Deriving phenotypes for molecular genetic studies of substance use disorders: A family study approach

Abstract

Although, family, twin, and adoption studies indicate that genes play a significant etiologic role in the development of substance use disorders (SUDs), the specific genes involved have been difficult to detect due, in part, to uncertainties about how best to define SUDs, the possibility of genetic heterogeneity and the variable phenotypic expression of SUD genotypes. The goal of the present work was to determine if phenotypes external to the diagnosis of SUD such as psychopathology and cognitive functioning would show evidence of utility as phenotypes for genetic studies of SUD. We did this by applying factor analysis to multiple measures collected from our family-study program and then determining if these factors were heritable and were co-familial with SUDs. We used data from families recruited into six contemporaneous studies of four psychiatric conditions in children and adults. We found evidence for two SUD related phenotypes. One was an index of Psychopathology and Psychosocial Impairment; the other was an index of school failure and cognitive dysfunction. Both factors showed evidence of heritability, longitudinal stability and familial association with Parental SUD but these findings were stronger for the index of school failure and cognitive dysfunction. Results provide some support for the idea that candidate SUD phenotypes such as psychopathology and cognitive functioning, which are external to the diagnostic criteria for SUDs, may be useful for genetic studies of SUD.

Introduction

Family, twin, and adoption studies indicate that genes play a significant etiologic role in the development of substance use disorders (SUDs) (Tsuang et al., 1998, Merikangas et al., 1985, Merikangas et al., 1998, Mirin et al., 1986, Cloninger, 1987, Pickens et al., 1991, Rounsaville et al., 1991, Luthar and Rounsaville, 1993). These studies have led to the widely accepted conclusion that much of the familial transmission of SUDs is due to genes. However, the specific genes involved have been difficult to detect. Major obstacles to identifying genes for SUDs are the lingering uncertainties about how best to define SUDs, the possibility of genetic heterogeneity and the variable phenotypic expression of SUD genotypes.

It is likely that multiple genes, each of small effect, combine to cause SUDs. If so, they may not be detectable without reducing measurement error and creating measures that more directly assess the genotype and its consequences (Tsuang et al., 1993, Faraone et al., 1995). Furthermore, it seems unlikely that there will be a one-to-one correspondence between genetically influenced processes in the brain and the clinical phenomena that define diagnostic categories. Molecular genetic studies may thus be more fruitful if they focus on alternative phenotypes explicitly developed to maximize the power to detect genes.

Given the extensive comorbidity between SUDs and many psychiatric disorders such as depression (Tsuang et al., 2001), anxiety (Mills et al., 2006), bipolar disorder (Lin et al., 2006) and attention deficit hyperactivity disorder (McGough et al., 2005), it is intriguing to speculate that, when SUDs occur in the presence of psychopathology, that psychopathology may reflect the genetic susceptibility to SUDs. This idea is consistent with two bodies of evidence. First, many forms of substance use have been shown through twin and family studies to share genetic risk factors with several forms of psychopathology (Compton et al., 2005, Kendler et al., 2003). Second, two twin studies have shown that, although individual SUDs may have some unique sources of genetic etiology, much of the genetic susceptibility to SUDs can be explained by a common genetic diathesis (Tsuang et al., 1998, Kendler et al., 2003). These findings suggest that the genetic predisposition to SUDs may be indexed by clinical features external to the diagnosis of SUDs, which are typically associated with psychopathology such as symptoms and evidence of adaptive and cognitive impairments.

The goal of the present work was to determine if phenotypes external to the diagnosis of SUD such as psychopathology and cognitive functioning would show evidence of utility as phenotypes for genetic studies of SUD. We did this by applying factor analysis to multiple measures collected from our family-study program and then determining if these factors were heritable and were co-familial with SUDs. Our approach addresses the complex genetics of SUDs by using a comprehensive set of assessment measures; we sought to develop refined SUD phenotypic measurements that are informative for genetic studies. Following the recommendation of Weinberg et al. (1998) we make use of dimensional measurement approaches. Also, because drug use disorders are heterogeneous (O’Brien and Jaffee, 1992, Cadoret, 1991, Glantz and Pickens, 1992), we sought a design that would reduce the genetic heterogeneity of SUDs and facilitate the search for alternative phenotypes. To do that, we have focused our work on families selected for psychopathology.