Because diagnosis and definition is more subjective and interpersonal in mental than in physical illness, the estimates of the proportion of people with HIV vary according to how strictly disorders are defined. But nearly all surveys find higher rates of certain disorders than in the general population, particularly mood, substance abuse and anxiety disorders, as well as adjustment disorder after diagnosis or if treatment failure happens.

One 2002 study from the University of North Carolina, (Hooshyar) looked at 726 people with HIV and 215 co-infected with HIV and Hepatitis C receiving HIV primary care from the university's clinic, which serves a rural population. It found a 57.3 percent prevalence of depression among HIV-infected patients and a 69.8 percent prevalence of depression among patients co-infected with HIV and hepatitis C (HCV), compared with an estimated 5-10 per cent in the general adult population of primary care patients.

Substance abuse, which affects an estimated 8.3 percent of the US population ages 12 and up, affected 28.8 percent of HIV-infected patients in the study, and 47.9 percent of HIV/HCV-infected patients.

The 2002 “What do you Need?” survey (Anderson) of 1800 HIV positive people in the UK found similar figures. Only a quarter of respondents were happy with their current quality of life across all needs. Forty per cent each complained of anxiety and/or depression and sexual problems. Thirty-three per cent reported sleep problems, 32 per cent problems with self-confidence, 15 per cent with relationships and 14 per cent with drugs and alcohol, among other more social needs such as housing and childcare. Two-thirds related they had experienced at least one depressive episode over the past year.

These are exceptionally high figures, even for HIV patients, and not all studies have shown such high rates. The US study did not control for depression caused by drug use as opposed to endogenous depression in this relatively poor and rural population, many of whom has intravenous drug use as their HIV acquisition factor. The UK survey was a social rather than psychiatric one and may have encouraged people to report general states of dissatisfaction rather than diagnosable mental ill-health.

Not all studies have found such high rates. A pre-HAART study of patients referred to the psychiatric liaison service at the Royal Free Hospital in London and compared 70 HIV patients with 70 from the general; population. It found that five times as many people in the HIV patient population were referred as in the hospital’s general patient population. However it found that the prevalence of each group of psychiatric diagnoses was not significantly different between the HIV and control groups, except in the case of alcohol dependence, which was more common in the control group (21%) than the HIV group (4%), and drug use where, conversely, 44% of the HIV group and 30% of the control group fulfilled DSM criteria for a diagnosis of non-alcohol psychoactive substance abuse. A diagnosis of borderline personality disorder was made more often in the HIV group.

A study from San Diego (Chan) caught a positive change in people’s mental health during the introduction of HAART. 2466 patients from the HIV Cost and Services Utilization Study were interviewed first in December 1996 and January 1997 and then again eight months later, after a number had started on HAART. It graded their severity of psychiatric disturbance on four scales rating major depression, dysthymia, generalized anxiety and panic tendencies. It did not find such high levels of alcohol and substance dependence as the North Carolina study, with 6% reporting heavy drinking and 11% substance abuse. And it found that the average number of psychiatric symptoms in patients declined by 20% during the eight-month follow-up period of the study.

Nonetheless it found that although the proportion of patients reporting any psychiatric problems declined by 10% during the study, it did so from 47% to 37% - indicating a prevalence of psychiatric disturbance of any kind of about double that in the general population, where one in six are estimated to suffer from at least one of the four conditions measured at any one time.

One particularly interesting study which teased out the contributions of age and drug taking to poor mental health among men with HIV was performed by the US Veteran’s Administration (Justice 2004). It enrolled 1803 patients, 1047 of them with HIV and 756 without. Nearly all (97.65) were men, with 15 women apiece who were HIV positive and negative. The majority (51.55%) were African American, with 10% more in the HIV positive group. The HIV positive group also had a lower level of educational attainment and were slightly younger.

They asked the patients’ physicians to provide reports of mental health problems (major depression, dysthymia, generalized anxiety disorder and panic attacks) and drink and drug use using a standardized brief interview called the Composite International Diagnostic Interview (CIDI). They then took a select group of 28 HIV positive and 22 HIV negative patients from the larger group and subjected them to a battery of neurocognitive and psychiatric tests (NCP) and calibrated these against the results reported from their physicians.

This gave an interesting insight into the imprecision of mental health diagnoses. The psychometric tests gave very reliable results but were impractical to perform in most clinical settings, taking on average three hours and as much as six hours to perform. For depression the researchers found the CIDI interview to have good negative predictive value – it agreed with the NCP diagnosis of no depression in 89% of those who did not have it. However although it was reasonably sensitive, correctly diagnosing depression in more than three-quarters of patients who according to the NCP tests had it, it was less specific, incorrectly diagnosing depression in more than a third of patients who according to the NCP test did not have it.

For heavy alcohol use the CIDI interviews were specific, misdiagnosing only 11% of patients as having drink problems when they in fact did not, it was poorly sensitive, missing fully two-thirds of those actually did.

For drug problems the results were similar with CIDI and NCP diagnoses of people who actually did have problem agreeing 90% of the time but missing 50% of those who did use recreational drugs.

And for organic cognitive impairment, the CIDI interviews picked up on memory problems very poorly with only a half of patients reporting problems who in fact had them, though they were better at diagnosing those who did not have problems.

In the NCP testing the sample sizes were too small for differences between HIV positive and negative patients to reach statistical significance. However there was a trend for positive patients to have more current major depression (20% vs 7%, p=0.2).

By calibrating results from the larger group of patients it was found that although the raw data suggested that HIV positive patients had higher rates both of mental problems and drink and drug use these statistics confounded each other. Positive patients did not have higher rates of mental problems independently of drink and drug use; it appeared that they had more mental problems because of drugs and alcohol and vice versa, not in so far as they had HIV.. The statistical significance also disappeared when in multivariate analysis rates were corrected for ethnicity and educational attainment.

In multivariate analysis after weeding out these confounders it was found that HIV positive patients in general were still found to have higher rates of drug use. However the most significant finding was that the reason positive patients had higher overall levels of mental poor health and drink and drug use was because, unlike the HIV negative patients, these rates did not tail off with age.

Not unexpectedly, drink and drug abuse tends to tail off as people get older, Contrary to some popular stereotypes, so do rates of anxiety and depression. But they tailed off much less in the positive patients. For every 10 years older, the positive patients were 30% more likely than negative patients of the same age to be currently depressed; 24% more likely to report lifetime depression; 18% more likely to be heavy drinkers; and 46% more likely to continue using recreational drugs. Only cognitive and memory performance, though it got poorer with age, got no poorer than it did in negative patients.

This study has one major limitation; although it controlled for ethnicity, it did not control for sexual orientation. Other studies have shown that gay men have higher rates of drink and drug use than the general population (Cochran 2004) and that whereas as heterosexual’s drug using ‘careers’ tend to be over by their late 20s, gay men continue using drugs into middle age.

The question thus remains: were the researchers seeing a greater level of mental distress and drink and drug use in older patients because they had HIV – or because they were gay?

Is it HIV or sexuality?

Gay men (and lesbians) are known to have higher rates of depression, drinking and suicide than heterosexuals. One study (Cochran and Mays 2000) found that 19.3% of US gay men under 35 had at some point attempted suicide compared with 3.6% of exclusively heterosexual men (the ‘exclusively’ is important, as we will see in the next study).

One fascinating study (Paul) of suicide in gay men (which has been strongly linked to both depression and alcohol use) not only found much higher rates of suicidal thoughts and attempted suicide in gay men but also suggested some psychosocial reasons, rooted in society’s views of homosexuality, why they might feel suicidal.

The study was part of the Urban Men’s Health Study, a household probability-based sample of MSM (defined as either having had sex with a male since age 14 or self-identifying as gay or bisexual) in 4 US cities (Chicago, Los Angeles, San Francisco and New York).

It used demographics to identify areas with larger gay populations and then used random-digit dialing, weighted more towards areas with higher gay populations but not excluding areas with low ones, to call up gay male households. 2,881 men were interviewed using this sampling method, which while it may still exclude ‘closeted’ gay men who do not wish to identify as such to a researcher, at list avoids the pitfall of sampling only gay men who use the commercial gay scene or the internet, as many other studies do.

It found very high levels for both having at some point thought about suicide and having actually attempted it. Twenty-one per cent of the men interviewed had at some point made plans to kill themselves. This contrasts with figures of 8.87% and 14.66% derived from two studies of the general male population (the National Institute of Mental Health Epidemiologic Catchment Area study and the National Comorbidity Survey, both reported in Weissman).

There was an even bigger difference in actual suicide attempts. Twelve per cent of the men interviewed had attempted suicide at some point, of whom 45% (5.4% of the whole sample) had attempted it more than once. This compares with rates of 1.52% and 3.19% in the two studies quoted by Weissman. Two-thirds of suicide attempts (8.3%) had taken place before the person’s 25th birthday – this question was asked because suicide rates are higher among male teenagers than men in general.

HIV positive gay men had somewhat higher suicide rates than HIV negative men; 29% had had suicidal thoughts and 16% had attempted suicide (five to 10 times the rate of men in general). However the rate of attempts before 25 was not higher, which may confirm that the higher rates among men with HIV were caused by the knowledge of and problems attendant on HIV rather than HIV infection being caused by suicidality/depression.

Men who identified as bisexual also had higher rates – 30% had planned suicide, 16% attempted it and 11% had attempted it before they were 25, possibly indicating higher rates of sexual confusion.

However the most fascinating (and discouraging) finding of the study was that suicide rates in gay men had increased in the last 25-30 years and that this increase was almost exclusively among gay youth. Those who had been 25 or under in 1971 (i.e. at least 44 by the time the survey was done, in 1998) were 33% less likely to have attempted suicide at some point than men who were 25 in the 1970s; 42% less likely than those who were 25 in the 1980s; and 69% less likely than those who had been 25 in 1991 or later or were under 25 now. When it looked specifically at suicide attempt when the respondent was under 25, an even stronger relationship with current youth was found.

“This evidence seems to run counter to expectations regarding the effect of an emergent gay culture and the prospect of earlier self-identification as gay or bisexual,” say the researchers.

“Identifying as gay, lesbian, or bisexual at an earlier age simultaneously places youths at risk of victimisation while potentially cutting them off from social supports available to ‘mainstream’ adolescents and young adults (their communities or families). Other studies show that the risk of suicide is greatest at the developmental point when youths come to identify as gay or bisexual but have not yet told anyone. Similarly, it was recently reported that if one controls for stress, social support, and coping resources, no differences are found between gay, lesbian, and bisexual youths and heterosexual youths in terms of current suicidality. The increased suicidal risk in this age range appears to be not simply a mental health concern but rather a broader issue of the effect of societal discrimination and harassment.”

So some of the lifetime history of depression and suicidality in gay men may come from being a troubled gay teenager. However when it comes specifically to HIV, older gay men with HIV show higher rates of depression and other mental health problems than HIV negative men. One study of patients aged over 54 (not just gay men) attending St Mary’s Hospital HIV Clinic and referred to the Imperial College psychiatry service found that the older adults were more socially isolated, had more employment worries and more sexual problems (primarily sexual dysfunction).

Africans with HIV and mental health

However we have to be careful not to make the assumption that higher rates of mental ill-health reported in gay men with HIV actually correlate with a higher real prevalence. Because mental ill-health is subjective, it is also dependent on cultural assumptions and taboos about mental distress and talking to strangers about it, and also about professionals’ assumptions about presentation.

A study from West Middlesex Hospital (Malanda) found that sub-Saharan African patients with HIV were almost three times less likely to be referred for specialist mental health care than other patients. A case control investigation of those referred to mental health services showed that those Africans who were referred were more likely to be referred for assessment of possible organic brain disease, and more likely to be found to be suffering from major depression or organic brain disease. In other words, only the most gross and obvious levels of dysfunction resulted in referral.

The researchers considered reasons for the lesser likelihood of referral to the mental health service, including the possible failure of staff to recognize psychological morbidity in Black Africans, or reluctance and fear on the part of patients to be referred to services that may be perceived as threatening.

A 2003 study of African living with HIV in the UK (Project Nasah – Weatherburn) was designed to look at the treatment information needs of the community. Ironically it found that treatment information was the one need they felt was provided for best, with it coming second to the bottom of a list of suggested issues that had caused problems in the previous 12 months – only 27% felt they had problems accessing treatment information. In contrast the top seven concerns were mentioned by over half of all respondents. They were money worries, depression and anxiety, sleeping problems, self-confidence, immigration status, housing problems and relationships. Four of these may be fairly said to be mental health problems or at least ones to do with happiness and quality of life – depression and anxiety were mentioned by 71% as a problem, sleeping by 57%, self-confidence by 56% and relationship problems by 51%.

Of these, three (depression and anxiety, sleeping problems and self-confidence) were mentioned as often by Africans as by a representative sample of HIV positive white people, while Africans mentioned relationships as a problem twice as much. In contrast some other areas like money – and also HIV treatment information – were very much more likely to be mentioned as problems by Africans (only 4% of the white HIV positive people mentioned treatment info as a need compared with 68% who reported depression). This may say something about the ‘medicalisation’ of people with HIV and the fact that other needs may be neglected.