Common genetic liability to major depression and posttraumatic stress disorder in men☆

1. Introduction 

Major depression (MD) and posttraumatic stress disorder (PTSD) are highly comorbid in epidemiological and clinical samples (Kessler et al., 1995, Orsillo et al., 1996). Pre-existing depression increases risk for both exposure to potentially-traumatic events and for developing PTSD among individuals exposed to such events (Breslau et al., 1998). The incidence of first-onset depression is elevated among individuals who develop PTSD following trauma compared to the trauma-exposed who do not develop PTSD (Breslau et al., 2000). Thus, the relation between MD and PTSD is multifaceted; MD appears to increase risk for PTSD and vice versa. However, the MD-PTSD association could also be explained by a common genetic vulnerability.

The twin method exploits the different level of genetic relatedness between monozygotic (MZ) and dizygotic (DZ) twin pairs to estimate the contribution of genetic and environmental factors to individual differences in the relation between two phenotypes, such as MD and PTSD (Plomin et al., 2001). Population covariance between MD and PTSD may be partitioned into an additive genetic component and two types of environmental components. The first is a shared environmental effect that is correlated between twins and has made siblings similar to each other. The second is an individual-specific environmental effect that is uncorrelated between siblings and includes measurement error.

This study is the first to use a twin design to quantify the degree to which a common genetic vulnerability explains the etiology of the association between MD and PTSD. Genetic influences explain a substantial proportion of the variance in MD (Lyons et al., 1998) and PTSD (True et al., 1993) individually and may explain their covariation. Moreover, emerging evidence supports a shared diathesis between anxiety and mood disorders. Dilsaver et al. recently demonstrated that a familial loading for mood disorders was associated with increased risk of a range of anxiety disorders. The authors also found a dose–response relationship between the number of comorbid anxiety disorders and a familial loading for mood disorders (Dilsaver et al., 2006). Prior research on a small subset of the sample used in the present study suggested that the familial association between MD and combat-related PTSD was genetically mediated, but this study was unable to quantify the extent to which MD-PTSD comorbidity is explained by a common genetic liability (Koenen et al., 2003). Genetic influences on both disorders overlap with those for generalized anxiety disorder (Chantarujikapong et al., 2001, Kendler, 1996) and with alcohol dependence (Fu et al., 2002, Xian et al., 2000). Given the attention and debate surrounding the etiology of the MD-PTSD association in the literature (Neria and Bromet, 2000), it is surprising the role of genetic influences on their comorbidity has not been quantified previously. The twin studies that have examined the genetic overlap between MD and anxiety disorders have not included PTSD (Hettema et al., 2006).

2. Methods 

3. Results 

The lifetime prevalence of MD was 9.2% (n=633). The lifetime prevalence of trauma exposure was 45.4% (n=3,065); 21.1% (n=649) of the exposed received a lifetime diagnosis of PTSD. Trauma-exposed subjects with lifetime MD were much more likely to meet criteria for PTSD (49.5%) than those without MD (16.5%; OR=4.98, 95% CI: 4.00–6.20).

3.1. Etiology of the association between MD and PTSD 

The best estimate of the phenotypic correlation between liability to MD and PTSD was .46. Table 1 presents the tetrachoric correlations for MD and PTSD by zygosity. The higher cross-twin cross-phenotype correlations among MZ compared to DZ twins suggested genetic influences on the MD-PTSD association.

Table 1. Tetrachoric correlation matrix for major depression (MD) and posttraumatic stress disorder (PTSD)

	MZ (n=1874)	MD twin 1	PTSD twin 1	MD twin 2	PTSD twin 2
	MD twin 1	1.00
	PTSD twin 1	.42	1.00
	MD twin 2	.42	.27	1.00
	PTSD twin 2	.25	.34	.44	1.00
	DZ (n=1498)
	MD twin 1	1.00
	PTSD twin 1	.48	1.00
	MD twin 2	.13	.07	1.00
	PTSD twin 2	.04	.22	.52	1.00

Abbreviations: MD, major depression; PTSD, posttraumatic stress disorder; MZ, monozygotic; DZ, dizygotic. Bold text, cross-twin, same phenotype correlations; Italic text, cross-twin, cross-phenotype correlations. p<.05.

Bivariate twin modeling was used to quantify the degree to which a common genetic liability contributed to the MD-PTSD association. The best-fitting bivariate model (Fig. 1) consisted of significant genetic and individual-specific environment factors influencing the covariation between phenotypes and provided excellent fit to the data (χ2(14)=14.70, ns; CFI=1.00; RMSEA=.005). Model fit did not deteriorate significantly if the shared environmental influences specific to MD and PTSD and those influencing their covariation were fixed to zero (▵χ2[3]=0.00, ns). However, model fit deteriorated significantly if additive genetic (▵χ2[1]=30.83, p<.001; CFI=.90; RMSEA=.04) or individual-specific environmental influences (▵χ2[1]=16.05, p<.001; CFI=.95; RMSEA=.03) were hypothesized to have no effect on the MD-PTSD association. A model where the genetic correlation between MD and PTSD was constrained to unity (rA=1.00; χ2(15)=17.94, ns; CFI=.99; RMSEA=.01) did not fit as well as one in which the genetic correlation was freely estimated.

View Large Image Download to PowerPoint

Fig. 1. Standardized variance estimates for the bivariate model of association between major depression (MD) and posttraumatic stress disorder (PTSD) in male twins. The model presents the standardized variance estimates of the AE model for MD and PTSD and their genetic (rA) and individual-specific environmental correlation (re). Amd and Emd refer to additive genetic and individual-specific environmental (including error) influences on major depression. Aptsd and Eptsd refer to additive genetic and individual-specific environmental (including error) influences on posttraumatic stress disorder. The model is displayed for twin 1 only; the model for twin 2 would look identical. The variances of the latent variables are fixed at 1. The parameter estimates were squared to determine how much of the variation in each phenotype is accounted for by latent factors. All parameter estimates are statistically significant p<.05.

Additive genetic influences and individual-specific environmental influences explain 40% and 60% of the variance in MD and 26% and 74% if the variance in PTSD, respectively. 24% of the variance in PTSD is explained by genetic and environmental influences common to MD. We also computed the amount of genetic variance in risk for PTSD that is common to MD, where rA is the genetic correlation coefficient and hptsd is the total genetic variance in risk for PTSD. Thus, [rA⁎(hptsd)1/2]2=15% of the total variance in risk for PTSD overlaps with the genetic risk for MD. Additive genetic influences common to MD account for 58% (15%/26%) of the genetic variance in PTSD. Individual-specific environmental influences common to MD explain only 8.5% of the total variance and 11% of the individual-specific environmental variance in PTSD. In addition, we can use the data in Fig. 1 to calculate the proportion of MD-PTSD covariance that is explained by a common genetic liability (15%/24%), 62.5%.

4. Discussion 

The overlap between MD and PTSD among middle-aged male twins was largely due to a common genetic liability that explained over 62% of the MD-PTSD covariance. In fact, genetic influences common to MD explained 15% of the total variation and 58% of the genetic variation in PTSD. Although it was not assessed in our study, the heritable personality trait of neuroticism may be a premorbid manifestation of this common genetic diathesis (Jang et al., 1996). Neuroticism has been associated with increased risk of developing MD (Khan et al., 2005) and PTSD individually (Breslau et al., 1995). Neurotism is moderately heritable and largely explains the genetic overlap between MD and other anxiety disorders such as generalized anxiety disorder and panic disorder (Hettema et al., 2006). The personality trait of neuroticism may also underlie MD-PTSD comorbidity.

The substantial genetic overlap between MD and PTSD implies that genes associated with MD are good candidates for PTSD and vice versa. Since PTSD requires exposure to a potentially-traumatic life event, genes shown to modify the effect of environmental adversity on risk of MD are particularly strong candidates for PTSD. Caspi et al. (2003) were the first to demonstrate that the short-version (s-allele) of a functional polymorphism in the promoter region of the serotonin transporter gene (SLC6A4) moderated the effect of adverse life events on risk of MD (Caspi and Moffitt, 2006). The s-allele has been associated with PTSD in a Korean sample (Lee et al., 2005).

A recent study by Kilpatrick et al. (under review) of adults exposed to the 2004 Florida hurricanes provides particularly compelling evidence that genetic variants shown to modify risk of developing MD in response to environmental adversity also modify risk of developing PTSD following trauma exposure. They found this same polymorphism in SLC6A4 moderated risk of both post-disaster PTSD and MD, but only under the high stress conditions of conditions of high hurricane exposure and low social support (Kilpatrick et al., 2007). In addition, polymorphisms in FKBP5, a glucocorticoid-regulating cochaperone of stress proteins, which were associated with recurrence of major depressive episodes and response to antidepressant treatment have also been associated with peri-traumatic dissociation, a risk factor for PTSD (Binder et al., 2004, Koenen et al., 2005). Further molecular genetic studies of PTSD are needed.

Although some of the same putative environmental risk factors have been linked to both MD and PTSD, our data indicate that the majority of their covariation arises largely from a genetically-transmitted process. Moreover, individual-specific environmental influences were only modestly correlated for both disorders. This implies that the environmental factors implicated in the etiology of MD and PTSD are largely disorder-specific. This result is consistent with previous research suggesting that life events involving threat are associated with anxiety disorders and those involving loss are more strongly associated with mood disorders (Eley and Stevenson, 2000). Further research is needed to identify environmentally-mediated risk factors that influence the development of MD versus PTSD.

This study has several limitations. Our participants were male Vietnam era veterans and our findings may not generalize to civilians, females or other cohorts. Our study relied on retrospective self-reported lifetime psychiatric diagnoses. Misclassification of MD and PTSD contributes to measurement error, resulting in an underestimate of the role of additive genetic factors and an overestimate of the role of individual-specific environmental factors in etiology. Thus, the contribution of genetic influences to the association between MD and PTSD in this study is likely to be an underestimate. Finally, our data are based on the DSM-III-R and results may not be the same under DMS-IV.

Our finding of a substantial common genetic liability to MD and PTSD has implications for theory, research, and practice. A common genetic liability to MD and PTSD is consistent with proposals for a quantitative-hierarchical approach to classification in DSM-V whereby PTSD is grouped with MD (and GAD) under the subcategory of internalizing-distress disorders (Watson, 2005). These findings also support R. Krueger's view that research on the etiology of psychiatric disorders will benefit from focusing on core processes underlying multiple forms of psychopathology rather than on discrete disorders (Krueger, 1999, Krueger et al., 2005). In practical terms, this means investigators interested in identifying causal genetic risk factors for common psychiatric disorders may benefit from using comorbidity to identify subtypes of disorder with a common etiology. Gene-hunters focused on PTSD need to carefully consider comorbidity in selecting controls for genetic studies. Given the MD-PTSD genetic overlap, the presence of MD in controls will attenuate an investigator's ability to detect a gene-PTSD association. Gene-hunting studies may benefit from distinguishing between comorbid MD-PTSD versus PTSD or MD-only in trauma-exposed samples. Finally, our data suggest that most of the environmental risk factors for MD and PTSD are disorder-specific. Prevention efforts will benefit from research aimed at identifying the environmental conditions that differentially shape their development in the context of common genetic liability.