Mental health articles

The results of recent twin studies of mood and anxiety disorder symptoms across the life span suggest some differences in the etiology of these symptoms in childhood versus adulthood. Only a few twin studies hav been conducted with children and adolescents, so we will begin with a review of the more extensive adult findings. We discuss the potential theoretical implications of discrepancies between the child and adult findings concerning mood and anxiety disorder symptoms in terms of genotypeenvironment correlation. In this section, we will review univariate findings (concerning the etiology of a single symptom dimension) before discussing multivariate findings (concerning common etiological influences acting on two or more different symptom dimensions) for mood and anxiety disorders.

Univariate Studies of Mood and Anxiety Disorder Symptoms Adults.Kendler, Heath, Martin, & Eaves examined the causes of individual differences in anxiety and depressive symptoms in adults recruited from a large, nationwide twin registry in Australia. Symptoms of anxiety and depression were assessed by using a self-report questionnaire, the Delusions-Symptoms-States Inventory (DSSI). The authors found that a model including only genetic and nonshared environmental influences provided a satisfactory fit to the observed data, and the heritability for most symptoms ranged from .33 to .46. Several studies have been based on assessments of DSM-III, DSM-III-R, or DSM-IV affective disorders. In a sample of adult female twins recruited from a large population-based twin registry in Virginia, Kendler and colleagues (1992b) found that liability to major depression was moderately heritable (h2 = .20 to .45), and the remainder of the variance in liability was once again attributable to nonshared environmental inf luences. The authors later conducted analyses on follow-up data from the same sample that allowed them to control for unreliability of measurement and obtained an increased heritability estimate of around 70% for major depression. This suggests that a sizable proportion of the nonshared environmental influences in the earlier study was due to measurement error rather than specific environmental risk factors. Kendler et al. extended these findings to twins and to non-twin relatives recruited from the American Association of Retired Persons as well as the Virginia Twin Registry and found that additive genetic influences explained 30–37% of the variance in symptoms, with no evidence for shared environmental influences.

Consistent with these results from unselected samples, McGuffin, Katz, Watkins, and Rutherford obtained moderate estimates of the heritability for unipolar depression by using a variety of diagnostic criteria in a selected sample of hospitalized adult probands and their co-twins, where h2 ranged from .48 to .75. Once again, the authors found no evidence for shared environmental influences. The higher heritability obtained in the McGuffin et al. study may be due to sampling characteristics because hospitalized probands are more likely to be severely ill than individuals who manifest depressive symptoms in an unselected sample. Kendler et al. were interested in determining whether various alternative manifestations of depression would show different etiologies. They performed a latent class analysis of the DSM-III-R symptoms of major depression in a population-based sample of female twins. The analysis yielded three classes of clinically significant depressive symptoms, which were termed ‘‘severe typical depression’’ (distinguished by frequent weight loss and decreased appetite rather than weight gain and increased appetite, insomnia rather than hypersomnia, and frequent co-occurrence of anxiety and panic), ‘‘mild typical depression’’ (characterized by similar but less severe symptoms than ‘‘severe typical depression’’), and ‘‘atypical depression’’ (characterized by frequent hypersomnia, increased appetite, and weight gain, associated with co-occurring bulimia). They found that among twins concordant for major depression, concordance for the specific subtype was more common than would be predicted by chance. The concordance for specific subtype of depression was also greater in monozygotic than in dizygotic twins, suggesting a genetic basis for the observed patterns of symptoms. They concluded that depression is not etiologically homogeneous but may consist of subtypes that have differing etiologies.

Finally, several studies examined the causes of affective illness defined broadly (including both depressive disorders and bipolar disorders). McGuffin and Katz reanalyzed twin data previously reported by Bertelsen, Harvald, and Hauge and obtained a high heritability of .86, with a very modest contribution of environmental influences (c2 = .07; e2 = .07). Kendler, Pedersen, Neale, and Math´e  assessed genetic and environmental influences on affective illness (defined using DSM-III-R diagnostic criteria as either major depression [with or without history of mania] or bipolar illness) in a combined hospital-ascertained and population-based (i.e., unselected) Swedish sample. The hospital-ascertained sample was combined with the unselected sample to increase the number of participants with rare diagnoses such as bipolar illness without sacrificing the greater generalizability afforded by including nonreferred participants. The authors did not find that the estimated heritability of affective illness (either bipolar or unipolar illness) differed between the hospital-ascertained and populationascertained samples. They found that liability to major depression (with or without mania) was attributable primarily to genetic influences (h2 = .60) and the remaining liability variance was attributable to nonshared environmental influences and measurement error.

Potential explanations for the higher heritability found in this study, compared with the results obtained by Kendler et al., include the older average age of subjects in the Swedish study, the use of questionnaire versus interview-based assessment, and possible population heterogeneity in the variance component estimates. Liability to bipolar illness was more highly heritable (h2 = .79) than liability to major depression, and liability to affective illness (defined as either bipolar or unipolar illness) was intermediate in heritability (h2 = .64), the remaining liability variance in all cases was attributable solely to nonshared environmental influences and measurement error. Univariate behavior genetic findings for anxiety disorders suggest a pattern of causal influences similar to those found for depression.

Tags: anxiety, anxiety disorders, disorders, mood, mood disorders