PART I.Syndromal Depression and HIV-1 (Summary Review)

There are eight studies in the literature that report the prevalence of current and/or lifetime major depression in HIV-infected adults. Each study used structured diagnostic interview instruments and Diagnostic and Statistical Manual of Mental Disorders, 3d ed. revised (DSM-IIIR), or ICD-10 criteria to define syndromal disorder. The results from these studies, which are summarized in table 1, indicate two general findings. The first is that none of the studies found an association of HIV-1 with lifetime depression, and only one found an association with current (1 month) depression (Baldeweg et al. 1993). In addition, none of the studies reported an association between either lifetime or current depression and early (presymptomatic) HIV-1 infection. The second finding is that, despite the general lack of association between major depressive disorders (MDD) and HIV-1, the rates for both current and lifetime depression in HIV-infected persons were significantly higher than the prevalence rates for depression in the general population reported in both the Epidemiologic Catchment Area (ECA) study (Regier et al. 1988) and in the more recent National Comorbidity Survey (NCS) (Kessler et al. 1994). The

average prevalence rates for lifetime MDD in HIV seropositive men (23.7 percent) was approximately fivefold higher than the average rate reported for men in the ECA (4.6 percent) and 1.8 times higher than reported for men in the NCS (13 percent). With respect to current depression, the observed rates were approximately 3.8 times higher than reported for 1-month ECA rates for men (2.3 percent). NCS rates for current major depression were available only for the past 12 months. Comparisons are presented for men only because the studies of HIV-1 included primarily well-educated, white, gay, male volunteers, which reflects the population most affected in the first wave of the disease. For example, in five of the studies the participants were described as gay or bisexual (Tross et al. 1987; Atkinson et al. 1988; Williams et al. 1991; Baldeweg et al. 1993; Perkins et al., in press). Given the population trends for this disease, it is very likely that the majority of the participants in these early studies were gay or bisexual. Despite the generally null findings regarding the association between MDD and HIV-1, one must note that few studies contrasted the spectrum of HIV-1 infection (Tross et al. 1987; Atkinson et al. 1988; Baldeweg 1993), while other studies pooled cases of presymptomatic and sympto- matic HIV-1 infection (Perry 1990; Pace et al. 1990). The pooling of early stage and advanced stage patients could attenuate the HIV-MDD association if the latter is more likely to be present in advanced cases. Also, most studies had small sample sizes, which restricts power to detect an association between these putative comorbid outcomes. The consistently high rates of MDD across studies, regardless of serostatus, raises the question of whether sexual orientation or other factors may be unexamined independent risk factors for major depression. Only one study (Atkinson et al. 1988) explored this hypothesis by including two small samples of noninfected gay (N = 11) and heterosexual (N = 22) controls. This study was the first to show an elevated rate of MDD in the gay and bisexual groups, independent of serostatus, suggesting that sexual orientation and lifestyle may be risk factors for major depression. It is also noteworthy that despite evidence of significant substance abuse among those at highest risk for HIV-1 infection (Donahoe 1990; Parker and Carballo 1992), none of these studies investigated whether the increased prevalence of depression in their samples may have been attributable, either directly or indirectly, to the widespread abuse of alcohol and other substances. TABLE 1. Prevalence (%) major depressive disorder (MDD) for

current (past month) and lifetime by studies versus ECA rates. Study

HIV-1

N

Sex. orient.

status Tross

Percent Current

SN

149

et al. 1987

ARC

40

AIDS

90

Atkinson

CONT

22

et al. 1988

SN

11

Percent

ECA

Life

ECA

4.7 G/B

15.0

2.2

No HIV effect

13.0

High % MDD

G/B

0.01

9.1

No HIV effect

9.1

HET

36.4

High % MDD in G/B

ASP

17

17.6

35.3

ARC

13

7.7

38.5

AIDS

15

6.7

13.3

SN

103

ASP

Perry1990

Comments

51

Mixed

4.9

30.1

3.9

2.2

25.5

No HIV effect 4.4

High % MDD Females higher % MDD

Pace

ASP

95

Mixed

5.8

2.8

5.3

7.0

et al. 1990

No HIV test No increase MDD

Williams

SN

84

4.0

33.0

et al. 1991

HIV+

124

G/B

4.0

2.2

32.0

3.0

No HIV effect High % MDD

Brown 1992

HIV+

442

Mixed

6.3

2.0

22.4

7.8

No HIV test High % MDD

Perkins

SN

71

et al., in

ASP

98

3.0

Baldeweg

SN

38

8.02

HIV effect

et al. 1993

ASP

59

3.0

(current)

SSP

48

G/B

32.0

8.0

2.2

28.0

No HIV effect 3.1

High % MDD

press

G/B

31.0

High % MDD

KEY: 1 = Heterosexual control; 2 = currrent equals 6 months.

In summary, these studies suggest that there does not appear to be a consistent association between HIV-1 serostatus and major depression, although groups at highest risk for this disease also appear to be at high concurrent risk for major depression. However, a more definitive test of the question of whether HIV-1 disease is associated with increased risk for depression will require larger study samples with adequate representation across the HIV-1 spectrum, including heterosexuals, gays, and bisexuals, to determine the independent and interactive effects of HIV disease and sexual orientation on major depression. Also, because most studies employed predominately welleducated, gay or bisexual, white male volunteers, results cannot be generalized to populations in developing countries, which now account for 85 percent of HIV-1 infection in the world and where primary transmission is through heterosexual intercourse (WHO 1992; Maj et al. 1994a). The results also cannot be generalized to ethnic minorities in the United States, especially to African Americans and Hispanic Americans, who constitute the groups at highest risk for infection in the second wave of the disease (Peterson and Marin 1988; Krueger at al. 1995). Finally, it is also possible that the widespread abuse of alcohol and other recreational drugs may enhance risk for both HIV-1 infection and major depression in high-risk groups (e.g., gays and bisexuals, intravenous (IV) drug users, and cocaine abusers). The next section reviews and reanalyzes data from two large cohort studies of HIV-infected men that afford an opportunity to investigate the hypothesized role of sexual orientation and substance abuse as cofactors in the increased prevalence of MDD in HIV-infected men.