCE:  in British culture it is considered to be polite to finish the food on your plate at mealtimes. However, in India, to finish all food from you plate is a sign that you are still hungry.  EVALUATION:  This is a limitation because shows that what is considered ‘normal’ in one culture may actually be ‘abnormal’ in another. This means that this definition does not consistently produce an accurate definition of abnormal behaviour.

(3) POINT:  The “deviation from Social norms” definition of abnormality is limited because it suggests that all behaviour which breaks social norms is “abnormal” and it doesn’t distinguish between socially deviant behaviour and mental abnormality.  EVIDENCE:  For example, someone found guilty of being a drink-driver may have broken a social norm, but they are not considered to have a mental abnormality.  EVALUATION:  This is a limitation because it shows that the deviation from social norms definition is inadequate in identifying abnormality as there are many behaviours that are clearly socially deviant but not psychological abnormalities.

Definition 2: Failure to Function Adequately (FFA)

“Failing to function” is generally taken to mean that a person is  unable to cope with everyday life . Their behaviour is seen as maladaptive (not good for survival), disrupting their ability to work and conduct satisfying relationships and causes personal distress, as well as distress to others.

Rosenhan and Seligman (1989)  have suggested some  key characteristics of ‘failing to function adequately’:

• Personal Distress   Most people who seek psychiatric help of any kind are suffering from a sense of psychological distress or discomfort ( Sue et al, 1994 ) and a recognition that they are failing to function adequately.
• Observer Distress (or Discomfort)   Where someone’s behaviour causes discomfort and distress to others observing the behaviour.
• Maladaptive Behaviour  Where someone’s behaviour interferes with their ability to lead a normal life (e.g. agoraphobia)
• Unpredictable Behaviour-  If behaviour is unpredictable if it does not fit the situation or if it is unexpected and uncontrolled (e.g. sobbing for no reason or laughing at bad news).
• Irrational Behaviour  If a person’s behaviour doesn’t make sense to other people (difficult to understand)

Evaluation, AO3 of Failure to Function Adequately:

(1) POINT:  Failing to function does not always indicate the presence of a psychological abnormality.  EVIDENCE:  For example, someone who looses their job will not be able to get up and go to work/earn money for their family etc This may cause them personal distress but it does not indicate an abnormality.  EVALUATION:  This is a limitation as environmental factors may cause a failure to function rather than any psychological abnormality and this therefore questions the validity of using this definition to identify abnormality.

(2) POINT:  However, the presence of an abnormality doesn’t always result in a “failure to function.”  EVIDENCE:  An individual with depression may still be able to hold a job and run a family successfully. There are many examples of celebrities who have suffered from psychological disorders who have had successful careers and who would not be considered as “failing to function adequately” e.g., David Beckham has OCD, Stephen Fry has bipolar disorder. there are people who suffer from abnormalities such as depression however, are able to carry out everyday tasks like going to work, looking after a family etc They do not necessarily display a failure to function.  EVALUATION:  This is a limitation because it shows that this definition is inadequate in truly identifying behaviours that may be considered abnormal.

(3) POINT:   Behaviour that looks as if it is a “failure to function” may actually be behaviour that is admired by some in society.  EVIDENCE:  Some political prisoners who go on a hunger strike as part of their political protest will often be respected.  EVALUATION:  This is a limitation because although starving yourself may be seen as irrational, maladaptive and unpredictable, it is understandable in a particular social context and therefore would not be classed as a mental abnormality.

AO1, Description:

Definition 3: Deviation from Ideal Mental Health  (DIMH)

This definition stands out by not defining abnormality directly. Instead, it attempts to define a state of  ideal mental health  (i.e., factors necessary for ‘optimal living’).

Marie Jahoda  (1958) defined ideal mental health through a list of characteristics indicating psychological health and therefore the  absence  of the characteristics (rather than presence; as with the FFA definition) suggests abnormality.

Characteristics for ‘Optimal Living’:

• Strong sense of self-Identity   A psychologically healthy individual should be in touch with their own identity and feelings.
• Resistant to stress   Someone with ideal mental health should be resistant to stress and its negative effects.
• Self-actualisation   The mentally healthy individual should also be focused on the future and on fulfilling their potential (developing their talents and abilities).
• Autonomy   They should function as autonomous individuals, recognising their own needs and with an accurate perception of reality (i.e., to have the ability to act independently, to make their own decisions and to find satisfaction from within themselves).
• Empathy   The individual should show understanding towards others.

The fewer of these qualities you have, the more abnormal you are seen to be.

Evaluation, AO3 of the Deviation from Ideal Mental Health Definition:

(1) POINT:  The definition is problematic, as the characteristics of Ideal Mental Health are seen as far too idealistic. EVIDENCE:  Evidence to support this comes from Maslow (1968) who argued that only a few people ever achieve ‘self-actualisation’ and so the criteria is set too high.  EVALUATION:  This is a limitation because if this is true, most people are unable to achieve self-actualisation which means that the majority would be defined as being abnormal. However, if the majority of people are ‘abnormal’ then not achieving self-actualisation becomes ‘normal’ therefore indicating that the DIMH definition is a poor definition of abnormality.

(2) POINT:   The definition requires a subjective judgement on how many criteria need to be lacking in order to define someone as ‘abnormal.’  EVIDENCE:  For example, one individual might consider a lack of 2 criteria to be abnormal, whereas another would consider lacking 4 criteria to be abnormal.  EVALUATION:  This is a limitation because using subjective judgement in this way, decreases both the reliability (i.e., consistency) and the validity (i.e., accuracy) of this method of defining abnormality.

(3)  POINT:   The criteria used in the “Ideal Mental Health” definition, are based on Western culture and therefore the definition is ethnocentric.

Individualistic :  Cultures that place their emphasis on personal achievement                                                Collectivist:  Cultures that strive for the greater good of the community.

EVIDENCE:  For example, her emphasis on personal growth and individual autonomy reflects Western,  individualistic  culture (e.g. North-American) and not  collectivist  cultures.  EVALUATION:  This is a limitation because it means that the definition is subjective and should only be used in the culture in which it has been developed, is not value-free and therefore may be biased.

Definition 4: Statistical Infrequency

This approach defines abnormality as a behaviour that  deviates  from the  ‘average.’  The idea behind this approach is that the less frequently a behaviour occurs in people, the more likely it is to be abnormal. The difference between normal and abnormal is therefore one of quantity rather than quality the majority of people are normal with the minority of people being abnormal. The judgement of whether the behaviour is statistically frequent of infrequent is based on the mathematical principal of the normal distribution. It so happens that when measured and plotted on a graph, most human characteristics fall within  normal distribution.

The majority of characteristics would cluster around the middle of the graph. Any characteristic in this region would be considered to be normal (statistically common). A minority of characteristics would be out at the edges, or tails, making them statistically rare and therefore a  deviation from statistical norm.

Evaluation, AO3 of the Statistical Infrequency Definition:

(1) POINT:  A strength of the Statistical Infrequency definition of abnormality is that offers the prospect of clear guidelines for identifying behaviours as normal and abnormal.  EVIDENCE:  For example, the definition introduces an element of  objectivity  into the process of defining abnormality so that different mental health care workers can all view the same kind of behaviour in the same kind of way.  EVALUATION:  This is a strength because it offers a more scientific way of measuring abnormality, reducing subjectivity and therefore leading to a more accurate way of defining abnormality.

(1) POINT:  A criticism of this definition is that it suggests that anyone who differs from the ‘average’ is technically abnormal.  EVIDENCE:  For example, it doesn’t take into account the desirability of behaviours, highly intelligent people are statistically rare and are thus, according to this definition, abnormal.  EVALUATION:  This is a weakness as just because somebody falls outside of the statistically common area of the distribution curve, it doesn’t necessary mean that they are able. Therefore, suggesting that the statistical infrequency definition is not always accurate in defining abnormality.

(2) POINT:  A weakness of this definition is that it can be criticised as being  ethnocentric.  EVIDENCE:  For example, cultures differ in terms of what they consider normal behaviour. It would be very unusual for someone in Great Britain to take an extended rest period during the working day; however, in other cultures this is very common. EVALUATION:  This is a weakness because the definition cannot be applied/generalised to every culture and therefore in using this definition it may lead to some individuals being inaccurately defined as being abnormal.

(3)   POINT:  A further weakness is that the statistical approach definition requires a decision about the point at which a behaviour becomes statistically abnormal.  EVIDENCE:  For example, it is not clear at which point energetic and excitable behaviour from a child becomes a ‘problem’ behaviour requiring specialist intervention.  EVALUATION:  This is a problem as the definition is left open to criticism, such apparent arbitrariness makes it difficult to accurately define abnormality.

• Psychopathology
• Social Psychology
• Approaches To Human Behaviour
• Biopsychology
• Research Methods
• Issues & Debates
• Teacher Hub
• Terms and Conditions
• Privacy Policy
• Cookie Policy
• [email protected]
• www.psychologyhub.co.uk

We're not around right now. But you can send us an email and we'll get back to you, asap.

Start typing and press Enter to search

Cookie Policy - Terms and Conditions - Privacy Policy

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• APA Open Access

Enduring Mental Health: Prevalence and Prediction

Jonathan d. schaefer.

1 Department of Psychology and Neuroscience, Duke University

Avshalom Caspi

2 Department of Psychology and Neuroscience, Center for Genomic and Computational Biology, and Department of Psychiatry and Behavioral Sciences, Duke University, and Social, Genetic, and Developmental Psychiatry Centre, Institute of Psychiatry, Psychology, and Neuroscience, King’s College London

Daniel W. Belsky

3 Social Science Research Institute, Duke University, and Department of Medicine, Duke University School of Medicine

Honalee Harrington

4 Department of Psychology and Neuroscience, Duke University

Renate Houts

L. john horwood.

5 Department of Psychological Medicine, University of Otago

Andrea Hussong

6 Department of Psychology and Center for Developmental Science, University of North Carolina at Chapel Hill

Sandhya Ramrakha

7 Dunedin Multidisciplinary Health and Development Research Unit, Department of Psychology, University of Otago

Richie Poulton

Terrie e. moffitt.

8 Department of Psychology and Neuroscience, Center for Genomic and Computational Biology, and Department of Psychiatry and Behavioral Sciences, Duke University, and Social, Genetic, and Developmental Psychiatry Centre, Institute of Psychiatry, Psychology, and Neuroscience, King’s College London

Jonathan D. Schaefer, Terrie E. Moffitt, and Avshalom Caspi developed the study concept. Terrie E. Moffitt, Avshalom Caspi, and Richie Poulton contributed to the study design. Honalee Harrington compiled the data, and Jonathan D. Schaefer performed the data analysis and interpretation under the supervision of Avshalom Caspi, Terrie E. Moffitt, and Renate Houts. Jonathan D. Schaefer drafted the article, and Avshalom Caspi, Terrie E. Moffitt, Daniel W. Belsky, L. John Horwood, Andrea Hussong, Sandhya Ramrakha, and Richie Poulton provided critical revisions. All authors approved the final version of the manuscript for submission.

The Dunedin Multidisciplinary Health and Development Research Unit is supported by the New Zealand Health Research Council and New Zealand Ministry of Business, Innovation, and Employment. This research received support from the U. S. National Institute on Aging (NIA) Grants AG032282, AG048895, AG049789, United Kingdom Medical Research Council Grant MR/K00381X, and Economic and Social Research Council Grant ES/M010309/1. Additional support was provided by the Jacobs Foundation and the Avielle Foundation. Jonathan D. Schaefer and Daniel W. Belsky were supported by NIA Grants T32-AG000139, and P30-AG028716. Jonathan D. Schaefer was also supported by NICHD grant T32HD007376.

We thank the Dunedin Study members, their parents, teachers, and peer informants, Unit research staff, and Study founder Phil Silva. Helpful comments on earlier drafts were provided by Tim Strauman.

Associated Data

We review epidemiological evidence indicating that most people will develop a diagnosable mental disorder, suggesting that only a minority experience enduring mental health. This minority has received little empirical study, leaving the prevalence and predictors of enduring mental health unknown. We turn to the population-representative Dunedin cohort, followed from birth to midlife, to compare people never-diagnosed with mental disorder ( N = 171; 17% prevalence) to those diagnosed at 1–2 study waves, the cohort mode ( N = 409). Surprisingly, compared to this modal group, never-diagnosed Study members were not born into unusually well-to-do families, nor did their enduring mental health follow markedly sound physical health, or unusually high intelligence. Instead, they tended to have an advantageous temperament/personality style, and negligible family history of mental disorder. As adults, they report superior educational and occupational attainment, greater life satisfaction, and higher-quality relationships. Our findings draw attention to “enduring mental health” as a revealing psychological phenotype and suggest it deserves further study.

General Scientific Summary

This study reviews evidence indicating that the experience of a diagnosable mental disorder at some point during the life course is the norm, not the exception. Our results suggest that the comparatively few individuals who manage to avoid such conditions owe their extraordinary mental health to an advantageous personality style and lack of family history of disorder, but not to childhood socioeconomic privilege, superior health, or high intelligence.

This article reports an investigation of individuals who manage to live for decades without experiencing a mental disorder: the phenomenon of “enduring mental health.” It has been widely assumed that individuals who experience mental disorder are relatively rare in the population, and, conversely, that individuals whose lives remain free from mental disorder are highly prevalent, commonplace, and therefore unremarkable. This assumption is reasonable if based on the point-prevalence of mental disorder in a cross-section of the population at any single point in time. However, new lifetime data are revealing that individuals who experience mental disorder are highly prevalent in the population and as a result of this high lifetime prevalence, individuals whose lives remain free from mental disorder are, in fact, remarkably few in number. Within the past decade, estimates from an array of population-representative samples have converged to suggest that a diagnosable disturbance in emotional or behavioral functioning at some point in the life course is near-universal. This novel observation led us to ask a question missing from the discussion of mental disorders in contemporary society: If nearly everyone will eventually develop a diagnosable mental disorder, what accounts for the distinct minority of individuals who manage to avoid such conditions?

As a result of the lack of awareness that enduring mental health is so statistically unusual, it has not previously attracted scientific interest, and thus it has not been a topic of investigation as a phenotype. To our knowledge, there are no prior studies of it. The consequent knowledge gap about enduring mental health should be filled by research, because if individuals who sustain enduring mental health have special characteristics or life experiences that distinguish them from individuals with more commonplace psychiatric histories, then such discerning characteristics might become interesting new targets for prevention and treatment research. We note a potential parallel to gerontologists’ study of rare individuals with unusually enduring physical health: centenarians. Much is being learned by comparing centenarians against individuals whose aging histories are more commonplace (i.e., characterized by age-related physical disorders). Researchers comparing centenarians to normative agers aim to uncover secrets to successful aging and identify new therapeutic targets. New therapeutic targets are likewise needed in mental health, because mental disorders are the leading cause of years lost to disability worldwide ( Whiteford et al., 2013 ), and are associated with higher health care utilization, a more-than-doubled mortality rate, and a loss of life expectancy of about 10 years ( Walker, McGee, & Druss, 2015 ).

This article has two overarching aims. First, we aim to draw attention to just how common mental disorders are, and, in doing so, inform discussions surrounding etiological theories of mental disorder, societal perceptions of stigma, and prevention efforts. Second, we aim to encourage researchers to shift scientific inquiry from an exclusive focus on the etiology of mental illness toward investigation of the etiology of enduring mental wellness. Just as research on the predictors and correlates of specific mental disorders has contributed substantially to the prediction, prevention, and treatment of these conditions, so too might research on the predictors of enduring mental health provide insight into how clinicians and policymakers can promote its spread in order to reduce both societal burden and individual suffering. This article addresses the knowledge gap about enduring mental health by reporting basic descriptive information about its prevalence, predictors, and correlates. Because readers may reasonably doubt our claim that the experience of diagnosable mental disorder is near universal, the first section of this article reviews existing prevalence findings that document the high lifetime prevalence of mental disorder and the logical basis for our claim that enduring mental health warrants scientific study. The second section then presents an empirical study in which we identified members of a repeatedly assessed, longitudinal cohort who experienced enduring mental health (i.e., an absence of disorder) for close to 3 decades, and analyzed their life circumstances, personal characteristics, and family histories.

A Qualitative Review of the Prevalence of Not Having a Mental Disorder

To date, researchers who have attempted to quantify the proportion of the population that suffers from any kind of diagnosable mental health problem have used data from three sources: (a) national registries, (b) retrospective surveys, and (c) prospective cohort studies.

Lifetime prevalence estimates generated by national registry data are shown as green bars in Figure 1 . These (sex-specific) prevalence rates drawn from the Danish Civil Registration System capture the proportion of the Danish population who received treatment in a psychiatric setting between 2000 and 2012, placing the overall lifetime risk of being treated for a mental disorder at approximately 1 in 3, in this country with a national health system ( Pedersen et al., 2014 ). However, because many people with a mental disorder either do not seek treatment or do so in nonpsychiatric medical settings, these estimates can be more accurately thought of as the lower boundary of the proportion of the population who experience a mental disorder during their lives ( The WHO World Mental Health Survey Consortium, 2004 ).

Proportion of cohort members in each study with a lifetime diagnosis of one or more mental disorders (see Table 1 for Study characteristics). Error bars represent 95% confidence intervals. Green bars represent estimates drawn from Danish registry data. Blue bars represent estimates from cross-sectional epidemiological surveys. Red bars represent estimates from prospective longitudinal studies with repeated mental health assessments. The estimates shown for the Christchurch Study and Dunedin Study are based on subsets ( N = 1,041 and 988, respectively) of the full cohorts ( N = 1,265 and 1,037, respectively) who contributed data to 3 + assessment waves. Age Range = age of cohort members at first mental health assessment, presented as a single number, range, or as “mean ( SD )” where appropriate. No. of assessments = number of assessment waves in each longitudinal study; Length of follow-up = duration of longitudinal follow-up across assessments.

A second group of prevalence estimates comes from nationally representative, retrospective epidemiological surveys, such as the Epidemiological Catchment Area (ECA) Study ( Regier & Robins, 1991 ), the National Comorbidity Survey (NCS; Kessler et al., 1994 ), the National Comorbidity Survey Replication (NCS-R; Kessler et al., 2005 ), and the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC; Compton, Thomas, Stinson, & Grant, 2007 ; Hasin, Stinson, Ogburn, & Grant 2007 ). As shown by the blue bars in Figure 1 , these large surveys have reported that roughly half of all citizens will develop a diagnosable mental disorder over the course of their lives ( Kessler et al., 1994 , 2005 ). An important advantage of these studies is that, unlike national registers, they count all cases of disorder irrespective of service use. However, because such surveys are cross-sectional (i.e., rely on a single retrospective report), the lifetime prevalence estimates drawn from these data are biased downward by methodological limitations such as recall failure ( Simon & VonKorff, 1995 ). Moreover, this undercounting of disorder cases may be exacerbated by selective participation, as individuals with mental disorders—particularly severe mental disorders that result in homelessness, institutionalization, or survey refusal—are less likely to be recruited and interviewed.

Finally, a third group of mental disorder prevalence estimates comes from prospective, longitudinal studies, which interview participants repeatedly about psychiatric symptoms and then aggregate disorders across multiple time points to calculate lifetime rates. Although such studies involve fewer participants than epidemiological surveys or national registers, they also boast several advantages that contribute to significantly higher prevalence estimates ( Haeny, Littlefield, & Sher, 2014 ; Moffitt et al., 2010 ; Takayanagi et al., 2014 ). Like surveys, longitudinal studies count cases irrespective of service use. In addition, they typically employ shorter recall periods (e.g., 6–12 months) than epidemiological surveys, thereby minimizing the odds of recall failure. Finally, repeated contact with research staff in the context of a longitudinal study may directly facilitate the disclosure of psychiatric symptoms through a heightened sense of trust that accumulates over multiple interviews.

The red bars in Figure 1 display prevalence estimates drawn from five longitudinal studies. In order to be included in Figure 1 , longitudinal studies had to (a) report cumulative mental disorder lifetime prevalence estimates aggregated across multiple assessment waves, (b) administer at least 3 separate diagnostic assessments over time, and (c) assess a wide variety of conditions, including those drawn from each of the three most common disorder “families”: depressive disorders, anxiety disorders, and substance-use disorders. As shown in Figure 1 , the proportion of participants in these studies diagnosed with a mental disorder ranged from 61.1% to 85.3%—between roughly 1.3 and 1.8 times as high as corresponding estimates drawn from the NCS/NCS-R, and more than twice as high as estimates drawn from Danish registry data, with no overlap in confidence intervals. There was also variation among longitudinal studies, with higher lifetime prevalence estimates tending to come from studies with more frequent assessments and lengthier follow-up periods ( Table 1 ).

Estimates from retrospective surveys and prospective cohort studies have been criticized for assessing only common Axis I disorders, omitting conditions such as personality disorders. The impact of this limitation on estimates of the lifetime prevalence of any diagnosable mental health problem is likely fairly small, however, given the high level of comorbidity between personality and common Axis I disorders ( Hayward & Moran, 2008 ).

Viewed together, the three types of studies represented in Figure 1 converge to indicate that the proportion of the population who lives through adolescence and adulthood without experiencing a mental disorder is surprisingly small. This observation is particularly striking given that even the longitudinal prevalence estimates shown in Figure 1 likely represent an underestimate of the true prevalence of mental disorders in the population due to factors such as gaps between assessment periods and the possibility of selective attrition. The experience of enduring mental health, therefore, may be substantially rarer than was previously thought. This realization prompted us to ask the following questions: Who, exactly, are these individuals who lead lives untouched by mental disorders? What sorts of environments did they grow up in? And does enduring mental health matter? That is, is a life free from mental disorders associated with more desirable life outcomes (i.e., greater attainment, increased life satisfaction, and higher-quality relationships)?

Empirical Study of Individuals With Enduring Mental Health

The second section of this article reports an analysis of early-life demographic, family environment, physical health, cognitive, temperamental/personality, and family history characteristics of individuals who have never been diagnosed with a mental disorder during the course of the Dunedin Longitudinal Study. In the absence of prior research or theory on enduring mental health, we selected from our data set measures available to us that have the best published evidence base as important risk factors for mental disorder. We have previously found that several of these measures correlate with scores on the ‘p-factor,’ which represents an individual’s propensity to develop any and all forms of common psychopathologies ( Caspi et al., 2014 ). We reasoned that individuals with enduring mental health (and, consequently, very low scores on the p-factor) thus ought to be exceptionally well-advantaged on these measures. We hypothesized, for example, that they would have well-to-do socioeconomic origins, exceptionally positive parent–child relations, robust physical health, high intelligence, adaptive personality styles from childhood, and nil histories of psychiatric illness in their families. To add to our descriptive data about individuals with enduring mental health, we also tested the hypothesis that they would enjoy exceptionally positive life outcomes (in the domains of educational attainment, socioeconomic status, life satisfaction, and the quality of their most recent romantic relations), as assessed at the end of our study observation period.

The Dunedin Study assessed Study members for a variety of common mental disorders beginning when they were 11 years of age, and repeated these assessments every few years up until the most recent wave, when Study members were all age 38. Because the predictors of most forms of severe and/or chronic mental disorders are well established, we chose to focus our analyses on the predictors and outcomes of extraordinary mental health—that is, what distinguishes Study members who were never diagnosed with a mental disorder (hereafter referred to as the “enduring-mental-health” group) from those who experienced a mental health history that could fairly be characterized as typical (i.e., at the mode) for the Dunedin cohort.

Participants are members of the Dunedin Multidisciplinary Health and Development Study (DMHDS), a 4-decade, longitudinal investigation of health and behavior in a complete birth cohort. Study members ( N = 1,037; 91% of eligible births; 52% male) were all individuals born between April, 1972 and March, 1973 in Dunedin, New Zealand who were eligible for the longitudinal study based on residence in the province at age 3, and who participated in the first follow-up assessment at age 3. The cohort represented the full range of SES in the general population of New Zealand’s South Island. On adult health, the cohort matches the NZ National Health & Nutrition Survey (e.g., body mass index, smoking, general practitioner visits; Poulton et al., 2015 ). The cohort is primarily white; fewer than 7% self-identify as having partial non-Caucasian ancestry, matching the South Island. Assessments were carried out at birth and at ages 3, 5, 7, 9, 11, 13, 15, 18, 21, 26, 32, and, most recently, 38 years, when 95% of the 1,007 Study Members still alive took part. At each assessment wave, each Study member is brought to the Dunedin research unit for a full day of interviews and examinations. This article examines Study members who were assessed for mental disorders at ages 11, 13, 15, 18, 21, 26, 32, and 38 years of age. The Otago Ethics Committee approved each phase of the Study and informed consent was obtained from all Study members.

Assessment of Mental Disorders

Mental disorders were ascertained in the Dunedin Study longitudinally using a periodic sampling strategy: Every 2 to 6 years, Study members were interviewed about past-year symptoms in a private in-person interview at the research unit by trained interviewers with tertiary qualifications and clinical experience in a mental health-related field such as family medicine, clinical psychology, or psychiatric social work (i.e., not lay interviewers). Interviewers used the Diagnostic Interview Schedule for Children (DIS-C) at the younger ages (11–15 years) and the Diagnostic Interview Schedule (DIS) at the older ages (18–38 years). At each assessment, interviewers were kept blind to Study members’ previous data, including mental health status. At ages 11, 13, and 15, diagnoses were made according to the then-current Diagnostic and Statistical Manual of Mental Disorders (3rd ed.; American Psychiatric Association [APA], 1980 ) and grouped for this article into a single wave reflecting the presence or absence of a juvenile mental disorder. At ages 18 and 21, diagnoses were made according to the Diagnostic and Statistical Manual of Mental Disorders (3rd ed, rev.; DSM–III–R ; APA, 1987 ) and at ages 26, 32, and 38 diagnoses were made according to the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM–IV ; APA, 1994 ). This method led to 6 waves in total representing ages 11–15, 18, 21, 26, 32, and 38. In addition to symptom criteria, diagnosis required impairment ratings ≥2 on a scale from 1 ( some impairment ) to 5 ( severe impairment ). Each disorder was diagnosed regardless of the presence of other disorders. Variable construction details, reliability and validity, and evidence of life impairment for diagnoses have been reported previously. Of the original 1,037 Study members, we included 988 (95.3%) Study members who had participated in at least half of the six mental health assessment waves from ages 11 to 38. Of these Study members, 849 (85.9%) contributed data to all 6 waves, 88 (8.9%) contributed data to 5 waves, 32 (3.2%) contributed data to 4 waves, and 19 (1.9%) contributed data to 3 waves.

Candidate Childhood Predictors

To test what distinguishes Study members who experienced enduring mental health from their peers, we report on 13 different predictors, selected because they are thought to be associated with risk of developing a mental disorder: parental socioeconomic status, positive family climate, negative discipline, maltreatment, parental loss, perinatal complications, childhood health, preschool IQ, middle childhood IQ, emotional difficulties, social isolation, self-control, and family psychiatric history. These measures are described in Table 1 in the online supplementary material.

Midlife Outcomes

Educational attainment.

Educational attainment at age 38 was measured on a four-point scale relevant to the New Zealand educational system: 0 = no secondary school qualifications , 1 = school certificate , 2 = high school graduate or equivalent , 3 = bachelor’s degree or higher .

Socioeconomic attainment (SES)

At age 38, Study members were asked about their current or most recent occupation. The SES of the study members was measured on a 6-point scale that assessed self-reported occupational status and allocates each occupation to 1 of 6 categories (1 = unskilled laborer , 6 = professional ) on the basis of the educational levels and income associated with that occupation in data from the New Zealand census. Homemakers and those not working were prorated based on their educational status according to criteria included in the New Zealand Socioeconomic Index ( Milne, 2012 ).

Life satisfaction

At age 38, Study members completed the 5-item Satisfaction With Life Scale (e.g., “In most ways my life is close to ideal”; “So far I have gotten the important things I want in life”; Pavot & Diener, 1993 ).

Relationship quality

At age 38, Study members who reported being in a relationship for at least one month during the past year reported on a 28-item scale about their current or most recent relationship, covering relationship characteristics such as shared activities and interests, the balance of power, respect and fairness, emotional intimacy and trust, and open communication. Each of these items was coded on a 3-point scale (0 = Almost never , 1 = Sometimes , 2 = Almost always ). We summed these ratings across items to create a composite measure reflecting overall relationship quality (α = .93). Of the 988 Study members who had participated in at least half of the six mental health assessments from ages 11 to 38, 841 (85.1%) reported a current or recent relationship at age 38.

Defining Mental Health Histories Over the First Half of the Life Course

Figure 2a displays the number of waves (from 0 to 6) in which Study members met criteria for one or more mental disorders. On average, cohort members met criteria for a mental disorder on 2.3 of the six assessment waves, but there was a great deal of variation. The most common mental health history in the cohort appeared to be one characterized by a relatively brief, episodic course of disorder, in which Study members met diagnostic criteria for a disorder at only 1 or 2 assessment waves ( N = 409). We also included in this group 9 Study members who were not diagnosed with a mental disorder by Dunedin Study staff, but reported receiving a psychiatric diagnosis while using mental health services in the gaps between assessment waves. 1 Study members who experienced enduring mental health (i.e., met diagnostic criteria at 0 waves), in contrast, were a distinct minority, comprising only 17.3% of the cohort ( N = 171). 2 The remainder of the cohort were Study members who had met criteria for one or more mental disorder diagnoses at 3 + waves ( N = 408). It is important to note that Study members were not classified as having enduring mental health simply because they participated in fewer waves: On average, Study members with enduring mental health had complete data on 5.7 (out of 6) waves, whereas Study members who met diagnostic criteria at 1–2 waves had complete data on 5.8 waves, and Study members who met diagnostic criteria at 3–6 waves had complete data on 5.8 waves.

Mental disorder diagnoses in the Dunedin Cohort ( N = 988). Panel A: Number of waves in which Dunedin Study members met criteria for a DSM diagnosis. The 6 waves represent ages 11–15, 18, 21, 26, 32, and 38. The red bar represents Study members with enduring mental health (those diagnosed at 0 waves). The light blue bars represent Study members with typical mental health histories (those diagnosed at 1–2 waves). The dark blue bars represent Study members diagnosed at 3 + waves. Panel B: Distribution of DSM diagnoses across assessment waves. Each thin horizontal line represents an individual Study member’s mental health history. Blue indicates that the Study member met criteria for a past-year DSM -defined psychiatric disorder during this assessment. Red indicates that the Study member did not meet criteria for a past-year DSM -defined psychiatric disorder during this assessment. Panel B shows that the largest proportion of Study members met diagnostic criteria at 1–2 waves, but that neither these individuals nor those diagnosed at 3 + waves were characterized by any particular developmental pattern (e.g., adolescent-limited course of disorder or late-onset forms of disorder).

Figure 2b displays the temporal pattern of psychiatric diagnoses across the life course of the cohort, from ages 11 to 38 years. The figure shows that the diagnosed groups were not dominated by any particular developmental pattern.

Table 2 displays indicators of disorder type, age-of-onset, and severity for individuals as a function of mental health-history group. Relative to the Study members diagnosed at 3 + waves, those with typical mental health histories (i.e., diagnosed at 1–2 waves) presented with a narrower set of disorders (primarily depression, anxiety, and substance dependence), an older age of onset, less comorbidity, and lower scores on a general factor of psychopathology ( Caspi et al., 2014 ).

Informant Reports: To What Extent Do They Confirm the Enduring Mental Health of Never-Diagnosed Study Members?

Given the high lifetime prevalence of mental disorders, it is reasonable to wonder whether Study members classified as experiencing “enduring mental health” are, in fact, simply those with a tendency to down-play or deny genuine past-year psychiatric symptoms during clinical interviews. As an additional “check” for evidence of mental disorder, we reviewed informant reports to see if these Study members showed any outwardly perceivable signs of common mental disorders. At ages 18, 21, 26, 32, and 38, we asked Study members to nominate someone who knew them well (e.g., best friends, partners, or other family members). These informants were mailed questionnaires which asked them “ To the best of your knowledge , did ________ have any of these problems over the last 12 months ?” Items included “Feels depressed, miserable, sad, or unhappy,” “Has unreasonable worries or fears,” “Has alcohol problems,” “Marijuana or other drug problems,” and (at ages 26, 32, and 38), “Talks about suicide.” Informants were asked to rate these items on a 3-point scale (0 = Not a problem , 1 = Bit of a problem , 2 = Yes, a problem ). In analyzing these data, we took a conservative approach, treating a rating of “2” by any informant during any assessment wave as evidence of symptomatic behavior. Informant report data were available for 987 (99.9%) of the 988 Study members reported here.

Although informant reports provide a useful complement to self-reported symptoms, endorsements of symptomatic behaviors must be interpreted with caution. The informant questionnaire was not designed to correspond directly with DSM diagnoses or diagnostic criteria. Therefore, many informants may have been inclined to endorse Study member “problems” (e.g., “feels depressed, miserable, sad, or unhappy”) even when these issues were not of sufficient severity to meet diagnostic criteria for a DSM -defined mental disorder (e.g., major depression).

As shown in the bottom panel of Table 2 , informant reports largely confirmed the absence of mental health problems among Study members with enduring mental health. From ages 18 to 38, only 36 (21.1%) Study members with enduring mental health had an informant report that they showed evidence of problems with depression, unreasonable fears, alcohol, drugs, or had talked about suicide (compared to 38.9% and 63.4% of Study members diagnosed at 1–2 and 3 + waves, respectively). According to informants, the most common problem for these Study members was feeling depressed (15.8%), with only a small handful of informants reporting problems with unreasonable fears (8.2%), alcohol (1.8%), drugs (1.8%), or talking about suicide (0.6%).

What Distinguishes Study Members Who Experienced Enduring Mental Health From Those Who Experienced “Typical” Mental Health Histories?

It has been repeatedly demonstrated that individuals with severe, persistent, or recurrent mental disorders differ from individuals without such disorders in multiple ways. This well-established finding was confirmed in our study: Table 3 shows that Study members diagnosed at 3 + waves had more childhood risk factors across each domain compared to both Study members with enduring mental health and Study members diagnosed at 1–2 waves.

The key comparison in this article, however, is between Study members who were never diagnosed with a mental disorder, and those who experienced a mental health history that resembles the histories of the majority of other Study members (i.e., the “1-2 wave” group). By comparing Study members with enduring mental health to those with more typical mental health histories across candidate predictor variables hypothesized to discriminate between them, we can distinguish factors predictive of enduring mental health from those that simply predict the absence of a severe, persistent, or recurrent disorder.

Although we had expected to find that Study members with enduring mental health were significantly advantaged across all 13 of our candidate predictors relative to Study members with typical histories, this hypothesis received only mixed support. First, we found that Study members with enduring mental health were surprisingly similar to Study members who met diagnostic criteria at 1–2 waves in terms of parental socioeconomic status, childhood physical health, and childhood cognitive ability (the observed distribution of mean predictor variable scores across the number of waves in which Study members received a diagnosis can be seen in Figure 1 in the online supplemental materials). Second, although we found some evidence to suggest that Study members in the two groups differed in their upbringing, analyses using these variables returned mixed results. Third, Study members with enduring mental health showed statistically significant advantages in childhood temperament/personality relative to Study members diagnosed at 1–2 waves, including fewer emotional difficulties, less social isolation, and superior self-control. Finally, Study members with enduring mental health also had significantly fewer first- and second-degree relatives who showed signs of mental disorder ( Table 3 ).

Thus far, we have characterized Study members’ mental health histories as a function of persistence or recurrence; that is, by the number of waves in our longitudinal study during which they received a diagnosis. We found that a mental health history in which the Study member met diagnostic criteria for a mental disorder at 1 or 2 waves was the most common pattern. Another way to characterize mental health histories, however, is as a function of comorbidity; that is, by the number of different types of disorder categories or “families” represented in Study members’ accumulated diagnostic histories. To ensure that the results in Table 3 were not dependent on the particular way in which we classified the severity of Study members’ mental health histories, we repeated these analyses using a classification scheme based on comorbidity rather than recurrence or persistence. As shown in Figure 2 in the online supplemental material, the same group of 171 Study members received no diagnosis throughout the course of the study, and were thus considered to experience enduring mental health by virtue of having no psychiatric comorbidity. Our new comparison group, however, consisted of 540 Study members who were diagnosed with disorders from 1–2 different diagnostic families, the cohort “comorbidity mode.” Similarly, our most severe group consisted of the 277 remaining Study members with mental health histories characterized by unusually high comorbidity, or diagnoses from 3 + different diagnostic families. Our substantive conclusions regarding the most and least effective predictors of enduring mental health remained almost entirely unchanged under this alternate classification scheme (see Table 2 in the online supplemental material). This stability is largely attributable to the fact that comorbidity and number of waves with disorder are highly correlated ( r = .80, p < .001), as are most indicators of disorder severity. The most common mental health history in our data thus appears to be characterized not only by disorders of relatively short duration but also those that are diagnostically “pure” (that is, with limited lifetime comorbidity).

Is Enduring Mental Health Associated With More Desirable Life Outcomes (i.e., Greater Educational and Occupational Attainment, Increased Life Satisfaction, and Higher Quality Relationships)?

As shown in Figure 3 , despite their comparable socioeconomic background, Study members with enduring mental health achieved higher levels of educational and socioeconomic attainment by age 38 than Study members who had experienced 1–2 waves of disorder. Study members with enduring mental health also expressed higher levels of life satisfaction when interviewed at age 38 than Study members diagnosed at 1–2 waves. Interestingly, although Study members with enduring mental health were just as likely to report being in a relationship at age 38 as Study members diagnosed at 1–2 waves (91.1% vs. 92.8%, respectively; χ 2 = 0.40, p = .528), they rated these relationships as being of higher quality (see Table 3, online supplemental material, for more detail).

Comparison of midlife outcomes for Dunedin cohort members in the 0 wave versus 1–2 wave mental health history groups. Error bars represent 95% confidence intervals. All outcome variables were standardized on the full cohort to a mean of 0 (representing the mean of the full cohort) and a standard deviation of 1. The means for the persistently diagnosed group are not shown here, but can be found in Table 3 in the online supplemental materials. Asterisks represent the statistical significance of the difference between groups, adjusted for sex. * p < .05.

Far from being the aberrant experience of a small, stigmatized subgroup, data from both the Dunedin Study and other longitudinal studies suggest that experiencing a diagnosable mental disorder at some point during the life course is the norm , not the exception. In our cohort, whose members have been repeatedly assessed for common mental disorders by trained professionals over a span of close to three decades, only 17% of repeatedly assessed Study members managed to reach midlife (age 38) without experiencing the psychiatric symptoms and resulting functional impairment necessary to meet criteria for the diagnosis of a mental disorder.

To some, the proportion of Dunedin Study members diagnosed with at least one mental disorder may seem unusually high, raising concerns about the representativeness of our sample. However, we have shown elsewhere that the past-year prevalence rates of mental disorders in the Dunedin cohort are similar to prevalence rates in nationwide surveys of the United States and of New Zealand. This observation indicates that the higher Axis-I-disorder lifetime prevalence rate in our study is due primarily to the advantage of our prospective assessment method rather than to an overabundance of mental disorder in New Zealand, or in our cohort ( Moffitt et al., 2010 ). Similarly, although Axis-I-disorder lifetime prevalence estimates drawn from the Dunedin Study and Christchurch Study are modestly higher than those of other longitudinal studies with similar methodologies ( Figure 1 ), this discrepancy is likely due to differences in study design. To our knowledge, the Dunedin Study is one of the only prospective, longitudinal studies with nearly three decades of mental health assessments that stretch from late childhood (when the earliest cases of most mental disorders first onset) through adolescence and young adulthood (the time of peak onset for many of these same disorders) and into midlife. The Christchurch Study captures a similar period of development with the additional advantage of mental health assessments that cover the full time period between assessments (rather than just counting symptoms experienced within the past 12 months). We anticipate that Axis-I-disorder lifetime prevalence estimates drawn from similar studies of younger cohorts (e.g., Copeland, Shanahan, Costello, & Angold, 2011 ) will eventually mirror (or exceed) the values obtained from these New Zealand studies as these cohorts are followed forward.

There is an extensive literature linking childhood attributes and experiences to later mental disorder. Usually, it is implicitly assumed that individuals without the disorder (“controls”) represent “normality,” whereas those who do develop the disorder (“cases”) represent “abnormality.” However, data reported here indicate that the statistically “typical” Study member is a person with at least some transient history of diagnosable psychopathology. Consequently, we sought to identify early life variables that differentiated between those with “typical” mental health histories and those with extraordinary histories marked by no episodes of diagnosable mental disorder whatsoever (at least, as far as we know). The utility provided by this type of comparison is that it helps to distinguish between variables that predict enduring mental health and those that predict the onset of severe mental disorders (but perhaps fail to distinguish between individuals on the opposite end of the spectrum). Our finding that relatively few early life measures seem to predict above-average mental health, whereas many predict very poor mental health is perhaps not surprising. Indeed, identifying measures that do distinguish between Study members with enduring mental health and those with typical mental health histories should be significantly more difficult than identifying measures that predict the more severely ill cases given that milder, more transient episodes of disorder are more likely to be attributable to situational, stochastic factors rather than enduring vulnerabilities.

Given the remarkably low prevalence of enduring mental health in the Dunedin cohort, we expected Study members with enduring mental health to come from backgrounds virtually free of each of our 13 well-established predictors. This expectation was strongly supported when we compared Study members with enduring mental health to Study members diagnosed at 3 + waves, but unsupported when comparing Study members with enduring mental health to Study members diagnosed at 1–2 assessment waves ( Table 3 ).

We identified only two childhood factors that clearly differentiated between Study members with enduring mental health and those diagnosed only at 1 or 2 waves: (a) a suite of advantageous personality traits and (b) a relative absence of family psychiatric history. Consistent with research that names a neurotic personality style as a risk factor for multiple different mental disorders ( Kendler, Gatz, Gardner, & Pedersen, 2006 ; Lahey, 2009 ), we found that Study members who showed little evidence of strong negative emotions in childhood were more likely to experience enduring mental health into their late 30s. Similarly, consistent with research that names abundant social support and sociability as “buffers” against stress (e.g., Ozbay et al., 2007 ), we also found that Study members with enduring mental health were significantly less socially isolated in childhood than peers with typical histories (or, alternatively, these exceptionally well-adjusted children were more attractive to peers, and thus acquired more childhood friends). In addition, we found that Study members with enduring mental health showed significantly higher levels of childhood self-control, in line with previous reports from this cohort demonstrating that higher self-control in childhood predicts other advantageous adult outcomes such as superior physical health, fewer financial problems, less criminal offending, and lower risk of substance dependence ( Israel et al., 2014 ; Moffitt et al., 2011 ). Finally, consistent with research indicating substantial familial aggregation of common psychiatric and substance-use disorders ( Kendler, Davis, & Kessler, 1997 ), we found that Study members who experienced enduring mental health had fewer first- and second-degree relatives with mental health issues relative to Study members diagnosed at 1–2 waves.

Our analyses of family factors returned mixed results. We found evidence to indicate that, relative to Study members with typical mental health histories, those with enduring mental health experienced a family environment characterized by less negative discipline and a reduced likelihood of parental loss. Surprisingly, however, the remainder of our childhood predictors did not seem to differ between the two groups. For example, we found that individuals with enduring mental health were not more socioeconomically advantaged than those with typical histories, despite evidence linking low childhood socioeconomic status to multiple mental disorders ( Reiss, 2013 ). In addition, Study members with enduring mental health showed no evidence of fewer perinatal complications or superior physical health in childhood, despite evidence linking perinatal complications and poor health in childhood to multiple mental disorders ( Buka & Fan, 1999 ; Foley, Thacker, Aggen, Neale, & Kendler, 2001 ; Merikangas et al., 2015 ). And finally, Study members with enduring mental health were not found to possess higher childhood intelligence than Study members diagnosed at 1–2 waves, even though multiple studies have confirmed low IQ as a risk factor for a wide array of psychiatric conditions ( Batty, Mortensen, & Osler, 2005 ; Gale et al., 2008 ; Koenen et al., 2009 ). These observations suggest that although childhood poverty, compromised physical health, and low cognitive ability are robust predictors of persistent mental disorder, their absence is unlikely to guarantee enduring mental health.

The predictive strength of our temperament/personality measures is perhaps unsurprising, given that they capture, in part, behaviors that could be viewed as juvenile manifestations of adult disorders (e.g., our measure of childhood emotional difficulties captures behaviors like frequent worrying, and often appearing sad or tearful). Nevertheless, our finding that these measures are capable of predicting which Study members will reach their late 30s without ever experiencing a diagnosable disorder suggests that the path to enduring mental health begins early in development, as is the case with many mental disorders ( Kessler et al., 2005 ; Kim-Cohen et al., 2003 ).

The present study is characterized by several limitations. First, although findings about the low prevalence of enduring mental health have appeared across studies, our findings regarding the correlates of enduring mental health were drawn from a single, largely Caucasian, New Zealand cohort born in the 1970s, and thus may not generalize to other populations. Second, assessment of mental disorder in the Dunedin cohort is both left- and right-hand censored, which means we cannot count episodes of disorder that occurred prior to age 11, or future cases that may onset after our most recent assessment at age 38. Third, gaps between the Dunedin Study’s 12-month assessment windows did not allow us to count individuals who experienced an episode of disorder between windows. Although we were able to use life history calendar interviews to reclassify 9 Study members who were not diagnosed by Study staff but reported being diagnosed and treated during these gaps into the “1–2 wave” group, the number of cohort members we classified incorrectly because their only episodes of disorder occurred between Study windows and went untreated is unknown. However, it is worth noting that the Dunedin Study’s Axis-I-disorder lifetime prevalence estimate is very similar to the Axis-I-disorder lifetime prevalence estimate drawn from the Christchurch Study ( Figure 1 ), which asks Study members at each assessment to extend their recall of psychiatric symptoms back to the previous assessment (thus avoiding gaps in assessment windows). This observation suggests that the number of Dunedin Study members who did experience a mental disorder but were “missed” by our eight past-year assessments is likely to be relatively small.

Replication of this study is needed. However, the study of enduring mental health poses a challenge for researchers, since classifying individuals as having experienced “enduring mental health” on the basis of a single clinical interview assessing lifetime psychiatric symptoms may result in substantial misclassification. One possibility suggested by our results is to further refine phenotyping by screening this group to also be free of a family history of psychiatric disorder.

The comparative rarity of the enduring-mental health phenotype has implications for etiological research into mental disorders. Studies of individuals with enduring mental health can complement studies of mental disorders in much the same way studies of centenarians complement studies of age-related disease (e.g., Galioto et al., 2008 ; Sebastiani & Perls, 2012 ). One way is by identifying targets for prevention. For example, our study suggests the hypothesis that interventions to promote children’s development of self-control skills might prevent subsequent mental disorder. Nonetheless, a limitation of the Dunedin Study is that it was not originally designed to study predictors of enduring mental health, because no one anticipated that it would be so rare as to be an interesting phenotype. As a result, our investigation was constrained by our set of pre-existing early life risk factors for mental disorder, suggesting that studies with richer sets of early-life, mental-health-promoting factors are needed.

Perhaps unsurprisingly, ours is not the first study to attempt to identify a “completely psychiatrically healthy” group of people. Indeed, control groups consisting of individuals who were screened to be free from any history of either psychiatric diagnosis or treatment were commonly used in early studies of psychiatric genetics, particularly those examining familial aggregation (e.g., Coryell & Zimmerman, 1988 ; Weissman et al., 1984 ). The extent to which these earlier studies were successful in screening out all individuals who may have at one point met criteria for a psychiatric diagnosis, however, is unclear, especially given that the pitfalls of retrospective psychiatric assessments did not become clear until a few years later.

Our findings add weight to the suggestion that research psychologists and psychiatrists should be cautious whenever they attempt to define and assemble a “healthy control group,” particularly when participants are categorized solely on the basis of a single retrospective assessment of lifetime psychiatric symptoms (see Streiner, Patten, Anthony, & Cairney, 2009 , for a thoughtful review of this issue). Because of the extremely high rates of lifetime disorder, it is likely that any “control group” defined without the use of repeated assessments will contain (a) participants with enduring mental health who have never met criteria for the disorder of interest nor any psychiatric comorbidities, (b) participants who have never met criteria for the disorder of interest, but who have met (or currently do meet) criteria for psychiatric comorbidities, and (c) participants who do not currently meet criteria for the disorder of interest, but did meet criteria in the past and have since forgotten or reframed this experience. The inadvertent inclusion of group (c) into the larger control group could lead to an attenuation of observed case-control differences, potentially reducing power to detect real effects. Conversely, a more stringent assessment process (e.g., repeated assessments of psychiatric status over time), could increase statistical power by bolstering researchers’ ability to correctly categorize study participants.

A final, intriguing question is whether enduring mental health is associated with exceptional psychological “well-being,” in addition to minimal psychological distress. Research in the fields of positive psychiatry and psychology indicates that measures of “mental health” and “mental illness” are at best moderately correlated ( Keyes, 2005 ), and that true well-being or “flourishing” (i.e., feeling good about and functioning well in life) is more than merely the absence of a diagnosable disorder ( Jeste, Palmer, Rettew, & Boardman, 2015 ; Keyes, 2002 ; Seligman & Csikszentmihalyi, 2000 ). Our data suggest that Study members with enduring mental health (as defined here) share many similarities with individuals who are described as “flourishing” in other studies, including superior adult functioning (as measured by midlife educational and occupational attainment) as well as greater life satisfaction and higher-quality relationships. This overlap suggests the hypothesis that the absence of disorder may facilitate the acquisition of other desirable psychosocial traits and outcomes across the life course. Nevertheless, it is worth noting that our never-diagnosed Study members were not universally satisfied with life—indeed, approximately one quarter (22.5%) scored below the cohort mean on our measure of life satisfaction. This observation indicates that “enduring mental health” and “flourishing” should not be used interchangeably, and suggests that additional research is needed to clarify the nature of the relationship between these two constructs.

In conclusion, the observations that mental disorder affects the overwhelming majority of persons at some point in life and that its course is often transient suggest a need to alter our conception of what it means to be mentally ill. For many, an episode of mental disorder is like influenza, bronchitis, anemia, kidney stones, or a fractured bone—these conditions are highly prevalent, sufferers experience impaired functioning in social and occupational roles, and many seek medical care, but most recover. Put another way, such research affirms that discussions of “abnormal psychology” should recognize that “normality” refers to the absence of a diagnosable disturbance in emotional or behavioral functioning at the present time— not across the life course. It is our hope that increased public recognition of this fact will reduce the stigma experienced by individuals diagnosed with a mental disorder, perhaps leading to higher rates of treatment uptake as well as better clinical outcomes.

Supplementary Material

1 Because it is possible that past-year reports separated by 1 to 5 years miss episodes of mental disorder occurring only in gaps between assessments, we reviewed life-history calendar interviews of Study members to ascertain indicators of mental disorder occurring in these gaps, including inpatient treatment, outpatient treatment, or spells taking prescribed psychiatric medication (indicators that are salient and recalled more reliably than individual symptoms). Life-history calendar data indicated that all but 9 Study members who experienced a disorder consequential enough to be associated with treatment (many of whom had a brief postnatal depression) were detected in our net of past-year diagnoses made at ages 11 to 38.

2 Four of these 171 Study members met symptom criteria for a mental disorder at some point during the Study, but rated their impairment as a 1 out of 5, thus avoiding a diagnosis.

• Angst J., Paksarian D., Cui L., Merikangas K. R., Hengartner M. P., Ajdacic-Gross V., & Rössler W. (2015). The epidemiology of common mental disorders from age 20 to 50: Results from the prospective Zurich Cohort Study . Epidemiology and Psychiatric Sciences, FirstView , 1–9. 10.1017/S204579601500027X [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• American Psychiatric Association (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. [ Google Scholar ]
• American Psychiatric Association (1987). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. [ Google Scholar ]
• American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. [ Google Scholar ]
• Batty G. D., Mortensen E. L., & Osler M. (2005). Childhood IQ in relation to later psychiatric disorder: Evidence from a Danish birth cohort study . The British Journal of Psychiatry , 187 , 180–181. 10.1192/bjp.187.2.180 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Belsky J., Jaffee S., Hsieh K.-H., & Silva P. A. (2001). Child-rearing antecedents of intergenerational relations in young adulthood: A prospective study . Developmental Psychology , 37 , 801–813. 10.1037/0012-1649.37.6.801 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Buka S. L., & Fan A. P. (1999). Association of prenatal and perinatal complications with subsequent bipolar disorder and schizophrenia . Schizophrenia Research , 39 , 113–119. 10.1016/S0920-9964(99)00109-7 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Caspi A., Houts R. M., Belsky D. W., Goldman-Mellor S. J., Harrington H., Israel S., et al.Moffitt T. E. (2014). The p factor: One general psychopathology factor in the structure of psychiatric disorders? Clinical Psychological Science , 2 , 119–137. 10.1177/2167702613497473 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Caspi A., McClay J., Moffitt T. E., Mill J., Martin J., Craig I. W., et al.Poulton R. (2002, August 2). Role of genotype in the cycle of violence in maltreated children . Science , 297 , 851–854. 10.1126/science.1072290 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Compton W. M., Thomas Y. F., Stinson F. S., & Grant B. F. (2007). Prevalence, correlates, disability, and comorbidity of DSM–IV drug abuse and dependence in the United States: Results from the national epidemiologic survey on alcohol and related conditions . Archives of General Psychiatry , 64 , 566–576. 10.1001/archpsyc.64.5.566 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Copeland W., Shanahan L., Costello E. J., & Angold A. (2011). Cumulative prevalence of psychiatric disorders by young adulthood: A prospective cohort analysis from the Great Smoky Mountains Study . Journal of the American Academy of Child and Adolescent Psychiatry , 50 , 252–261. 10.1016/j.jaac.2010.12.014 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Coryell W., & Zimmerman M. (1988). The heritability of schizophrenia and schizoaffective disorder. A family study . Archives of General Psychiatry , 45 , 323–327. 10.1001/archpsyc.1988.01800280033005 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Dunn L. M. (1965). Expanded manual for the Peabody Picture Vocabulary Test . Minneapolis, MN: American Guidance Service. [ Google Scholar ]
• Elley W. B., & Irving J. C. (1976). Revised Socio-Economic Index for New Zealand . New Zealand Journal of Educational Studies , 11 , 25–36. [ Google Scholar ]
• Farmer R. F., Kosty D. B., Seeley J. R., Olino T. M., & Lewinsohn P. M. (2013). Aggregation of lifetime axis I psychiatric disorders through age 30: Incidence, predictors, and associated psychosocial outcomes . Journal of Abnormal Psychology , 122 , 573–586. 10.1037/a0031429 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Foley D. L., Thacker L. R. II, Aggen S. H., Neale M. C., & Kendler K. S. (2001). Pregnancy and perinatal complications associated with risks for common psychiatric disorders in a population-based sample of female twins . American Journal of Medical Genetics , 105 , 426–431. 10.1002/ajmg.1402 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Gale C. R., Deary I. J., Boyle S. H., Barefoot J., Mortensen L. H., & Batty G. D. (2008). Cognitive ability in early adulthood and risk of 5 specific psychiatric disorders in middle age: The Vietnam Experience Study . Archives of General Psychiatry , 65 , 1410–1418. 10.1001/archpsyc.65.12.1410 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Galioto A., Dominguez L. J., Pineo A., Ferlisi A., Putignano E., Belvedere M., et al.Barbagallo M. (2008). Cardiovascular risk factors in centenarians . Experimental Gerontology , 43 , 106–113. 10.1016/j.exger.2007.06.009 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Haeny A. M., Littlefield A. K., & Sher K. J. (2014). Repeated diagnoses of lifetime alcohol use disorders in a prospective study: Insights into the extent and nature of the reliability and validity problem . Alcoholism: Clinical and Experimental Research , 38 , 489–500. 10.1111/acer.12237 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hasin D. S., Stinson F. S., Ogburn E., & Grant B. F. (2007). Prevalence, correlates, disability, and comorbidity of DSM–IV alcohol abuse and dependence in the United States: Results from the National Epidemiologic Survey on Alcohol and Related Conditions . Archives of General Psychiatry , 64 , 830–842. 10.1001/archpsyc.64.7.830 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hayward M., & Moran P. (2008). Comorbidity of personality disorders and mental illnesses . Psychiatry , 7 , 102–104. 10.1016/j.mppsy.2008.01.010 [ CrossRef ] [ Google Scholar ]
• Israel S., Moffitt T. E., Belsky D. W., Hancox R. J., Poulton R., Roberts B., et al.Caspi A. (2014). Translating personality psychology to help personalize preventive medicine for young adult patients . Journal of Personality and Social Psychology , 106 , 484–498. 10.1037/a0035687 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Jaffee S. R., Moffitt T. E., Caspi A., Fombonne E., Poulton R., & Martin J. (2002). Differences in early childhood risk factors for juvenile-onset and adult-onset depression . Archives of General Psychiatry , 59 , 215–222. 10.1001/archpsyc.59.3.215 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Jeste D. V., Palmer B. W., Rettew D. C., & Boardman S. (2015). Positive psychiatry: Its time has come . The Journal of Clinical Psychiatry , 76 , 675–683. 10.4088/JCP.14nr09599 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kendler K. S., Davis C. G., & Kessler R. C. (1997). The familial aggregation of common psychiatric and substance use disorders in the National Comorbidity Survey: A family history study . The British Journal of Psychiatry , 170 , 541–548. 10.1192/bjp.170.6.541 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kendler K. S., Gatz M., Gardner C. O., & Pedersen N. L. (2006). Personality and major depression: A Swedish longitudinal, population-based twin study . Archives of General Psychiatry , 63 , 1113–1120. 10.1001/archpsyc.63.10.1113 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kessler R. C., Berglund P., Demler O., Jin R., Merikangas K. R., & Walters E. E. (2005). Lifetime prevalence and age-of-onset distributions of DSM–IV disorders in the National Comorbidity Survey Replication . Archives of General Psychiatry , 62 , 593–602. 10.1001/archpsyc.62.6.593 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kessler R. C., McGonagle K. A., Zhao S., Nelson C. B., Hughes M., Eshleman S., et al.Kendler K. S. (1994). Lifetime and 12-month prevalence of DSM–III–R psychiatric disorders in the United States. Results from the National Comorbidity Survey . Archives of General Psychiatry , 51 , 8–19. 10.1001/archpsyc.1994.03950010008002 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Keyes C. L. M. (2002). The mental health continuum: From languishing to flourishing in life . Journal of Health and Social Behavior , 43 , 207–222. 10.2307/3090197 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Keyes C. L. M. (2005). Mental illness and/or mental health? Investigating axioms of the complete state model of health . Journal of Consulting and Clinical Psychology , 73 , 539–548. 10.1037/0022-006X.73.3.539 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Kim-Cohen J., Caspi A., Moffitt T. E., Harrington H., Milne B. J., & Poulton R. (2003). Prior juvenile diagnoses in adults with mental disorder: Developmental follow-back of a prospective-longitudinal cohort . Archives of General Psychiatry , 60 , 709–717. 10.1001/archpsyc.60.7.709 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Koenen K. C., Moffitt T. E., Roberts A. L., Martin L. T., Kubzansky L., Harrington H., et al.Caspi A. (2009). Childhood IQ and adult mental disorders: A test of the cognitive reserve hypothesis . The American Journal of Psychiatry , 166 , 50–57. 10.1176/appi.ajp.2008.08030343 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Lahey B. B. (2009). Public health significance of neuroticism . American Psychologist , 64 , 241–256. 10.1037/a0015309 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Merikangas K. R., Calkins M. E., Burstein M., He J.-P., Chiavacci R., Lateef T., et al.Gur R. E. (2015). Comorbidity of physical and mental disorders in the neurodevelopmental genomics cohort study . Pediatrics , 135 ( 4 ), e927–e938. 10.1542/peds.2014-1444 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Milne B. J. (2012). New Zealand Socio-economic Index 2006 (NZSEI-06): An introduction for social science researchers . New Zealand Sociology , 27 , 117–127. [ Google Scholar ]
• Milne B. J., Caspi A., Harrington H., Poulton R., Rutter M., & Moffitt T. E. (2009). Predictive value of family history on severity of illness: The case for depression, anxiety, alcohol dependence, and drug dependence . Archives of General Psychiatry , 66 , 738–747. 10.1001/archgenpsychiatry.2009.55 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Moffitt T. E., Arseneault L., Belsky D., Dickson N., Hancox R. J., Harrington H., et al.Caspi A. (2011). A gradient of childhood self-control predicts health, wealth, and public safety . Proceedings of the National Academy of Sciences of the United States of America , 108 , 2693–2698. 10.1073/pnas.1010076108 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Moffitt T. E., Caspi A., Taylor A., Kokaua J., Milne B. J., Polanczyk G., & Poulton R. (2010). How common are common mental disorders? Evidence that lifetime prevalence rates are doubled by prospective versus retrospective ascertainment . Psychological Medicine , 40 , 899–909. 10.1017/S0033291709991036 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Molfese V. J. (2013). Perinatal risks across infancy and early childhood: What are the lingering effects on high and low risk samples? In DiLalla L. F. & Dollinger S. M. C. (Eds.), Assessment of biological mechanisms across the life span (pp. 53–70). Hillsdale, NJ: Erlbaum. [ Google Scholar ]
• Moos R. H., & Moos B. (1981). Family Environment Scale manual . Palo Alto, CA: Consulting Psychologists Press. [ Google Scholar ]
• Ozbay F., Johnson D. C., Dimoulas E., Morgan C. A., Charney D., & Southwick S. (2007). Social support and resilience to stress: From neurobiology to clinical practice . Psychiatry , 4 , 35–40. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Pavot W., & Diener E. (1993). Review of the Satisfaction With Life Scale . Psychological Assessment , 5 , 164–172. 10.1037/1040-3590.5.2.164 [ CrossRef ] [ Google Scholar ]
• Pedersen C. B., Mors O., Bertelsen A., Waltoft B. L., Agerbo E., McGrath J. J., et al.Eaton W. W. (2014). A comprehensive nationwide study of the incidence rate and lifetime risk for treated mental disorders . Journal of the American Medical Association Psychiatry , 71 , 573–581. 10.1001/jamapsychiatry.2014.16 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Poulton R., Moffitt T. E., & Silva P. A. (2015). The Dunedin Multidisciplinary Health and Development Study: Overview of the first 40 years, with an eye to the future . Social Psychiatry and Psychiatric Epidemiology , 50 , 679–693. 10.1007/s00127-015-1048-8 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Regier D. A., & Robins L. N. (1991). Psychiatric disorders in America: The Epidemiologic Catchment Area Study . New York, NY: Free Press. [ Google Scholar ]
• Reiss F. (2013). Socioeconomic inequalities and mental health problems in children and adolescents: A systematic review . Social Science & Medicine , 90 , 24–31. 10.1016/j.socscimed.2013.04.026 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Reynell J. (1969). The Reynell Developmental Language Scales . London, England: National Foundation for Educational Research. [ Google Scholar ]
• Rutter M., Tizard J., & Whitmore K. (1970). Education, health and behaviour . London, England: Longman Green. [ Google Scholar ]
• Schaefer J. D., Caspi A., Belsky D. W., Harrington H., Houts R., Israel S., et al.Moffitt T. E. (2016). Early-life intelligence predicts midlife biological age . The Journals of Gerontology, Psychological Sciences and Social Sciences , 71 , 968–977. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Sebastiani P., & Perls T. T. (2012). The genetics of extreme longevity: Lessons from the New England centenarian study . Frontiers in Genetics , 3 , 277 10.3389/fgene.2012.00277 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Seligman M. E. P., & Csikszentmihalyi M. (2000). Positive psychology. An introduction . American Psychologist , 55 , 5–14. 10.1037/0003-066X.55.1.5 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Shalev I., Caspi A., Ambler A., Belsky D. W., Chapple S., Cohen H. J., et al.Moffitt T. E. (2014). Perinatal complications and aging indicators by midlife . Pediatrics , 134 , e1315–e1323. 10.1542/peds.2014-1669 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Simon G. E., & VonKorff M. (1995). Recall of psychiatric history in cross-sectional surveys: Implications for epidemiologic research . Epidemiologic Reviews , 17 , 221–227. [ PubMed ] [ Google Scholar ]
• Streiner D. L., Patten S. B., Anthony J. C., & Cairney J. (2009). Has ‘lifetime prevalence’ reached the end of its life? An examination of the concept . International Journal of Methods in Psychiatric Research , 18 , 221–228. 10.1002/mpr.296 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Takayanagi Y., Spira A. P., Roth K. B., Gallo J. J., Eaton W. W., & Mojtabai R. (2014). Accuracy of reports of lifetime mental and physical disorders: Results from the Baltimore Epidemiological Catchment Area study . Journal of the American Medical Association Psychiatry , 71 , 273–280. 10.1001/jamapsychiatry.2013.3579 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Terman L. M., & Merrill M. A. (1960). Stanford-Binet Intelligence Scale: Manual for the third revision form L-M . Boston, MA: Houghton Mifflin. [ Google Scholar ]
• Walker E., McGee R. E., & Druss B. G. (2015). Mortality in mental disorders and global disease burden implications: A systematic review and meta-analysis . JAMA Psychiatry , 72 , 334–341. 10.1001/jamapsychiatry.2014.2502 [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Wechsler D. (1974). Manual for the Wechsler Intelligence Scale for Children, Revised . New York, NY: Psychological Corp. [ Google Scholar ]
• Weissman M. M., Gershon E. S., Kidd K. K., Prusoff B. A., Leckman J. F., Dibble E., et al.Guroff J. J. (1984). Psychiatric disorders in the relatives of probands with affective disorders: The Yale University—National Institute of Mental Health collaborative study . Archives of General Psychiatry , 41 , 13–21. 10.1001/archpsyc.1984.01790120015003 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Weissman M. M., Wickramaratne P., Adams P., Wolk S., Verdeli H., & Olfson M. (2000). Brief screening for family psychiatric history: The family history screen . Archives of General Psychiatry , 57 , 675–682. 10.1001/archpsyc.57.7.675 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Whiteford H. A., Degenhardt L., Rehm J., Baxter A. J., Ferrari A. J., Erskine H. E., et al.Vos T. (2013). Global burden of disease attributable to mental and substance use disorders: Findings from the Global Burden of Disease Study 2010 . The Lancet , 382 , 1575–1586. 10.1016/S0140-6736(13)61611-6 [ PubMed ] [ CrossRef ] [ Google Scholar ]
• The WHO World Mental Health Survey Consortium (2004). Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys . JAMA: Journal of the American Medical Association , 291 , 2581–2590. 10.1001/jama.291.21.2581 [ PubMed ] [ CrossRef ] [ Google Scholar ]

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Review Article
• Open access
• Published: 14 June 2019

Conceptualizing mental disorders as deviations from normative functioning

• Andre F. Marquand 1 , 2 , 3 ,
• Seyed Mostafa Kia   ORCID: orcid.org/0000-0002-7128-814X 1 , 2 ,
• Mariam Zabihi 1 , 2 ,
• Thomas Wolfers 1 , 2 ,
• Jan K. Buitelaar   ORCID: orcid.org/0000-0001-8288-7757 1 , 2 , 4 &
• Christian F. Beckmann 1 , 2 , 5  

Molecular Psychiatry volume  24 ,  pages 1415–1424 ( 2019 ) Cite this article

22k Accesses

149 Citations

106 Altmetric

Metrics details

• Diagnostic markers
• Neuroscience
• Prognostic markers

A Correction to this article was published on 26 June 2019

This article has been updated

Normative models are a class of emerging statistical techniques useful for understanding the heterogeneous biology underlying psychiatric disorders at the level of the individual participant. Analogous to normative growth charts used in paediatric medicine for plotting child development in terms of height or weight as a function of age, normative models chart variation in clinical cohorts in terms of mappings between quantitative biological measures and clinically relevant variables. An emerging body of literature has demonstrated that such techniques are excellent tools for parsing the heterogeneity in clinical cohorts by providing statistical inferences at the level of the individual participant with respect to the normative range. Here, we provide a unifying review of the theory and application of normative modelling for understanding the biological and clinical heterogeneity underlying mental disorders. We first provide a statistically grounded yet non-technical overview of the conceptual underpinnings of normative modelling and propose a conceptual framework to link the many different methodological approaches that have been proposed for this purpose. We survey the literature employing these techniques, focusing principally on applications of normative modelling to quantitative neuroimaging-based biomarkers in psychiatry and, finally, we provide methodological considerations and recommendations to guide future applications of these techniques. We show that normative modelling provides a means by which the importance of modelling individual differences can be brought from theory to concrete data analysis procedures for understanding heterogeneous mental disorders and ultimately a promising route towards precision medicine in psychiatry.

Similar content being viewed by others

The normative modeling framework for computational psychiatry

Saige Rutherford, Seyed Mostafa Kia, … Andre F. Marquand

Identifying biological markers for improved precision medicine in psychiatry

Erin Burke Quinlan, Tobias Banaschewski, … IMAGEN Consortium

Brain charts for the human lifespan

R. A. I. Bethlehem, J. Seidlitz, … A. F. Alexander-Bloch

In most areas of medicine, biomarkers that objectively indicate disease state have revolutionized diagnosis and treatment allocation. In contrast, psychiatric disorders are still diagnosed exclusively on the basis of symptoms and biological tests to assist diagnosis or treatment allocation remain to be developed [ 1 ]. This yields clinical groups that are highly heterogenous, both in terms of clinical presentation and underlying biology, which is a major barrier to understanding underlying mechanisms and developing better treatments [ 1 , 2 , 3 ]. This is widely recognized at a theoretical level and over the years, many different theoretical models have been proposed to explain the heterogeneity of psychiatric disorders [ 4 , 5 , 6 , 7 , 8 , 9 ]. These emphasize the myriad pathological mechanisms that may converge on the same symptoms in different participants [ 4 , 5 ] and that the same underlying biological risk factors may result in a different clinical phenotype in different individuals depending on the context, genetic background and critical time window [ 4 , 9 ]. Heterogeneity remains the dominant theme even in recent large-scale theoretical initiatives such as the Research Domain Criteria (RDoC) [ 6 , 7 ] and the European Roadmap for Mental Health Research (ROAMER) [ 8 ]. All these theoretical models are founded on an implicit recognition of the importance of modelling individual differences within and across clinical cohorts [ 10 , 11 ]. It is very important to recognize, however, that this broad theoretical recognition is not reflected in the data analysis strategies employed in practice. Instead, the overwhelming majority of analysis approaches remain focussed on group averages (e.g. the ‘average patient’) and regard individual differences principally as noise. Footnote 1

Normative modelling is an emerging approach that can address this challenge by providing statistical inferences at the level of the individual with respect to an expected pattern [ 12 ]. This is analogous to the widespread use of normative growth charts in paediatric medicine to map child height or weight as a function of age with respect to centiles of variation in a reference population [ 13 ]. Normative modelling generalizes this notion by substituting these variables for clinically relevant variables then applying automated statistical techniques to map centiles of variation across the cohort. This is increasingly used to map variation between cognitive, clinical or demographic variables and quantitative biomarkers derived from neuroimaging [ 12 , 14 , 15 , 16 , 17 ]. The key feature of normative modelling that makes it useful for stratifying cohorts is that it permits the detection and mapping of distinct patterns of abnormality in individuals without requiring a consistent neurobiological signature across all individuals.

Whilst early applications focused on brain development and ageing [ 16 , 18 ], normative modelling has recently been shown to be highly promising for psychiatry [ 12 , 14 , 17 , 19 , 20 , 21 ]. First, to map variation related to brain development and ageing in psychiatric disorders, which is appealing given the neurodevelopmental basis of mental disorders [ 22 ]. For example, neurodevelopmental normative models have been used in the context of schizophrenia [ 17 , 19 ], attention deficit/hyperactivity disorder (ADHD) [ 20 , 23 ] and autism [ 21 , 24 ] to help understand the emergence of mental disorders as deviations from an expected developmental trajectory and identify individuals following an atypical trajectory. Second, normative modelling can also be abstracted beyond development to chart the spectrum of functioning across any cognitive domain. It has been used, for example, to chart variation in reward systems via mappings between trait measures of reward sensitivity and reward-related brain activity [ 12 ]. Finally, normative models can help to understand healthy variation and move beyond simple dimensional theories of mental disorders [ 25 ].

Here, we provide a unifying review of normative modelling for charting individual variation across different behavioural, demographic and biological dimensions, thereby helping to understand heterogeneity within clinical cohorts. We first provide a statistically grounded, yet non-technical overview of its conceptual underpinnings and propose a framework to link the many different methodological approaches that have been proposed. Second, we outline connections between normative modelling and existing approaches for tackling heterogeneity including clustering [ 26 ] and ‘brain age’ approaches that characterize subjects in terms of a difference between a brain-derived predicted age and true chronological age [ 27 , 28 ]. Third, we survey the literature employing normative modelling in clinical conditions. Fourth, we discuss normative modelling as a tool for finding structure in large cohorts, which is important given the recent shifts towards ‘big data’ neuroscience and large population-based cohorts [ 29 , 30 , 31 ]. Finally, we provide recommendations for future studies and critically evaluate the limitations of normative modelling.

Introduction to normative modelling for understanding heterogeneity in clinical cohorts

Normative modelling is a statistical framework for mapping between behavioural, demographic or clinical characteristics and a quantitative biological measure, providing estimates of centiles of variation across the population (Fig.  1a ). Normative modelling provides a concrete method for studying individual differences and parsing heterogeneity across cohorts because it provides statistical inferences at the level of the individual participant as to the degree to which each individual deviates from the normative pattern and allows these deviations to be mapped in each individual. In other words, normative modelling provides a way to quantify and characterize the manner in which different individuals deviate from the expected pattern, and from one another. Importantly, this does not require that atypicalities overlap across participants (e.g. in the same brain regions) or even that a consistent pattern of deviation exists. Therefore, this accommodates the convergence of multiple pathological pathways on the same symptoms in different individuals [ 4 ]. This is clearly different from case-control analyses, which all focus on first order statistics (group means), thereby seeking a consistent pattern of atypicality (i.e. the ‘average patient’). In a case-control context, heterogeneity becomes apparent via inflation of the model residuals and ultimately decreases sensitivity for detecting disorder-related effects. In contrast, normative modelling explicitly models heterogeneity because it focuses on modelling individual variation around the mean using second order statistics (variances). Therefore, normative modelling explicitly characterizes and quantifies the heterogeneity underlying clinical conditions at a finer grained level than is afforded by group averages. Normative modelling also does not require that the clinical group can be cleanly partitioned into subtypes [ 26 ] although it can be used to generate features for clustering. Normative modelling can be used to estimate many different kinds of mappings based on the variables chosen, but here we focus on mappings between behavioural or demographic measures and a quantitative biological readout, most commonly derived from neuroimaging.

Conceptual overview of normative modelling. a Normative modelling is similar to the use of growth charts in paediatric medicine, except the conventional response variable (e.g. height or weight) is substituted for a quantitative biological readout (e.g. regional brain activity). The classical covariates (age and sex) can also be substituted for clinically relevant variables. Normative modelling provides statistical inference at the level of each subject with respect to the normative model (red figure). b Procedural overview of normative modelling. After the choice of reference cohort and variables, the normative model is estimated, before being validated out of sample on new response variables and covariates (y* and x*, respectively). Finally, the estimated model can be applied to a target cohort (e.g. clinical cohort). c A common configuration for normative modelling of neuroimaging data, where a separate normative model is estimated for each sampled brain location. This can be described by a set of functions (y = f(x)) predicting neurobiological response variables (y) from clinical covariates (x). d Normative models can also be estimated for the opposite mapping, where brain measures are chosen as covariates and age or other covariates are chosen as a response variable. See text for further details

Procedurally, normative modelling involves four steps (Fig.  1b ): First, a reference cohort and a set of variables are chosen to define the mapping and population over which variation is measured. Second, a statistical model is estimated to model variance in a response variable (a.k.a. target or dependent variable) from a set of clinically relevant covariates (predictor or independent variables) across the reference cohort. For example, one may estimate a normative model for cortical thickness as a function of age and gender using a population-based reference cohort. Third, it is necessary to assess the accuracy of the normative model for predicting the response variable (e.g. mean-squared error, explained variance). To ensure accurate estimates of generalizability, this must be performed on withheld data (e.g. under cross-validation). Finally, this model can be applied to quantifying the deviations of samples from a target cohort (e.g. clinical cohort) with respect to this reference model.

Many regression models have been proposed for normative modelling, including hierarchical linear models, polynomial regression, quantile regression, support vector regression and Gaussian process regression (Table  1 ). The estimation of normative models is conceptually similar to classical growth charts, for which many approaches have been proposed [ 13 ]. In both cases, the data make multiple demands from the regression model including ensuring precise estimation of outer centiles (where data are sparsest), ensuring centiles vary smoothly as a function of the covariates (and do not cross) and the ability to estimate deviations for individual samples via analytical formulae (e.g. Z -scores) [ 32 ]. In order to provide a conceptual framework linking these approaches, we categorize different approaches according to three criteria: (i) the choice of covariates and response variables; (ii) the degree to which the model separates different sources of variation and (iii) the degree to which the model permits statistical inference the individual level. However, other features are also important, for instance ability to model non-linear relationships.

Choice of covariates and response variables

A simple way to categorize different approaches is in terms of the variables that define the mapping. One common configuration (Fig.  1c ) uses age as a covariate, often in combination with other clinical or demographic variables to predict a quantitative biological readout. However, there are other possibilities: for example, the mapping can be inverted such that age is the response variable which is predicted from clincal or demographic variables (Fig.  1d ). This is the approach used by ‘brain age’ models [ 27 , 28 ] which use multivariate regression to predict age from a pattern of brain-derived measures. For linear models, it is obvious that an association can be detected in either direction simply by inverting the linear model. However, the interpretation of the centiles of variation and regression coefficients differs and we consider that charting variation over the biological readout is more appealing because it directly mirrors the use of growth-charting in paediatric medicine.

As noted, normative models are also not restricted to charting variation across development. By substituting age for other variables, normative models can chart variation in any kind of mapping, for example to link cognitive scores with brain activity patterns [ 12 ].

Separating different sources of variation across the cohort

Normative modelling principally aims to model variation across the cohort over and above estimation of mean effects. To achieve this effectively, it is important to separate different sources of variation, most importantly to differentiate actual variation within the data (i.e. across participants) from variability due to parameter and model uncertainty (i.e. induced variability due to a lack of data). In normative modelling, we quantify these variabilities using two types of uncertainty commonly defined in machine learning [ 33 ]: (i) irreducible (or ‘aleatoric’) uncertainty that reflects true underlying variability that cannot be reduced with more data; (ii) reducible (or ‘epistemic’) variation that reflects parameter uncertainty or ignorance about the true model and can be reduced by more data. Aleatoric uncertainty is of primary interest for stratification because it reflects variation across subjects whereas epistemic uncertainty is nuisance variation that it is desirable to minimize. The degree to which different approaches account for these sources of variability can be classified hierarchically (Fig.  2 ): the simplest approach involves estimating the mean effect only and assessing deviations from the expected pattern using the model residuals (Fig.  2a, b ) [ 19 , 27 ]. Whilst this is appealing in its simplicity, it provides no estimate of variation across the cohort and cannot provide statistical inferences at the individual level (see below). This has been addressed in different ways, for example via estimating confidence intervals via a post hoc regression between the residuals of the model against the true response variable [ 14 ] or using quantile regression to directly estimate centiles of variation in the data [ 34 ]. These approaches provide statistical estimates of variation within the population and can indicate if a particular participant deviates from the expected pattern at a given confidence level, but they do not fully account for different sources of uncertainty, e.g. uncertainty in the estimation of the centiles (Fig.  2c ). Bayesian methods such as Gaussian process regression [ 12 , 16 , 35 ] provide one solution to this problem (Fig.  2d ) by estimating distinct variance components and providing predictions for each participant that account for all sources of uncertainty. This is important for two reasons: first, it provides estimates of centiles of variation within the reference cohort that are not influenced by data density; second, it allows all sources of uncertainty to be taken into account when making predictions. This provides the desirable property that inferences become more conservative in regions of the input space where data are sparse. With these complementary purposes in mind, it may be desirable to report different variance components separately.

Separating different sources of uncertainty in normative modelling. Panels a and b show the simplest approach for normative models which do not quantify uncertainty at all ( a : linear model, b : non-linear model). Instead, deviations from the model (red figures) are assessed via the residuals from a regression function (blue lines). In red, the corresponding equation for assessing deviations from the model is shown where deviation from the normative model are assessed simply as the difference between the true ( y ) and predicted ( \(\widehat y\) ) normative response variable for each subject. c Some models estimate centiles of variation explicitly either via separate model fits or post hoc to the initial regression fit (blue dotted lines). This captures ‘aleatoric’ or irreducible variation in the cohort which shows how subjects vary across the population ( \(\sigma _a^2\) ). However, there is also uncertainty associated with each of these centiles of variation (shaded blue regions), which is highest in regions of low data density and should be accounted for. d Some models separate and take all sources of variation into account (i.e. also including ‘epistemic’ uncertainty ( \(\sigma _e^2\) ), which can be reduced by the addition of more data). This allows the model to automatically adjust predictions, becoming more conservative in regions where data are sparse. This is shown by a widening of the statistical intervals, although note that these intervals now have a different interpretation to those in ( c ). For example, the right-most figure in ( d ) would not be judged as an outlier, whereas the same figure may be judged as an outlier in models that do not account for all sources of uncertainty ( c ). This is important to prevent a subject being declared as ‘atypical’ simply because of data sparsity. See text for further details

Degree of individual prediction

The ability of the model to perform single participant inference can be classified hierarchically. At the simplest level (‘numerical inference’; Fig.  2a, b ), the model only provides numerical deviations (i.e. residuals of the target cohort from the reference model). This is the approach taken by most brain age approaches [ 27 ] and permits group-level inferences about how deviations correlate—for example—with symptoms. In contrast, some models provide estimates of centiles of variation within the population (‘statistical inference’; e.g. quantile regression [ 34 ] and Bayesian techniques [ 12 , 16 ]). This provides inferences as to whether each individual deviates from the model at a given statistical significance level (Fig.  2c ). Some studies have derived variance estimates via post hoc regression on the model residuals [ 14 ] and atypicality cutoffs could also be defined post hoc, although this must be done on unseen data to remain unbiased. However, as described above, it is important to recognize whether predictions account for all sources of variance. If they do not, they may yield overly optimistic inferences. This can be addressed using models that estimate separate variance components for different types of variance and account for all uncertainty in the predictions (Fig.  2d ).

In addition to mapping deviations in individuals, it is often desirable to estimate participant-level summary statistics for the overall deviation from the normative pattern. Different methods have been proposed for this, e.g.: combining deviations across all voxels [ 16 ] or modelling the most extreme deviations in each subject using extreme value statistics [ 12 ].

Relationships to other approaches for parsing heterogeneity

Normative modelling is complementary to the predominant approach for tackling heterogeneity in mental disorders, i.e. subtyping using clustering algorithms. Clustering has been widely applied [ 26 ] and is often useful. However, it also suffers from limitations: first, clustering assumes that the clinical group can be cleanly partitioned into subtypes. However, this assumption is seldom evaluated, which is problematic because clustering algorithms always yield a result, regardless of whether clusters are ‘really there’ [ 26 , 36 ]. Second, clustering focuses on group averages and does not fully model individual variation within clusters. In other words, most clustering algorithms regard piece-wise constant clusters as atomic units. Whilst some algorithms provide ‘soft’ cluster assignments that capture some variation within clusters, this cannot accurately model, for example, a spectrum of functioning. In contrast, normative modelling shifts the analytical focus: (i) away from group means to understanding cohort variation (i.e. from first- to second-order statistics); (ii) towards understanding variation across individuals and (iii) towards mapping deviations at the level of individual.

Normative modelling is complementary to alternative techniques for individual prediction; for example, supervised discriminative models [ 3 ] can assess the degree of group separation in a case-control sense and therefore provide predictions that are specific for certain disorders. On the other hand, normative modelling can be used to understand the variation across the cohort independently of the clinical labels.

Brain age models are related and complementary to normative modelling. They can be considered as a type of normative model which estimate the opposite mapping (i.e. brain readouts as covariates and age as the response variable; Fig.  1d ). As noted, choosing age as a covariate mirrors growth-charting in paediatric medicine and allows the regional deviations in each subject to be mapped, which is desirable for interpretation. In contrast, brain age models condense a complex multivariate pattern into a single number (a deviation between true and predicted age). This is often useful because it summarizes a complex pattern by an interpretable score. On the other hand, it provides limited ability to stratify individuals or identify which brain regions underlie any observed deviation. This is important because different subjects may have the same predicted brain age because of distinct underlying abnormality patterns [ 27 ].

Applications of normative modelling in psychiatry

Normative modelling has been applied to many clinical phenotypes: unsurprisingly, many applications have focused on studying changes in brain organization across the lifespan. More recently, studies have emerged applying normative modelling to map the biological heterogeneity underlying mental disorders (Table  1 ). Footnote 2 Taken together, the applications reviewed here show the flexibility of normative modelling for many different clinical phenotypes, on the basis of different clinical and biological measures. Moreover, they highlight the value of normative modelling for studying individual differences in that they show that: (i) a potentially small number of patients have alterations in the same brain regions and (ii) that the pattern of individualized regional differences detected by normative models can be very different to case-control differences. For example, normative deviations may be partially consistent with case control-effects [ 17 ], very different [ 21 ] or evident in the absence of case-control effects [ 24 ].

Normative modelling for big data

Normative modelling is useful for understanding variation in ‘big data’ cohorts. This parallels an increasing focus in clinical neuroimaging towards acquiring large population-based cohorts that capture a wide range of clinically relevant variation [ 29 , 30 , 31 ]. The conventional motivation is to increase statistical power for the detection of subtle effects [ 37 ]. While this is undoubtedly important, such cohorts also provide an excellent opportunity to understand structure in heterogeneous clinical populations. Normative modelling is ideal for this due to its focus on understanding variation rather than detecting mean effects and can be used to find distinct and potentially non-overlapping patterns of abnormality.

On the other hand, big data cohorts introduce challenges, including computational scaling of analytical methods to many data points, requiring availability of modelled variables across all subjects and dealing with nuisance variation. For example, most large datasets include data from multiple study sites, which increases the risk of missing data and introduces the possibility that observed deviations could be related to site variance. To address these concerns, careful stratification procedures during model fitting and cross-validation and explicitly modelling different sources of variance are important (e.g. using hierarchical models). Alternatively, for predictions on new sites, normative models may be recalibrated on held-out normative control data to ensure that the normative model remains appropriate for the new data sample. There are many ways to achieve this [ 38 ], but a simple approach involves adjusting the mean, slope and variance after fitting a post hoc regression to withheld subjects.

Another use for normative models is for calibrating measures on different scales to a common normative reference. In other words, separate normative models can be estimated for different cognitive and biological mappings. This has the effect of rescaling different variables to a common reference range (for example, Z -statistics reflecting the number of standard deviations each subject is from the population norm). This forms an ideal set of features for the application of clustering algorithms, in the spirit of precision medicine. Relative to application to the raw data, this increases interpretability by scaling diverse data to population norms and also can tease apart correlated symptom domains more clearly than using clinical or biological data alone.

Ultimately, estimating normative models to link multiple phenotypic measurements with their multifaceted biological underpinnings is likely to be very important to: (i) understand disorders across multiple domains in the spirit of RDoC and ROAMER; (ii) identify different groups of patients with different atypical mechanisms; (iii) to better understand healthy variation and how this relates to the mechanisms of mental disorders and (iv) to move beyond simple dimensional theories of mental disorders [ 25 ].

Study design considerations

The applications above show that normative modelling is very flexible given the choice of covariates, response variables, target cohort and reference cohort. The choice of reference cohort is particularly crucial; it is important that it captures a wide range of variation in the reference population. In paediatric medicine, the typical choice is a population-based cohort containing thousands of participants. In psychiatry, it is also common to estimate normative models using population-based cohorts that include participants across the full range of functioning (i.e. both healthy and with disorders) [ 19 , 20 ], although this is not the only option and may not always be the optimal choice. For neuroimaging, several large population-based cohorts are being acquired [ 29 , 31 , 39 ], however these often focus on specific lifespan periods [ 29 ] and are frequently enriched for individuals ‘at-risk’ for mental disorders [ 31 , 39 ]. Moreover, since these cohorts aim to address multiple questions, they may lack rich clinical phenotyping measures that are valuable for characterizing deviations (see below). Another option is to apply normative models to existing cohorts (e.g. based on case-control designs) [ 17 , 23 , 24 ]. In such cases, either the whole cohort or only the healthy participants can be used for the reference cohort. If the whole cohort is used, it is important to remember that under a case-control paradigm the frequency of the clinical phenotype is usually much higher than the population prevalence (e.g. equal numbers of cases and controls). If only the healthy subjects are used as the reference, this can be considered an approximation to a population-based cohort, which is reasonable if the prevalence of the clinical phenotype in the wider population is relatively low. In both cases, the deviations should be interpreted with respect to the cohort chosen. Regarding the choice of the target cohort, this can be the same as the reference cohort, provided the predictions are derived in an unbiased manner, for example, under cross-validation.

Limitations

Normative modelling is a bottom-up approach to map variation and should not be considered a substitute for hypothesis testing or top-down theory driven approaches. Rather, it is desirable to combine the benefits of both. For example, combining top-down theory driven approaches with supervised discriminative models [ 3 ] and normative models. These can be used, respectively, to assess the degree of group separation in a case-control sense and to map the variation across individuals with respect to the theory-driven model.

Another important consideration is that normative models do not directly indicate whether the deviations obtained are clinically relevant. For example, deviations may be biologically meaningful yet unrelated to psychopathology, or they may be a result of artefactual variation (e.g. site variation). Therefore, external validation of derived deviations on external measures such as symptoms, genotype and environmental factors is crucial, as are careful data preprocessing and checking procedures.

Future developments: towards precision psychiatry

Normative modelling provides a method to map deviations from an expected pattern at the individual level. This has been helpful to understand individual variation in the context of brain development or ageing and also for mapping variation across multiple cognitive domains and measures of brain organization. This provides the potential to understand individual variation across a multi-dimensional cognitive space, consistent with initiatives such as RDoC [ 6 , 7 , 8 ]. However, work remains to be done as to how these can be integrated and to determine the clinical relevance of individual variability. In other words, how, when and why individual variability turns into vulnerability or resilience. Indeed, a high priority is to develop methods that can convert deviations from normative models to a stratification of individuals. As noted, training clustering algorithms on the deviations from multiple normative models is one option [ 12 , 26 ]. However, conventional ‘hard’ clustering algorithms allocate each subject to a single cluster and do not accommodate the possibility of multiple overlapping mechanisms operating in different individuals. Therefore, a more promising route may be using ‘soft’ clustering algorithms or other latent variable models that allow subjects to be allocated to multiple potentially overlapping clusters or risk profiles [ 40 ]. Another area of future work involves explicitly modelling spatial information, which may increase sensitivity for detection of spatially distributed patterns of abnormality [ 41 , 42 , 43 ].

Conclusions and outlook

We have surveyed the emerging literature employing normative modelling to mental disorders. We have shown that these methods are highly flexible: they can naturally applied to estimating centiles of variation in brain growth and to mappings between many aspects of behaviour, cognition and biology. The most important feature of normative models is their ability to make predictions at the level of individual with respect to a normative pattern and that they shift emphasis from studying mean effects to understanding individual variation. They provide a means by which modelling individual differences can be brought from theory to concrete data analysis procedures and therefore a promising route towards precision medicine in psychiatry.

Change history

26 june 2019.

A correction to this paper has been published and can be accessed via a link at the top of the paper.

This includes classical analysis methods based on the general linear model (e.g. t -tests and analysis of (co-) variance) plus classical and supervised machine learning classification models (e.g. linear discriminant analysis, logistic regression and support vector machines) and unsupervised clustering methods (e.g. k-means), where—at best—confound regressors are used to remove some of the cross-subject variation.

Note that we include some studies that can also be considered brain age studies but that share salient features with normative models (e.g., ref. [ 14 ]), we refer the reader elsewhere for more detailed reviews of brain age studies [ 27 , 28 ].

Kapur S, Phillips AG, Insel TR. Why has it taken so long for biological psychiatry to develop clinical tests and what to do about it? Mol Psychiatry. 2012;17:1174–9.

Article   CAS   Google Scholar  

Scarr E, Millan MJ, Bahn S, Bertolino A, Turck CW, Kapur S, et al. Biomarkers for Psychiatry: The Journey from Fantasy to Fact, a Report of the 2013 CINP Think Tank. Int J Neuropsychoph. 2015;18:pyv042.

Article   Google Scholar  

Wolfers T, Buitelaar JK, Beckmann C, Franke B, Marquand AF. From estimating activation locality to predicting disorder: a review of pattern recognition for neuroimaging-based psychiatric diagnostics. Neurosci Biobehav Rev. 2015;57:328–49.

Cicchetti D, Rogosch FA. Equifinality and multifinality in developmental psychopathology. Dev Psychopathol. 1996;8:597–600.

Cannon TD, Keller MC. Endophenotypes in the genetic analyses of mental disorders. Annu Rev Clin Psychol. 2006;2:267–90.

Insel T, Cuthbert B, Garvey M, Heinssen R, Pine DS, Quinn K, et al. Research Domain Criteria (RDoC): toward a new classification framework for research on mental disorders. Am J Psychiatry. 2010;167:748–51.

Insel TR, Cuthbert BN. Brain disorders? Precisely. Science. 2015;348:499–500.

Schumann G, Binder EB, Holte A, de Kloet ER, Oedegaard KJ, Robbins TW, et al. Stratified medicine for mental disorders. Eur Neuropsychopharmacol. 2014;24:5–50.

Marin O. Developmental timing and critical windows for the treatment of psychiatric disorders. Nat Med. 2016;22:1229–38.

Foulkes L, Blakemore SJ. Studying individual differences in human adolescent brain development. Nat Neurosci. 2018;21:315–23.

Seghier ML, Price CJ. Interpreting and utilising intersubject variability in brain function. Trends Cogn Sci. 2018;22:517–30.

Marquand AF, Rezek I, Buitelaar J, Beckmann CF. Understanding heterogeneity in clinical cohorts using normative models: beyond case-control studies. Biol Psychiatry. 2016;80:552–61.

Cole TJ. The development of growth references and growth charts. Ann Hum Biol. 2012;39:382–94.

Erus G, Battapady H, Satterthwaite TD, Hakonarson H, Gur RE, Davatzikos C, et al. Imaging patterns of brain development and their relationship to cognition. Cereb Cortex. 2015;25:1676–84.

Eavani H, Hsieh MK, An Y, Erus G, Beason-Held L, Resnick S, et al. Capturing heterogeneous group differences using mixture-of-experts: application to a study of aging. Neuroimage. 2016;125:498–514.

Ziegler G, Ridgway GR, Dahnke R, Gaser C, Alzheimer’s Dis N. Individualized Gaussian process-based prediction and detection of local and global gray matter abnormalities in elderly subjects. Neuroimage. 2014;97:333–48.

Wolfers T, Doan NT, Kaufmann T, Alnæs D, Moberget T, Agartz I, et al. Mapping the heterogeneous phenotype of schizophrenia and bipolar disorder using normative models. JAMA Psychiatry. 2018;75:1146–55.

Brewer JB. Fully-automated volumetric MRI with normative ranges: translation to clinical practice. Behav Neurol. 2009;21:21–8.

Gur RC, Calkins ME, Satterthwaite TD, Ruparel K, Bilker WB, Moore TM, et al. Neurocognitive growth charting in psychosis spectrum youths. JAMA Psychiatry. 2014;71:366–74.

Kessler D, Angstadt M, Sripada C. Growth charting of brain connectivity networks and the identification of attention impairment in youth. JAMA Psychiatry. 2016;73:481–9.

Bethlehem R, Seidlitz J, Romero-Garcia R, Lombardo M. Using normative age modelling to isolate subsets of individuals with autism expressing highly age-atypical cortical thickness features. BioRxiv 2018:1–23.

Insel TR. Mental disorders in childhood shifting the focus from behavioral symptoms to neurodevelopmental trajectories. JAMA. 2014;311:1727–8.

Wolfers T, Beckman CF, Hoogman M, Buitelaar JK, Franke B, Marquand A. Individual differences v. the average patient: mapping the heterogeneity in ADHD using normative models. Psychol Med. 2019;311:1727–8.

Zabihi M, Oldehinkel M, Wolfers T, Frouin V, Goyard D, Loth E, et al. Dissecting the heterogeneous cortical anatomy of autism spectrum disorder using normative models. Biol Psychiatry Cogn Neurosci Neuroimaging. 2018.

Holmes AJ, Patrick LM. The myth of optimality in clinical neuroscience. Trends Cogn Sci. 2018;22:241–57.

Marquand AF, Wolfers T, Mennes M, Buitelaar J, Beckmann CF. Beyond lumping and splitting: a review of computational approaches for stratifying psychiatric disorders. Biol Psychiatry: Cogn Neurosci. 2016;1:433–47.

Google Scholar  

Cole JH, Franke K. Predicting Age Using Neuroimaging: Innovative Brain Ageing Biomarkers. Trends Neurosci. 2017;40:681–90.

Cole JH, Marioni RE, Harris SE, Deary IJ. Brain age and other bodily ‘ages’: implications for neuropsychiatry. Mol Psychiatry. 2019;24:266–81.

Miller KL, Alfaro-Almagro F, Bangerter NK, Thomas DL, Yacoub E, Xu JQ, et al. Multimodal population brain imaging in the UK Biobank prospective epidemiological study. Nat Neurosci. 2016;19:1523–36.

Van Essen DC, Smith SM, Barch DM, Behrens TEJ, Yacoub E, Ugurbil K, et al. The WU-Minn Human Connectome Project: an overview. Neuroimage. 2013;80:62–79.

Volkow N, Koob G, Croyle R, Bianchi D, Gordon J, Koroshetz W, et al. The conception of the ABCD study: from substance use to a broad NIH collaboration. Dev Cogn Neurosci. 2017;32:4–7.

Borghi E, de Onis M, Garza C, Van den Broeek J, Frongillo EA, Grummer-Strawn L, et al. Construction of the World Health Organization child growth standards: selection of methods for attained growth curves. Stat Med. 2006;25:247–65.

Gal Y. Uncertainty in deep learning. Cambridge: University of Cambridge; 2016. PhD thesis.

Huizinga W, Poot D, Vernooij M, Rothschupkin G, Ikram M, Rueckert D, et al. A spatio-temporal reference model of the aging brain. Neuroimage. 2018;169:11–2.

Rezek I, Beckmann C. Models of disease spectra. arXiv:1207.4674 [stat.ML]: arXiv preprint, 2012.

Liu Y, Hayes DN, Nobel A, Marron JS. Statistical significance of clustering for high-dimension, low-sample size data. J Am Stat Assoc. 2008;103:1281–93.

Smith SM, Nichols TE. Statistical challenges in “big data” human neuroimaging. Neuron. 2018;97:263–8.

Sansom PG, Ferro CAT, Stephenson DB, Goddard L, Mason SJ. Best practices for postprocessing ensemble climate forecasts. Part I: selecting appropriate recalibration methods. J Clim. 2016;29:7247–64.

Calkins ME, Merikangas KR, Moore TM, Burstein M, Behr MA, Satterthwaite TD, et al. The Philadelphia Neurodevelopmental Cohort: constructing a deep phenotyping collaborative. J Child Psychol Psychiatry. 2015;56:1356–69.

Ruiz FJR, Valera I, Blanco C, Perez-Cruz F. Bayesian nonparametric comorbidity analysis of psychiatric disorders. J Mach Learn Res. 2014;15:1215–47.

Kia SM, Marquand A. Normative modelling of neuroimaging data using scalable multi-task gaussian processes. ArXiv. 2018;1–12.

Kia SM, Beckmann CF, Marquand AF. Scalable multi-task Gaussian process tensor regression for normative modeling of structured variation in neuroimaging data. ArXiv. 2018;1808.00036.

Kia SM, Marquand A. Neural processes mixed-effect models for deep normative modeling of clinical neuroimaging data. arXiv. 2018;1812.04998.

Lefebvre A, Delorme R, Delanoe C, Amsellem F, Beggiato A, Germanaud D, et al. Alpha waves as a neuromarker of autism spectrum disorder: the challenge of reproducibility and heterogeneity. Front Neurosci. 2018;12:662.

Ordaz SJ, Foran W, Velanova K, Luna B. Longitudinal growth curves of brain function underlying inhibitory control through adolescence. J Neurosci. 2013;33:18109–24.

Alexander-Bloch AF, Reiss PT, Rapoport J, McAdams H, Giedd JN, Bullmore ET, et al. Abnormal cortical growth in schizophrenia targets normative modules of synchronized development. Biol Psychiatry. 2014;76:438–46.

Download references

Acknowledgements

This work was supported by the Netherlands Organisation for Scientific Research via Vernieuwingsimpuls VIDI fellowships to Andre Marquand (grant number 016.156.415) and Christian Beckmann (864.12.003). The authors also gratefully acknowledge support from a Wellcome Trust Strategic Award (098369/Z/12/Z).

Author information

Authors and affiliations.

Donders Centre for Cognitive Neuroimaging, Donders Institute for Brain, Cognition and Behaviour, Radboud University, Nijmegen, the Netherlands

Andre F. Marquand, Seyed Mostafa Kia, Mariam Zabihi, Thomas Wolfers, Jan K. Buitelaar & Christian F. Beckmann

Department of Cognitive Neuroscience, Radboud University Medical Centre, Nijmegen, the Netherlands

Department of Neuroimaging, Centre for Neuroimaging Sciences, Institute of Psychiatry, King’s College London, London, UK

Andre F. Marquand

Karakter Child and Adolescent Psychiatric University Centre, Nijmegen, the Netherlands

Jan K. Buitelaar

Oxford Centre for Functional Magnetic Resonance Imaging of the Brain (FMRIB), University of Oxford, Oxford, UK

Christian F. Beckmann

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Andre F. Marquand .

Ethics declarations

Conflict of interest.

JKB has been a consultant to/member of advisory board of and/or speaker for Janssen Cilag BV, Eli Lilly, Medice, Roche, and Servier. He is not an employee of any of these companies. He is not a stock shareholder of any of these companies. He has no other financial or material support, including expert testimony, patents, or royalties. CFB is director and shareholder of SBGneuro Ltd. The other authors declare that they have no conflict of interest.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Marquand, A.F., Kia, S.M., Zabihi, M. et al. Conceptualizing mental disorders as deviations from normative functioning. Mol Psychiatry 24 , 1415–1424 (2019). https://doi.org/10.1038/s41380-019-0441-1

Download citation

Received : 29 December 2018

Revised : 15 April 2019

Accepted : 29 April 2019

Published : 14 June 2019

Issue Date : October 2019

DOI : https://doi.org/10.1038/s41380-019-0441-1

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Nuclei-specific hypothalamus networks predict a dimensional marker of stress in humans.

• Daria E. A. Jensen
• Klaus P. Ebmeier
• Miriam C. Klein-Flügge

Nature Communications (2024)

Brain-based classification of youth with anxiety disorders: transdiagnostic examinations within the ENIGMA-Anxiety database using machine learning

• Willem B. Bruin
• Paul Zhutovsky
• Moji Aghajani

Nature Mental Health (2024)

Immersive virtual reality gameplay detects visuospatial atypicality, including unilateral spatial neglect, following brain injury: a pilot study

• David R. Painter
• Michael F. Norwood
• Heidi Zeeman

Journal of NeuroEngineering and Rehabilitation (2023)

Developmental prediction modeling based on diffusion tensor imaging uncovering age-dependent heterogeneity in early childhood autistic brain

• Xinyue Huang
• Yating Ming

Molecular Autism (2023)

A canonical trajectory of executive function maturation from adolescence to adulthood

• Brenden Tervo-Clemmens
• Finnegan J. Calabro
• Beatriz Luna

Nature Communications (2023)

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Final dates! Join the tutor2u subject teams in London for a day of exam technique and revision at the cinema. Learn more →

Reference Library

Collections

• See what's new
• All Resources
• Student Resources
• Assessment Resources
• Teaching Resources
• CPD Courses
• Livestreams

Study notes, videos, interactive activities and more!

Psychology news, insights and enrichment

Currated collections of free resources

Browse resources by topic

• All Psychology Resources

Resource Selections

Currated lists of resources

Deviation from Ideal Mental Health

Deviation from ideal mental health is a definition of abnormality, which suggests that abnormal behaviour should be defined by the absence of particular (ideal) characteristics. Jahoda proposed six principles of ideal mental health, including having a positive view of yourself and being resistant to stress. Therefore, if an individual does not demonstrate Jahoda’s criteria, they would be classified as abnormal.

• Share on Facebook
• Share on Twitter
• Share by Email

Introduction to ​Psychopathology

Study Notes

Failure to Function Adequately

Phobias, depression and ocd - emotional & cognitive characteristics.

Quizzes & Activities

Definitions of Abnormality

Deviation from ideal mental health (online lesson).

Online Lessons

Psychopathology - take the Yes/No challenge

Example answers for psychopathology: a level psychology, paper 1, june 2019 (aqa).

Exam Support

Psychopathology: MCQ Revision Test 1 for AQA A Level Psychology

Topic Videos

Example Answers for Psychopathology: A Level Psychology, Paper 1, June 2018 (AQA)

Psychopathology "revision q&a" flip, example answer for question 19 paper 1: a level psychology, june 2017 (aqa), free unit assessment: psychopathology.

22nd September 2017

Free Key Term Glossary & Crosswords: Psychopathology

21st September 2017

Free Topic Essays: Psychopathology

20th September 2017

Free Topic Companion: Psychopathology

Model answer for question 6 paper 2: as psychology, june 2016 (aqa).

Mental Health - Time To Act?

30th March 2016

Our subjects

• › Criminology
• › Economics
• › Geography
• › Health & Social Care
• › Psychology
• › Sociology
• › Teaching & learning resources
• › Student revision workshops
• › Online student courses
• › CPD for teachers
• › Livestreams
• › Teaching jobs

Boston House, 214 High Street, Boston Spa, West Yorkshire, LS23 6AD Tel: 01937 848885

• › Contact us
• › Terms of use
• › Privacy & cookies

© 2002-2024 Tutor2u Limited. Company Reg no: 04489574. VAT reg no 816865400.