Is the current approach to diagnosing mental health conditions holding back our research? Jerome Kagan asks clinicians to approach diagnosis in a different way.
- The current reluctance to question popular diagnostic categories for mental illnesses has slowed progress in discovering the diverse causes of symptoms of the same category.
- Reliance on a single measure of a mental state or a theoretical concept is also retarding deeper understanding because no single measure reveals all the features of a psychological state or a construct.
- The social class of the child’s family of rearing makes important contributions to most mental illnesses and clinicians should try to assess the effects of childhood experiences.
Psychologists and biologists favour different premises and strategies. Biologists are splitters who assume that most of the phenomena they study are composed of a number of different kinds and they have to parse what they observe into a set of more homogenous kinds. The biologists who first discovered that some individuals had an abnormally high number of white blood cells called this state leukemia. Subsequent investigators analyzed this phenomenon and found many different kinds of genetic abnormalities that were accompanied by the same high white blood cell count. These were different diseases requiring different treatments.
By contrast, many psychologists are lumpers who resist unpacking an observation into its distinctive kinds. Clinicians, as well as investigators, continue to write about depression, anxiety, or impulsivity as if each label named a unitary phenomenon. The concept of autistic spectrum implies that autism is one condition that varies in severity, where the evidence reveals that the symptoms are usually the result of a large number of different genetic changes, some rare, that occurred during gestation. These are different conditions. No physiologist would posit a headache spectrum. The clinician’s loyalty to retaining these labels is due, in part, to the need to use them to be reimbursed. But many researchers who cannot use this excuse also honor these categories.
A reliance on a single measure of a concept, for example a questionnaire to evaluate changes in depressed mood, a parental interview to assess attention deficit hyperactivity disorder (ADHD) in a child, or salivary cortisol to index anxiety, is a second error. This practice violates a fundamental scientific principle. There are very few psychological or biological phenomena that are the result of only one cascade of preceding events. The task is to discover the pattern of features that defines each of the cascades. For example, difficulty breathing can be caused by either allergic changes in bronchi or lung cancer. Each measure of the hidden elements of the cascade is like a filter as it reveals only some features of the hidden process. Hence, clinicians ought to use more than one method to measure a patient’s mental state. Sole reliance on a person’s verbal report cannot reveal all the features that define the psychological state of the person with a mental illness. Neither can reliance on one measure of anxiety, memory, or impulsivity reveal all the properties of these concepts.
Most therapists rely only on a patient’s report of their mood or symptoms to evaluate the level of improvement, even though many patients report improvement to rationalize the effort they expended and to please the therapist. Investigators use this evidence to decide whether one therapeutic regimen for an illness category is superior to others. However, studies find that, over a period of years, no particular form of psychotherapy is better than any other for most illnesses. This fact would make Jerome and Julia Frank smile for they wrote a slim book in 1991 titled Persuasion and healing: A comparative study of psychotherapy. They argued that most patients in psychotherapy improve when the patient and therapist: share the same ideas about the cause of, and cure for, their symptoms; the patient trusts the competence and empathy of the therapist; and the therapist communicates sincere concern with the patient. That is why shamans can be effective.
There are very few psychological or biological phenomena that are the result of only one cascade of preceding events.
The balance among conditions
Current diagnoses for a mental disorder include individuals with different patterns of properties and life histories. The three most critical properties are vulnerabilities created by the person’s biology, childhood experiences, and current life circumstances. The biological vulnerabilities include temperamental biases.
A temperamental bias is usually defined as an inherited brain state that renders an infant susceptible to varied feelings and actions. I suspect there are a large number of temperamental biases to be discovered. These biases can include irritability, ease of soothing, high or low activity, a predictable or labile circadian rhythm, as well as variation in the level of arousal to unfamiliar or unexpected events.
My colleagues and I have studied this last bias for more than 40 years. We discovered that four-month-old, healthy, middle-class, Caucasian infants born in the Boston area varied in their tendency to display vigorous limb movement and crying in response to several unfamiliar but innocent events. One event was a set of three colorful mobiles, with a varied number of objects, that were moved in front of the babies’ face for twenty second intervals. A second was tape recordings of human voices speaking sentences, with no actual human present. We called the 20% of infants who began to thrash their limbs and cry to these simple, non-threatening but unexpected events high reactive. The 40% who were motorically still and rarely cried were called low reactive.
Variation in the excitability of the amygdala to unfamiliar and unexpected events probably contributes to the differences between high and low reactives. That hypothesis implies that high reactive infants should become shy, timid toddlers, whereas low reactive infants are apt to become bold and sociable. We recognized that the family and peer environments had the power to suppress shyness in some high reactive children and establish a timid posture in a low reactive. That expectation was confirmed. Although a vigilant, cautious persona was always more common among high reactive one, two, five, and seven-year-olds, as well as eleven and eighteen-year-old adolescents, the behavioral differences between the two groups became smaller as the children developed. However, the two groups differed more consistently in several biological features at every age. For example, the high reactive adolescents had higher and less variable heart rates, and larger event-related potentials in the electroencephalogram (EEG), as well as greater blood flow to the amygdala in reaction to unexpected, unfamiliar pictures, e.g. a baby’s head on a horse’s body. It appears that the two temperamental groups preserved their biology while the behavioral profiles they displayed to others were more likely to change. Carl Jung would not have been surprised by this result. He believed that the traits a person displays in public, called a persona, did not reveal their inner, private feelings, or anima. Many of our high reactive adolescents, whose persona did not contain signs of timidity, confessed that they often felt anxious even though they did not display any observable clues to this state.
An emotional identification with a parent, family pedigree, or one’s class, ethnic, or religious group is a second significant childhood experience that affects adult personality.
The social class of the child’s family of rearing is one of the most important sources of childhood experiences because class is correlated with parental practices, values, diet, health, quality of schools attended, IQ, grades, years of education, occupation, and sense of agency. No set of genes can predict these outcomes as well as class of rearing. A low reactive infant growing up in a supportive, middle-class family that encouraged achievement might choose a challenging occupation, such as neurosurgery, trial law, or portfolio manager. The same child raised in a poor, single parent family in a city with gangs is likely to choose a criminal career.
I was studying development in a small, Mayan Indian village located on the shores of Lake Atitlan in northwest Guatemala in 1972. The small houses, with dirt floors and no plumbing, were identical in appearance. However, the children of the parents who owned the tiny lot on which the house stood, whose families had higher status, performed better on our cognitive tests than children of families who rented the land.
An emotional identification with a parent, family pedigree, or one’s class, ethnic, or religious group is a second significant childhood experience that affects adult personality. A large number of children experience vicarious pride or shame because of their belief that they share some distinctive features of a parent, relative, or the typical member of their class, ethnic, or religious group. The Israeli writer Amoz Oz described the pride he enjoyed from being the son of a famous writer. On the other hand, 12 year old Rainier Hoess experienced corrosive shame when he first learned that his grandfather had been the Commandant at Auschwitz.
The American writer John Wideman, who grew up poor in a black ghetto in Pittsburgh, Pennsylvania, confessed in his book Brothers and keepers that his identification with a disadvantaged class and ethnic group created self-doubt that was accompanied by the thought that today was the day many critics would question his talent. The British literary critic Sir Frank Kermode, also from a poor home on the Isle of Man, suffered similar thoughts which he described in his book Not entitled: A memoir. Of course, not every poor child feels this way and many black Americans are proud of their ethnic identity.
Clinicians should try to gather indirect clues to a person’s possible temperamental biases.
An adolescent’s identification with a disadvantaged class is more likely today than it was 500 years ago when 80–90% of the population were peasants. The fact that being poor is more distinctive today makes it easier to identify with this group. The greater the income inequality in a city or region, the higher the frequency of crime and civil unrest.
Identifications with an ethnic group are also accompanied by vicarious emotion. Michael McDonald, who grew up in a mainly Irish neighborhood in South Boston in the 1960s, described the vicarious pride he felt when the adults in South Boston marched against a 1974 judge’s order to bus Irish children to distant schools to aid integration. Agnostic Jewish professors, who denied an ethnic identification in an interview, confessed to feeling surprised by the joy they felt after Israel won the seven-day war. But the polymath Norbert Wiener wrote in a 1996 memoir that his Jewish identity ‘forced on me a sense of inferiority’.
I now turn to the person’s local circumstances, the third major condition and the one most often ignored. Consider a pair of monozygotic twin sisters growing up in a middle-class family in a small town in northern Scotland who were born with a high reactive temperament and, therefore, were at a higher risk for developing an anxiety disorder. One twin remained in the town, attended a nearby college, became the town librarian, married, had two children, and lived a happy life. Her sister decided to go to a university in London, even though she had no friends there. Her first year was marked by loneliness, anxiety, and poor grades. As a result, she developed the symptoms of social anxiety disorder because she found herself in circumstances that were inimical to her temperament. Many successful, happy Jewish businessmen who fled Germany with their families in the 1930s arrived in the United States during the depression. Jobs were scarce and, since they did not have the skills needed at the time, they became depressed. Homicides are more common in cities with high levels of income inequality. Auto fatalities are more frequent in states with low population densities and fewer cars on the roads. These, and many other examples, affirm that local circumstances affect behavior, mood, and the probability that a person will develop a mental illness.
What to do?
These ideas have important implications. Clinicians should try to gather indirect clues to a person’s possible temperamental biases. Some candidate measures include heart rate and heart rate variability as indexes of sympathetic tone, EEG power bands and event-related potentials to unexpected events as a sign of a vigilant trait, and the ratio of the lengths of the index over the ring finger because it is a crude index of the amount of testosterone secreted by the male embryo. Verbal reports should be supplemented with direct observations, reports of close friends, and objective tests. A diagnosis of ADHD should not rely only on a parental or teacher report.
Finally, investigators have to develop additional ways to measure the effects of any therapeutic intervention. These will probably involve some biological measures. A report of less anxiety after 16 weeks of therapy that is not accompanied by some biological signs fails to distinguish between patients who enjoyed a less vigilant mood and those who did not.
The current reluctance to question the unity of the popular diagnostic categories and the reliance on single measures as indexes of an abstract psychological concert are frustrating major advances in understanding. We need more courageous scientists to stand up and announce that the emperor is not wearing any clothes. A small number of investigators have criticized current practices, but their message is drowned out by the much larger number of clinicians and investigators who rely on popular practices. Barbara McClintock, a geneticist who studied maize, discovered in the 1950s that some genes changed their locations on the chromosome. A majority of the world’s eminent geneticists rejected her claim because they assumed the genome was stable. She persisted and won a Nobel Prize. Will the Barbara McClintocks of 2019 please stand up.
- K. A. Degnan, A. A. Hane, H. A. Henderson, O. L. Moas, B. C. Reeb-Sutherland, N. A. Fox. ‘Longitudinal stability of temperamental exuberance and social-emotional outcomes in early childhood’, Developmental psychology, 47, 2011.
- J. Frank, J. B. Frank. Persuasion and healing: A comparative study of psychotherapy, Johns Hopkins University Press, 1991.
- J. Kagan. Galen’s prophecy: Temperament in human nature, Basic Books, 1994.
- J. Kagan. N. Snidman, The long shadow of temperament, Harvard University Press, 2004.
- F. Kermode. Not entitled: A memoir, Farrar Straus and Giroux, 1993.
- P. Kim, G. W. Evans, E. Chen, G. Miller, T. Seeman. ‘How socioeconomic disadvantages get under the skin and into the brain to influence health development across the lifespan’, Springer, 21, 2018.
- M. P. McDonald. All souls, Beacon Press, 1997.
- J. L. Taylor, J. P. G. Debost, S.U. Morton, E. M. Wigdor, H. O. Heyne, D. Lal, D. P. Howrigan, A. Bloemendal, J. T. Larsen, J. A. Kosmicki, D. J. Weiner, J. Homsy, J. G. Seidman, C. E. Seidman, E. Agerbo, J. J. McGrath, P. B. Mortensen, L. Petersen, M. J. Daly, E. B. Robinson. ‘Paternal-age-related de novo mutations and risk for five disorders’, Nature communications, 10, 1, 2019.
- E. F. Torrey, R. H. Yolken. ‘Schizophrenia as a pseudogenetic disease: A call for more gene-environmental studies’, Psychiatric research, 278, 2019.
- J. W. Wideman. Brothers and keepers, Mariner Press, 1984.
- A. Zhang, C. Franklin, S. Jing, L. A. Bornheimer, A. H. Hai, J. A. Himle, D. Kong, Q. Ji, ‘The effectiveness of four empirically supported psychotherapies for primary care depression and anxiety: A systematic review and meta-analysis’, Journal of affective disorders, 245, 2019.