In its search for aetiological factors, psychiatric epidemiology has worked mainly with psychosocial variables. The dominant paradigm has been that the causes of most mental disorders are likely to be found in the social environment and the experiences it imposes. A major contribution towards correcting this has come from Kendler and his collaborators. Instead of the general population, they used a large sample of adult twins from the Virginia Twin Registry. This enabled them to estimate how much of a disorder is attributable to genetic inheritance; how much to a common environment that both twins shared; and how much to each twins' unique environment. This method has generated findings of fundamental importance for aetiology.(9 91 and 92> It is in this context that the technical advances described below become highly attractive.
It has long been recognized by psychiatric epidemiologists that the addition of biological measures would be theoretically desirable, but these have not been highly practicable for administration to large numbers of persons in field surveys. The situation has recently changed and exciting new opportunities have become available. These lie in the unprecedented advances being made in molecular genetics. What has become available is a new set of predictor variables in the form of genotypes. It is not a matter of psychiatric epidemiology turning its back on psychosocial variables. Instead, genes can be assessed in their interaction with a full range of experiential and social factors. Two complementary strategies are being followed. The first is a continuing search for genes associated with discrete diseases such as bipolar affective disorder or schizophrenia. The second strategy is quite different: to search not for genes that may cause or be directly related to disorders, but for genes that confer vulnerability to them. Such an approach is well suited to epidemiological methods.
Cloninger et al.(93) have commented on the difficulties of replicating genetic associations with complex psychiatric disorders, and have argued that it may be more fruitful to map genes contributing to temperament, not symptoms. Within populations, personality traits are distributed as continua and probably involve many genes, each of which is neither necessary nor sufficient for the trait. In recent years, there has been increasing interest in genes that contribute to variation in quantitative traits. These quantitative trait loci (QTLs) may vary in the size of their effect on a trait from small to modest. The aim of current QTL research is to find the loci that have the largest effect. There are a number of techniques available for the study of QTLs, including linkage analysis in families, allele-sharing methods between relatives, and association studies in population samples. It is argued that allelic association studies are at present the strategy of choice for detecting QTLs. T.able.4 Because a large number of candidate markers are being investigated, there is a high risk of type I errors.
Table 4 The search for cause: a matrix for epidemiological studies
In the next few years, it can be expected that some personality traits will be found to occur more often in people with particular alleles in genes related to brain function. Amongst these, a preferred group of candidates are genes with known polymorphisms that alter the function of neurotransmitter systems, either by affecting the metabolism of a transmitter or some aspect of its function such as transport, receptor binding, or signal transduction. The attraction for psychiatric epidemiology is twofold: the promise of introducing to population studies a biological variable of fundamental significance; and the possibility of looking for interaction between biologically based vulnerability and life experiences. Although several studies reporting associations between personality traits and polymorphisms related to serotonin and dopamine metabolism have already been published, it would be premature to accept any association as established until replications have been achieved in sizeable samples across different populations.(94)
Other strategies for finding quantitative trait loci are emerging. One is to obtain general population samples not of individuals but of sib pairs or triads of first-degree relatives. Genomic scanning and gene expression technology are also now practicable. The prospects have been critically assessed by Henderson and Blackwood/95 but the field is moving so rapidly that interested readers will need to maintain close vigilance on the literature.
Was this article helpful?