Mental health in HIV-positive pregnant women: Results from Angola

Abstract

Our objective was to assess the mental health status of pregnant women who are HIV-positive, compared with other groups of pregnant women. We evaluated pregnant HIV-positive women attending the Medecins Sans Frontiers (MSF) HIV clinic in Malanje, Angola (N = 23). The control group consisted of pregnant women coming for antenatal clinic consultations who were not known to be HIV-positive (N=134). To assess mental health, the 12-item General Health Questionnaire (GHQ-12) was used. A score of three or greater was considered to indicate significant emotional distress. We also examined determinants of emotional distress in logistic multivariate regression models. We found that the mean score on the GHQ-12 for the HIV-positive group was more than twice the mean score of the controls, indicating poorer mental health in the HIV-positive group. Two-thirds of HIV-positive women had significant emotional distress, more than twice that in the control group. As well as HIV status, marital status was a strong independent predictor of mental health status, with married women experiencing less emotional distress. Thus, in our sample, pregnant women who were HIV-positive had a much poorer mental health status than the controls. Strategies to improve the mental health of HIV-positive mothers must be implemented and evaluated; efforts to decrease the levels of stigma and discrimination in this population are of key importance.

Introduction

It has been calculated that almost 25 million adults and children were living with HIV in sub-Saharan Africa at the end of 2005. During 2005, an estimated two million AIDS-related deaths occurred in Africa, leaving behind some twelve million orphans (Joint UN Programme on HIV/AIDS, 2004).

For many reasons, pregnant women represent a particularly important aspect of those living with HIV/AIDs (McIntyre, 2005). There have been very few studies regarding how HIV-positive status affects the psychological status of these women and the majority of previous studies have been done in developed countries (Larrabee et al., 1996; Swartz et al., 1998). The purpose of our study was to assess the mental health status of pregnant women who are HIV-positive compared with a control group in the sub-Saharan African country of Angola.

All participants gave informed consent to participate and local health authorities in Malanje sanctioned the study as part of the evaluation of HIV services in the region.

Methods

Our sample of pregnant HIV-positive women was comprised of all those who attended consultations at the MSF HIV clinic in Malanje, Angola (N=23). The control group consisted of pregnant women coming for antenatal clinic consultations at a rural health post in Malanje (N=134). Subjects who attended the clinics over a three-week period (May 30–June 22, 2005) were invited to participate in a structured interview and all but one individual (in the HIV-positive group) were recruited. With the given number of subjects, for a one-tailed test of the hypothesis that emotional distress is greater in the HIV-positive group compared to the controls, we would be able to detect a two-fold difference in the prevalence of emotional distress in the two groups (at an alpha level of 0.05) with a statistical power of over 95%.

To assess mental health, a brief mental health checklist—GHQ-12 (Goldberg & Williams, 1991)—was used. The GHQ-12 is a frequently-used and well-standardized measure of recent emotional distress containing twelve items related to ability to concentrate, loss of sleep, self-esteem, ability to make decisions, feelings of stress, ability to overcome difficulties, enjoyment of activities, ability to face problems, feelings of depression, loss of confidence, feelings of worthlessness and happiness. We pilot-tested the survey instrument among clinicians and healthcare services researchers to assess its face validity. We scored the response categories on the GHQ-12 items using the method recommended by Goldberg and Williams. As reported in the literature (Jesus & Williams, 1986; Paice et al., 2002; Snelgrove, 1998), we considered a score of three or greater as indicative of significant emotional distress.

Additional questions were asked of the HIV-positive women, regarding knowledge and practices related to HIV-positivity. Finally, we also examined determinants of emotional distress in a logistic multivariate regression model that included HIV status, number of pregnancies, duration of current pregnancy, age, marital status and income.

Results

Demographics and factors related to the current pregnancy were very similar between the HIV-positive group and the controls. The HIV-positive and control groups did not differ in terms of mean age (23.8 versus 22.9 years); number of previous pregnancies (three for both groups) or duration of current pregnancy (5.8 versus 5.6 months). A similar percentage judged that their current pregnancy was going well (58.3% in the HIV-positive group versus 47.7% in the controls). The women reported a number of occupations (in farming, retail etc.), with about a fifth of both groups indicating no occupation. There was a trend for better education in the HIV-positive women and their mean income was slightly higher, though not significantly (4021.74 Kwanza; US$ 45.70 per month in the HIV-positive group versus 3657.46 Kwanza; US $41.56 per month in the controls).

Of the HIV-positive women, WHO staging for HIV infection (Lifson et al., 1995) was as follows: 16 women (66.7%) were stage one (asymptomatic); seven women (29.2%) were stage two (minor symptoms, normal activity) and one woman (4.2%) was stage three (more symptoms, sometimes bedridden); none were stage four. Only four of these women (16.7%) were currently on Highly Active Anti-Retroviral Therapy.

The comparison of the GHQ-12 scores showed that pregnant women who were HIV-positive had much poorer mental health status than the controls. The mean GHQ-12 score for the HIV-positive group (mean 4.25±1.2; 95%CI: 3.13, 5.37) was more than twice that of the control group (mean 1.85±0.38; 95%CI: 1.47, 2.23). Two-thirds (66.7%) of HIV-positive women had significant emotional distress (95%CI: 47.7%, 85.7%), which was more than twice that of controls (30.6%; 95%CI: 22.8%, 38.4%).

The majority (87.5%) of the HIV-positive women indicated that they did not know the HIV status of their partner. Moreover, 42.6% had not disclosed their HIV status to anyone and only 37.5% had disclosed their HIV status to their current sexual partner. Of the women who reported having sexual relationships since they had tested HIV-positive (N=15), only three reported using condoms all of the time, with the majority of the remaining indicating that they never used condoms.

The results of the multivariate logistic regression model are presented in Table I. As well as HIV status, marital status was a strong independent predictor of mental health status, with married women experiencing less emotional distress. No other factor could be demonstrated as an independent predictor.

Table I. Independent effect* of variables in predicting emotional distress in pregnant women in Angola.

	Odds Ratio*	95%CI
Age (years)	1.04	0.94, 1.16
Positive HIV status	3.26	1.19, 8.93
Married	0.14	0.04, 0.47
Duration of pregnancy (months)	0.86	0.71, 1.04
Gravida (number of pregnancies)	1.04	0.76, 1.43
Monthly income (US dollars)	1.00	0.99, 1.01
*Adjusted odds ratio, outcome is based on emotional distress as indicated by a score.
≥3 points on the General Health Questionnaire (GHQ-12).

Discussion

AIDS is a formidable problem in sub-Saharan African countries, including Angola. The adult HIV prevalence rate in Angola has been calculated at 8.6%; by 2009, it is estimated that 430,000 adults in Angola will have died of AIDS, leaving behind 315,000 orphans (Joint UN Programme on HIV/AIDS, 2006).

In our sample, HIV-positive women scored significantly higher on the GHQ-12 scale than the controls, indicating poorer mental health. The finding that two-thirds of HIV-positive women had significant emotional distress, corroborates the findings from studies in other populations showing that the majority of HIV-positive pregnant women suffer from poor mental health (Kwalombota, 2002). A recent American study of depressive symptoms in HIV-positive women indicated the need for interventions to address such factors as personal support and social resources (Moneyham et al., 2000).

The GHQ-12 is a measure of recent emotional distress, asking how the subject has been ‘over the past few weeks’, prefacing the questions with the phrase ‘have you recently…’ (e.g. been able to concentrate, etc). The women in our sample were tested for HIV as part of their initial antenatal care and most of the women were in their sixth month of pregnancy or beyond. Thus, we do not believe the GHQ-12 scores reflect primarily acute stress linked to HIV testing and finding out the diagnosis; however, it is difficult to comment definitively.

Many women in our sample had not disclosed their HIV status to anyone; this may have been due to a fear of abandonment, as has been suggested in other studies (Lester et al., 1995). In our logistic regression model, married women appeared to experience less emotional distress. This is in keeping with a study of Asian HIV-positive women, where high depression scores were seen in women who were no longer married or in a relationship with their partner (Bennetts et al., 1999).

Our provisional interpretation of the findings would be that having a diagnosis of HIV-positivity likely leads to poor mental health in pregnant women and we believe that this is the most likely reason for our results. However, we admit that causality cannot be determined in a cross-sectional sample and therefore it is possible (as an alternate explanation) that women with low self-esteem and pre-existing high emotional distress may actually be most susceptible to risk taking and/or subsequent HIV infection (for the reasons mentioned or possibly due to other issues e.g. being single etc).

To summarize, HIV-positivity appears to have major implications in the lives of pregnant women. Our findings shed some light on the difficulties of being HIV-positive in rural Angola and how these factors might contribute to poor mental health. Specifically, a large number of HIV-positive women in our sample had not told anyone about their status; moreover, the lack of openness in communication was evident in the fact that most women did not know the HIV status of their current partner. Strategies to improve the mental health status of HIV-positive mothers must be implemented and evaluated; efforts to decrease the levels of stigma and discrimination in this population are of key importance (Bennetts et al., 1999).