Testosterone level, androgen receptor polymorphism, and depressive symptoms in middle-aged men

Abstract

Background: Testosterone (T) level declines progressively with age. Psychiatric symptoms of T deficiency (e.g., dysphoria, fatigue, irritability, low libido) are also symptoms of depression, and appear to be variably expressed.

Methods: We assessed independent measures of hypothalamic-pituitary-gonadal axis functioning, i.e., total T level and androgen receptor (AR) CAG repeat length (CAG RL), a genetic trait marker associated with AR function; and depression (diagnosed by above-threshold score on the Center for Epidemiologic Studies-Depression Scale [CES-D]) in 1000 men (mean age = 62.6 years; SD = 8.3) who participated in the Massachusetts Male Aging Study.

Results: There were 110 (11%) men with “depression” (CES-D score ≥ 16) in the analysis sample. Neither total T level nor CAG RL was associated with depression in bivariate analyses. Among men with shorter CAG RLs, the percentage of men with depression was 21.6% in the lowest subgroup of total T (defined by quintiles) and 4.2% in the highest subgroup of total T. This was confirmed in simple logistic regression models with depression as the dependent variable and continuous total T as the predictor, run separately within the three CAG RL subgroups: depression was significantly and inversely associated with total T in men with shorter CAG RLs but not in men with moderate and longer CAG RLs.

Conclusions: CAG isotype, a genetic trait marker of androgen receptor function, may mediate the expression of the central nervous system effects of T deficiency in men.

Introduction

Testosterone (T) level declines linearly with age, and approximately one fourth of elderly men have mild-to-moderate T deficiency Field et al 1994, Vermeulen and Kaufman 1995. Symptoms of profound T deficiency in young adult men include loss of libido, dysphoria, fatigue, and irritability (Wang et al 1996). The development of depressive symptoms in men with mild T deficiency is not well studied but appears to be variable Morales et al 2000, Sternbach 1998. Similarly, T administration to eugonadal men has psychiatric effects only in susceptible sub-populations Pope and Katz 1994, Pope et al 2000. Factors that may mediate the psychiatric effects of androgens are not known. We hypothesized that variability in androgen receptor (AR) transactivation may contribute to variability in the expression of androgen-mediated psychiatric symptoms. To test this hypothesis, we assessed the relation between AR isotype, total T level, and depression in a large community-based sample of middle-aged and elderly men.

T is the most potent and abundant androgen. It binds to the intranuclear AR, which is distributed throughout the body and the central nervous system. The AR is a typical steroid receptor: it contains an N-terminal domain, a DNA-binding domain, and a hormone-binding domain. The steroid-receptor complex binds to specific androgen-responsive sequences of genomic DNA and thereby influences messenger RNA production and modulates synthesis of a wide array of enzymatic, structural, and receptor proteins Lubahn et al 1988, Simental et al 1991, Simental et al 1992.

The AR gene has a polymorphic CAG microsatellite encoding variable-length glutamine repeats in the N-terminal of the AR protein Chamberlain et al 1994, Lubahn et al 1988. The CAG repeat length (CAG RL) ranges normally from 6 to 39 repeats, with a mean of 22 Edwards et al 1992, Krithivas et al 1999. Shorter CAG repeat length appears to be correlated with greater AR transactivation: recent studies have indicated that short CAG repeats are related to higher risk of prostate cancer, younger age at diagnosis, and poor response to endocrine therapy, as well as to increased growth of benign prostatic hypertrophy Bratt et al 1999, Chamberlain et al 1994. The relation between AR isotype and the symptoms of T deficiency has not been studied.

The Massachusetts Male Aging Study (MMAS) is a large, ongoing epidemiologic study of middle-aged men that includes T level, AR isotype (i.e., CAG RL), and the Center for Epidemiologic Studies-Depression Scale (CES-D). Previous MMAS analyses have not demonstrated an association between T level and CES-D-defined depression (Araujo et al 1998), though men with T ≤ 250 ng/dL, compared to all others, were significantly more likely to report low libido, erectile dysfunction, physical inactivity, sleep disturbance, gastrointestinal upset, and headache (unpublished data). In an initial report that utilized baseline and 9-year follow-up data, the magnitude of the 9-year T level decline (i.e., from baseline to follow-up) was inversely associated with the number of CAG repeats (e.g., total T decreased 0.74% ± 0.36% per CAG repeat decrement) (Krithivas et al 1999), suggesting that T levels are modulated by AR genotype. In this report, we assess the interaction between this AR genotype marker, total T level, and depression to test the hypothesis that this trait marker defines a vulnerable group in whom depressive symptoms are expressed when T level declines.