Genetic perspectives on crime

This article examines the empirical evidence for a genetic influence in the etiology of antisocial behavior. This review relates the results of studies from 3 approaches to genetic investigation. The first, family studies, provides valuable information about the increased risk for deviance among the family members of affected individuals. Family studies provide few conclusions about genetic etiology, however, because members of families share environments as well as genes. A second approach, the study of twins, offers a somewhat better separation of genetic and environmental effects. The twin studies compare monozygotic (MZ) twins, who are genetically identical, to fraternal, same-sex, dizygotic (DZ) twins who have no more genes in common than other siblings (50%). The research design assumes that the effect of hereditary factors is demonstrated if the MZ twins have more similar outcomes (concordance for deviance) than DZ twins. The twins are reared together in almost all studies, and the environmental influences for MZ pairs may be more similar than for DZ pairs. A third approach, the adoption study, largely overcomes the possibility of confounding genetic and environmental factors which limit inferences from the results of twin studies. In this method, the deviant outcomes of adopted children (separated early in life from their biological parents) are compared with the outcomes of their adoptive parents and their biological parents. Similarity in outcome between adoptees and biological parents indicates a genetic effect.

A 30-year follow-up study of the sons of alcoholic men

The nature-nurture question in the etiology of alcoholism is discussed. The research results from twin and adoption studies indicate a considerable genetic (= biological) component in the etiology of alcoholism. A longitudinal high-risk study of alcoholism is presented. The sons of alcoholic men and matched controls have been followed prospectively since before birth. The main results from previous phases of the study and a recent 30-year follow-up assessment are presented.

Third ventricle enlargement and lower heart rate levels in a high-risk sample

Heart rate activity and computed tomographic measures of structural brain abnormalities were evaluated in 32 individuals with a genetic risk for schizophrenia (offspring of schizophrenic mothers). Heart rate activity was assessed in 1962 when the subjects were a mean age of 15.1 years. Diagnostic and computed tomography assessments were conducted in 1980. Compared to individuals with normal third ventricles, individuals with enlarged third ventricles evidenced significantly lower heart rate levels overall and significantly lower heart rate during rest and during the periods preceding conditioning and test for conditioning stimulus trials. These effects were independent of age, psychiatric diagnosis, and abnormalities in other brain regions. Difficulties in interpretation posed by the index of brain abnormality employed and by the 18-year time interval between the heart rate and computed tomography assessments are discussed. Together with prior evidence of a relationship between third ventricle enlargement and reduced electrodermal responsiveness in the same subjects, these findings provide a preliminary indication that enlargement of the third ventricle may involve damage to diencephalic structures involved in autonomic nervous system activity.

Adult third ventricle width and infant behavioral arousal in groups at high and low risk for schizophrenia

Previous research has demonstrated: (1) Subjects who are at genetic high risk for schizophrenia and who suffer delivery complications are at increased risk to evidence a widened third ventricle. (2) A widened third ventricle is related to decreased ANS arousal and, among schizophrenics, is related to negative symptom schizophrenia. (3) Adult schizophrenics evidence behavioral analogues of negative symptom schizophrenia premorbidly. This study compared adult CT scans to ratings of infant behavior in 179 subjects (104 at high genetic risk for schizophrenia) with the hypothesis that widened third ventricles would be related to underaroused infant behavior. Results of an ANOVA suggest that subjects who are at genetic high risk for schizophrenia and who evidence a widened third ventricle are more likely to have shown signs of behavioral underarousal as infants. Possible explanations, implications and limitations of the study are discussed.

Ventricular enlargement and premorbid deficits in school-occupational attainment in a high risk sample

Using a subsample of subjects from the Copenhagen schizophrenia high-risk project, this study examined the relationship between ventricular enlargement and prospectively assessed deficits in premorbid school/occupational adjustment. Subjects with enlarged ventricles evidenced poorer premorbid school-occupational adjustment across three times of measurement spanning a period of 10 years. This effect was independent of the age and sex of the subjects and remained significant after controlling for psychiatric diagnosis. The results provide prospective evidence for an association between ventricular enlargement and poor premorbid history.

Childhood antecedents of antisocial behavior: parental alcoholism and physical abusiveness

Hierarchical logistic regression was used to assess the independent and interactive effects of paternal alcoholism and physical child abuse on antisocial behavior in young adult men. Men with alcoholic fathers (N = 131) did not report or exhibit more antisocial behavior than comparison subjects (N = 70). Men with physical abuse histories, however, reported more aggressive and antisocial behaviors during a clinical interview and were rated by a clinical interviewer as more likely to act out aggression. Arrest records did not distinguish the groups. There was no evidence that paternal alcoholism and childhood victimization interacted to increase the risk of antisocial behavior.

Exposure to influenza epidemics during gestation and adult schizophrenia. A 40-year study

We attempted to replicate earlier findings of an association between exposure to influenza in the second trimester of gestation and adult schizophrenia. The number of live births, of births of future schizophrenics, and of cases of influenza reported to the Ministry of Health in Denmark was ascertained by month from 1911 to 1950. The relationship between fetal exposure to influenza and adult schizophrenia was examined. It is possible that unknown factors produce excesses of both influenza and schizophrenia in the winter, creating an artifactual association. To control for this coincidence, the effects of season were removed from the monthly influenza and schizophrenic birth-rates by several methods. Using the residual scores, it was found that influenza rates higher than seasonally expected, occurring in the sixth month of gestation, were associated with rates of births of schizophrenics greater than seasonally expected. This association was not attributable to some winter-related, third factor or to climatic variables.

Antecedents of predominantly negative- and predominantly positive-symptom schizophrenia in a high-risk population

We reanalyzed the Copenhagen schizophrenia high-risk project data set to test recently developed models of the antecedents of predominantly negative and predominantly positive forms of schizophrenia. Among a group of 138 high-risk individuals, those at elevated genetic risk who suffered severe delivery complications and who were autonomic nonresponders during adolescence were significantly more likely than those without this pattern to evidence outcomes of schizophrenia with predominantly negative symptoms (86% vs 0.8%, respectively). Among a group of 160 high-risk subjects, those who escaped delivery complications, who evidenced a high degree of autonomic responsiveness in adolescence, and who experienced severe disruption of the early family rearing environment were significantly more likely than those without this pattern to evidence outcomes of schizophrenia with predominantly positive symptoms (40% vs 1.2%, respectively). In late childhood and early adolescence, predominantly negative-symptom schizophrenics were rated by their teachers as passive, socially isolated, and unresponsive to praise; predominantly positive-symptom schizophrenics were rated as overactive, irritable, distractible, and aggressive. The study is limited by the fact that the hypotheses were based in part on previous analyses of the same data set, by the small number of schizophrenic subjects of each subtype, and by the use of simplified theoretical and statistical models that do not address the multidetermination of negative and positive symptoms.

Genetic and perinatal determinants of structural brain deficits in schizophrenia

Using a subsample from the Copenhagen schizophrenia high-risk project, we examined the contributions of schizophrenic genetic liability and perinatal complications to computed tomographic (CT) measurements of ventricular enlargement and cortical and cerebellar abnormalities. A factor analysis of six CT measurements yielded two significant factors. One factor reflected multisite neural deficits as evidenced by abnormality of the cerebellar vermis and widening of the sylvian and interhemispheric fissures and cortical sulci. The other factor reflected periventricular damage as evidenced by enlargement of the third and lateral ventricles. Because all of the subjects had schizophrenic mothers, the major source of genetic variation is contributed by the diagnostic status of their fathers. In a stepwise multiple-regression analysis, it was determined that the multisite neural deficits factor was significantly related to genetic risk for schizophrenia (as measured by schizophrenia spectrum illness in the subjects' fathers) but was unrelated to pregnancy or delivery complications or to weight at birth. Periventricular damage was highly and significantly correlated with the number of complications suffered at delivery, but only among subjects with an elevated genetic risk. Although limited by a small sample size, these results suggest that the two types of CT abnormalities in schizophrenia may reflect partially independent processes based on different combinations of genetic and perinatal influences.

Fetal neural development and schizophrenia

The conference on Fetal Neural Development and Schizophrenia which was held in Washington, DC, May 31-June 1, 1988, focused on factors of possible etiological significance in fetal development. Schizophrenia researchers joined experts in brain imaging, neuropathological, and neurochemical changes in brain development and investigators of potential genetic and neurobehavioral causes of psychosis. The combined evidence suggested dysfunction in frontal and parieto-occipital neocortex, basal ganglia, hippocampus, and amygdala. Dopamine transmission was implicated both in basal ganglia deficits and in widespread neocortical disturbances. Viral infection, or excessive stress, during the second trimester of pregnancy, as well as obstetrical complications, minor physical anomalies, and brain defects, correlated positively with incidence of adult schizophrenia. Autonomic nonresponding, birth complications, and ventricular enlargement were found to be closely related to negative symptom schizophrenia in high-risk populations. A dual factor model of schizophrenia was suggested, where genetic and environmental influences combine to produce psychosis.

Minor physical anomalies and recidivistic adult violent criminal behavior

Minor physical anomalies (MPA) result from disruptions of gestation, and may be used as signs of central nervous system defects in development. Utilizing a Danish birth cohort, we tested the hypothesis that MPA predict adolescent and adult recidivistic violent criminal behavior. The number of MPA was measured at 11 to 13 years of age and police records of criminal behavior were ascertained at 20 to 22 years of age. Recidivistic violent offenders evidenced an elevated level of MPA compared with subjects with one violent offense or subjects with no violent offenses.

Pattern reversal visual evoked potentials after alcohol administration among men at risk for alcoholism

The P100 component of the pattern reversal visual evoked potential was used to compare men at high risk for alcoholism and control subjects before and after a low (0.5 g/kg) dose of ethanol. The high risk and control subjects did not differ in age, self-reported ethanol consumption, or estimates of ethanol metabolism rates, but changes in the occipital P100 latency differentiated them following ethanol administration. The P100 latency changes that distinguished high risk from control subjects were lateralized and provide preliminary evidence that perceptual visual stimulus processing is differentially affected in the two groups following ethanol administration.

EEG identification of subgroups of men at risk for alcoholism?

Biological sons of male alcoholics constitute one group at high risk (HR) for the development of alcoholism, and were the subjects of this study. A low dose of alcohol (0.5 g/kg) was administered to HR and control subjects. On the basis of changes in the electroencephalographic (EEG) mean alpha frequency that occurred following alcohol administration, two HR subgroups were identified. Measures obtained after alcohol administration, comprising self-ratings and an observer's assessment, distinguished the HR subgroups and control subjects; measures of visuomotor performance did not. The findings are discussed in relation to two current etiological theories bearing on the development of alcoholism: a biopsychological perspective and an initial tolerance for alcohol effects.

Pattern reversal visual evoked potential among men at risk for alcoholism

The biological sons of male alcoholics, deemed to be at high risk (HR) for the development of alcoholism, were compared to control males, aged 18 to 21, using measures of the visual evoked potential elicited by checkerboard pattern reversal. Overall, the HR and control groups were not distinguished on the basis of visual evoked potential measures acquired from the occipital scalp region; however, when comparisons were restricted to right-handed subjects, the HR subjects showed more symmetry in a positive component with approximate latency of 242 ms compared with control subjects. The results are discussed in relation to hemispheric differences and alcoholism.

Temporal course of symptoms and social functioning in relapsing schizophrenics: a 6-year follow-up

As a part of the Copenhagen High Risk project, a 6 year follow-up of relapses after first hospitalisation of 67 schizophrenic women was performed. Four dimensions of psychopathology were examined: positive symptoms, negative symptoms, formal thought disorder and depressive symptoms. Each symptom group was its own best predictor over time. Only depressive symptoms diminished significantly at a relapse 6 years after first hospitalisation. Broadly defined subtyping of schizophrenia into paranoid and non-paranoid, retained stability over a 6 year period. It is concluded that there is a stability of psychopathology across onset episodes and relapses.

Motor impairment and antisocial behavior in adolescent males at high risk for schizophrenia

A familial link between schizophrenia and antisocial behavior has been established (e.g., Silverton, 1985). This study examined this relationship in a Danish cohort. The subjects were 36 high-risk males (offspring of a schizophrenic parent) and 36 low-risk males (offspring of parents without psychopathology). This high-risk subjects exhibited more antisocial behavior than the low-risk subjects. We tested the hypothesis of a correlation between neurointegrative deficits, as defined by motor impairment, and antisocial behavior, rated at ages 10-13, in subjects at genetic risk for schizophrenia. Path analyses were conducted from motor impairment at 1 year and motor impairment at 10-13 years to antisocial behavior separately for high-risk and low-risk subjects. Adolescent motor impairment was a significant predictor of antisocial behavior for high-risk subjects. Motor impairment at 1 year was also associated with antisocial behavior for these subjects, although the association was partly due to the indirect effects of motor impairment at 1 year on motor impairment at 10-13 years, which, in turn, was associated with changes in antisocial behavior. As predicted, none of the path coefficients nor the effect coefficient was significant for low-risk subjects.

Adult schizophrenia following prenatal exposure to an influenza epidemic

In the context of a Finnish birth cohort, we tested the hypothesis that viral infection during the latter two thirds of fetal development would increase the risk of adult schizophrenic outcome. Psychiatric hospital diagnoses were recorded for all individuals in greater Helsinki who were fetuses during the 1957 type A2 influenza epidemic. Those exposed to the viral epidemic during their second trimester of fetal development were at elevated risk of being admitted to a psychiatric hospital with a diagnosis of schizophrenia. This was true for both males and females and independently in several psychiatric hospitals. The second-trimester effect was seen in the elevated proportion of schizophrenics among those admitted to a psychiatric hospital and also in higher rates of schizophrenia per 1000 live births in the city of Helsinki. The study has several limitations: (1) We have no direct evidence that the subjects actually suffered a viral infection. (2) The psychiatric data were obtained only for subjects up to the age of 26 years, 56 days. (3) The findings are based on hospital diagnoses. (4) The determination of stage of gestation at time of exposure to the epidemic is based on date of birth. The viral infection might have occurred outside the official epidemic window; the infant may have had a preterm or postterm delivery. These sources of error, however, should not serve to enhance the findings. The observed viral effect is interpreted as being one of many potential perturbations of gestation. We suggest that it is less the type than the timing of the disturbance during fetal neural development that is critical in determining risk for schizophrenia.

The Copenhagen high-risk project. The diagnosis of maternal schizophrenia and its relation to offspring diagnosis

The Copenhagen longitudinal high-risk study of offspring of 129 schizophrenic mothers commenced in 1962. At that time, the mothers were diagnosed according to contemporary Danish criteria. We have re-examined all of the hospital records of these mothers: 108 (84%) fulfil present-day DSM-III criteria for schizophrenia and 95 (74%) were diagnosed as paranoid schizophrenic according to ICD-8 criteria. In a follow-up at mean age 24, the offspring of the paranoid schizophrenic mothers were themselves found to be less frequently schizophrenic (5%) than were the offspring of non-paranoid schizophrenic mothers (29%).