In the long run...longitudinal studies of psychopathology in children. Committee on Child Psychology. Group for the Advancement of Psychiatry. Report no.143.

Longitudinal studies are difficult to do well. Too short a study and the results may be meaningless. Too long a study and the subjects cannot be found, money runs out, and research methods become seriously out of date. Despite these problems, there have been some longitudinal studies done that have greatly advanced our understanding of the nature and the treatment of psychopathology in childhood. Without these studies, much less would be known about the clinical course of important disorders, the effects of treatments, and the various risk and protective factors. None of these studies has been perfect. Some longitudinal studies did not focus on quite the right questions, some produced contradictory results, and others produced results that were hard to interpret. What have we learned from the longitudinal studies reviewed in this Group for the Advancement of Psychiatry (GAP) report? Many of the things that we have learned have been surprising--even counterintuitive. Pre- and perinatal insults need not necessarily lead to serious consequences in later life. Premature infants, if raised in nondeprived settings, are not likely to be mentally retarded or learning disabled. Today, of course, premature infants who would not have been kept alive 15 years ago are surviving. Will this advancement led to an untoward outcome? We do not know. New longitudinal studies need to be done. Certain serious illnesses emerging later in childhood may be associated with a greater risk of psychopathology. This risk is true at least for those with asthma. Psychological factors, such as psychological stress, also may lead to exacerbation of asthmatic attacks. Whether other illnesses are associated with a greater risk of psychopathology simply has not been studied adequately. Infant temperamental characteristics can be classified and measured; however, they appear to predict little in terms of later personality development or psychopathology. Although temperamental characteristics measured in infancy correlate poorly with temperamental characteristics measured at age 4 or 5 years, they appear to be much more stable and correlate well with temperament at ages 8 and 12 years. Some of these temperamental characteristics, particularly aggressivity and negativity at age 5 years, are serious risk factors for behavioral psychopathology at age 8 or 12 years and, possibly, at later ages as well. Aggressive, negative 5-year-olds do not necessarily grow out of their problems. Their problems are important and should be dealt with when children are age 5 years. What does appear important is early language development. Delays in language development or developmental language disorders are risk factors for later psychopathology. Fifty percent of school-age children with language problems requiring speech and language therapy also have diagnosable psychopathology. Hyperactivity and short attention span are characteristics that tend to persist in later childhood--even adulthood. Alone, each of these characteristics may not be a risk factor for later psychopathology, but, when accompanied by oppositional behavior and conduct disorder, the outcome is detrimental to the child. Many studies have shown that children with these comorbid problems are at risk later for becoming delinquent and antisocial. Conduct disorder in childhood, which is much more common in males, is a risk factor for later delinquency, even if it is not associated with hyperactivity. Again, early intervention is important. The effects of comorbidity in this instance can be serious. Comorbidity in general, especially as children grow older, leads to adverse social, academic, and psychopathological outcomes. Depression does occur in children, in all of its adult manifestations. Mania, in contrast, is rarely seen in children younger than age 12 years. Depression in school-age children can be severe or mild. In contrast to adolescents with depression, depressed children are most likely to have somatic compla