Sexual risk behaviour and HAART: a comparative study of HIV-infected persons on HAART and on preventive therapy in Kenya

Abstract

Summary

Unprotected sex (UPS) among persons receiving highly active antiretroviral therapy (HAART) remains a concern because of the risk of HIV-transmission. A cross-sectional study comparing the sexual risk behaviour of 179 people living with HIV/AIDS (PLHA) receiving HAART with that of 143 PLHA receiving preventive therapy (PT) with cotrimoxazole/isoniazid was conducted in Mombasa, Kenya. Forty-five percent of all participants were sexually active in the last six months. Participants receiving PT were more likely to report ≥2 partners (13% vs.1%; P = 0.006). Participants receiving PT reported more UPS with regular partners (odds ratio [OR]: 3.9; 95% confidence interval [CI]: 1.8–8.4) and also more sexually transmitted infections (STI) symptoms (OR: 1.7; 95% CI: 1.0–2.8; P = 0.059). More than 40% of all participants did not know the HIV-status of regular partners. Therefore, HAART was not associated with increased sexual risk behaviours though considerable risk of HIV-transmission remains. HIV-care services need to emphasize partner testing and consistent condom use with all partners.

Keywords

sexual behaviour HIV antiretroviral treatment sexually transmitted infections

INTRODUCTION

With an increased access to highly active antiretroviral therapy (HAART) there has been a dramatic decline in morbidity and mortality from HIV disease.¹ As of August 2006, 8800 HIV-infected persons were receiving antiretroviral therapy (ART) in Kenya.² Unprotected sex (UPS) among persons receiving HAART is of concern because of the risk of HIV-transmission to sero-discordant partners, possibly with resistant viruses or the risk of re-infection with new, drug resistant viral strains.³ An increase in risk behaviours has the potential to undo gains achieved by prevention and antiretroviral initiatives.

There is a widespread concern that a reduction in preventive behaviours may occur among HIV-infected persons once they feel better with ART. UPS and the incidence of sexually transmitted infections (STI) including HIV have increased among men who have sex with men (MSM) since HAART became more widely available.^4–5 Among HIV-infected MSM receiving ART, risk behaviour was associated with immunological and virological improvements, related in part to a perception of lower infectivity due to lower viral loads.⁶ An increased risk of acquiring STIs, an epidemiological marker of UPS, has been also reported among heterosexual HIV-infected persons receiving ART.⁷ A recent review found that while people's beliefs about lower infectivity with ART and undetectable viral loads promote UPS, HIV-positive patients receiving ART did not exhibit increased sexual risk behaviour, even when therapy achieved undetectable viral loads.⁸ Currently, limited evidence is available from resource poor settings.^9–11

Our study examines the sexual risk behaviour of HIV-positive persons accessing care and whether HAART is associated with increased sexual risk behaviour in Mombasa, Kenya.

METHODS

This cross-sectional study compares the sexual risk behaviour of HIV-infected adults ≥18 years of age who were receiving either HAART or preventive therapy (PT) with cotrimoxazole and/or isoniazid without HAART.

Study setting and subject recruitment

Participants receiving HAART comprised of adult HIV-infected persons completing six-months on ART; five persons refused the interview. Participants receiving PT comprised of HIV-infected persons completing at least five months on treatment; there were no refusals. Participants were recruited as they came for follow-up services.

Participants in both groups were in regular contact with health workers through monthly visits to the HIV-clinics. All patients received messages on positive prevention at the time of HIV testing, when starting their respective treatments and during follow-up; information provided consisted of routes of transmission, condom use, disclosure and partner testing. Those receiving HAART also had intensive counselling on treatment adherence.

Data collection

A structured questionnaire was used. Face-to-face interviews were conducted (September 2003 and June 2005) in English or Swahili as per patient preference. Interviewers were trained in rapport building and non-judgmental information gathering. The reference period for all measures was six months. Study participants provided a written informed consent. Ethical approval for the study was obtained from the Kenyatta National Hospital Ethics Committee and the Ethical Review Committee of the Population Council.

Study variables and measures

A regular partner was defined as a spouse or cohabiting sexual partner. A casual partner referred to a partner with whom the respondent had sex infrequently and was not living with or married to. A sex worker was a partner to whom money was paid in exchange for sex. Condom use at last sex referred to the most recent sexual act in the last six months. Consistent condom use was defined as always using condoms in the past six months; it was assessed as ‘always’, ‘sometimes’, ‘never’ and excluded last sex. UPS was defined as condoms not used at last sex or inconsistent condom use in the last six months. STIs were self-reported episodes of genital discharge (GD) or genital ulcer (GU) in the past six months; laboratory confirmation of STI was not available. Respondents were asked to report whether they knew their partner's HIV-status and whether they had disclosed their own HIV-status to their partners. Economic status was derived from type of housing and ownership of assets; the scores were categorized into quartiles.

Statistical methods

SPSS 11.0 (SPSS, Inc., Chicago, IL, USA) was used for statistical analysis. Chi-square tests were used to compare sociodemographic characteristics and behaviours between the groups. Logistic regression analysis was undertaken to determine predictors of sexual activity and UPS in the last six months. Variables found to be associated with sexual activity and UPS by having odds ratios (OR) that reached significance (P < 0.05) on univariate analysis were included in multivariate models.

RESULTS

Sociodemographic characteristics

We interviewed 179 HIV-infected persons six months after initiating HAART (first-line regimen: stavudine, lamivudine plus nevirapine or efavirenz) and 143 HIV-infected persons receiving PT with cotrimoxazole and/or isoniazid for median of six months (interquartile range [IQR]: 5, 7). Mean age of participants was 37 years (Standard deviation: eight years). There were no differences between groups with regard to age, marital status, sex or educational level (Table 1). Participants receiving PT were more likely to be employed (P = 0.007) and belong to higher economic status (P = 0.01) than those receiving HAART. For participants receiving HAART, the median CD4 cell-count at start of treatment was 100 cells/mm³ (IQR: 50, 152) and median increase in CD4 cell-counts over six months was 216 cells/mm³ (IQR: 128, 334). CD4 cell-counts were not available for those receiving PT.

Table 1

Sociodemographic characteristics of participants receiving HAART and participants receiving PT at HIV care clinics in Mombasa

	HAART% (n = 179)	PT% (n = 143)	P value
Age (years)
≤30	24 (42)	28 (40)	0.585
31–40	45 (81)	45 (64)
≥41	31 (56)	27 (39)
Sex
Male	35 (63)	34 (49)	0.862
Female	65 (116)	66 (94)	0.862
Education level
Never attended school	7 (13)	14 (20)	0.181
<7 years school	47 (84)	39 (55)
8–12 years school	40 (71)	41 (59)
College/technical/university	6 (11)	6 (9)
Marital status
Married/cohabiting	48 (86)	45 (64)	0.830
Single	12 (21)	12 (17)
Widowed/separated	40 (72)	43 (62)
Employment
Yes	18 (33)	32 (45)	0.007
No	82 (146)	69 (98)	0.007
Economic status
Lower ES (Q1)	18 (33)	7 (10)	0.010
Middle ES (Q2)	55 (99)	61 (87)
Higher ES (Q3 and Q4)	26 (47)	32 (46)

HAART = highly active antiretroviral therapy; PT = prevention therapy

SEXUAL ACTIVITY

Forty-five percent of study participants reported sexual intercourse in the reference period. There was no difference in self-reported sexual activity between participants receiving HAART and those receiving PT (44 and 47%; P = 0.476). Sex, education, employment and study group were not associated with sexual activity. On multivariate analysis, age and marital status emerged as independent predictors of sexual activity. Older participants were less likely to report sex than younger patients (OR: 0.94 per unit increase in age; 95% confidence interval [CI]: 0.91–0.97; P = 0.001). Married or cohabiting respondents were 8.3 times more likely to report sex than single/divorced/widowed respondents (95% CI: 4.96–14.14; P < 0.001).

SEXUAL PARTNERS

The majority of sexually active study participants reported sex with one partner (136/146). Patients receiving HAART were less likely to report multiple partners than those receiving PT (P = 0.006) (Table 2). The majority of respondents reported sex with regular partners. No male-to-male sex was reported. More participants receiving HAART reported sex with regular partners (P = 0.044) and fewer with casual partners (P < 0.001) than those receiving PT.

Table 2

Sexual risk behaviour variables by study group

	HAART% (n/N)	PT% (n/N)	P Value
Number of partners (n = 146)
1	99 (77/78)	87 (59/68)	0.006*
≥2	1 (1/78)	13 (9/68)
Type of partners (n = 146)
Regular	97 (76/78)	88 (59/68)	0.044
Casual	2 (2/78)	23 (15/68)	0.000*
Sex worker	1/78	1/68
Knowledge of partner status
Regular (n = 135)
HIV-positive	45 (34/76)	36 (21/59)	0.248
HIV-negative	13 (10/76)	24 (14/59)
Unknown	42 (32/76)	41 (24/59)
Casual (n = 17) ^†
HIV-positive	–	4/15
HIV-negative	–	2/15	–
Unknown	2/2	9/15
Sex worker (n = 2) ^†
Unknown	1/1	1/1	–
Disclosed status to
Regular (n = 135)	82 (62/76)	73 (43/59)	0.228
Casual (n = 17)^†	0/2	10/15	–
Sex worker (n = 2)^†	1/1	0/1	–
Condom use with
Regular (n = 135)
Last sex	93 (71/76)	77 (45/59)	0.012
Consistently	53 (40/76)	22 (13/59)	0.000
Casual (n = 17) ^†
Last sex	2/2	5/15	–
Sex worker (n = 2) ^†
Last sex	0/1	0/1	–

*Fisher's exact test

^†Absolute numbers, statistical testing not undertaken

Over 40% of respondents in both groups were unaware of the HIV-status of their regular partners. There was no difference between groups with regard to knowledge of partner's status or disclosure of HIV-status to regular partners (Table 2). Similarly, the majority of respondents in both groups did not know the HIV-status of and did not disclose their HIV-status to their casual and sex-worker partners (Table 2).

CONDOM USE WITH PARTNERS

Participants receiving HAART were more likely to report condom use at last sex (OR: 4.1; 95% CI: 1.37–12.28) and consistent condom use (OR: 3.9; 95% CI: 1.83–8.43) with regular partners than those receiving PT (Table 2). More participants receiving HAART reported consistent condom use with HIV-positive partners (56% vs. 14%; P = 0.004) and partners of unknown status (56% vs. 16%; P = 0.004) than those receiving PT. There was no difference between groups with regard to HIV-negative partners (30% vs. 42%).

Participants receiving HAART were more likely to report condom use with casual partners; there was no difference with regard to sex workers (Table 2).

SEXUALLY TRANSMITTED INFECTIONS

Participants receiving PT were more likely to report STI symptoms (22% vs. 13%; OR: 1.7; 95% CI: 0.98–2.81; P = 0.059) than those receiving HAART (GU: 27% vs. 18%; OR: 1.9; 95% CI: 1.41–3.45; P = 0.03; GD: 15% vs. 9%; P = 0.12).

Women respondents were more likely to report STI symptoms (30% vs. 8%; OR: 4.91; 95% CI: 2.33–10.31; P < 0.001) compared with men (GU: OR: 3.05; 95% CI: 1.43–6.51; P = 0.005; GD: OR: 22.310; 95% CI: 3.01–165.21; P < 0.001).

There were no differences between groups in seeking treatment (82% vs. 76%) and informing partners about the STI (59% vs. 58%).

FACTORS ASSOCIATED WITH UPS WITH REGULAR PARTNERS

Sixty percent of respondents reported UPS with regular partners in the last six months. Patients receiving PT were more likely to report UPS than those on HAART (78% vs. 47%; P < 0.001).

On univariate analysis, male respondents, married respondents and those receiving PT were more likely to report UPS. Partner's HIV-status, disclosure of HIV-status, number of partners, reporting an STI were not found to be associated with UPS. On multivariate analysis (Table 3), married or cohabiting respondents were three times more likely to report UPS with regular partners. After controlling for marital status and gender, patients receiving PT were four times more likely to report UPS than those receiving HAART (P < 0.001).

DISCUSSION

There has been widespread concern that providing ART to HIV-infected patients may lead to an increase in sexual activity and sexual risk behaviour. Fifty-five percent of our study population was sexually inactive; the proportions were similar among patients receiving ART and those receiving PT. Despite marked improvements in health status of patients on HAART (mean CD4 cell count increase: 239 cells/mm³) there was no difference in self-reported sexual activity between the two groups six months after treatment. Our findings are similar to those reported from Uganda⁹ and Cote d'Ivoire.¹¹

We found fewer multiple partners and fewer casual partners among PLHA receiving HAART compared with those receiving PT, consistent with findings from Cote d'Ivoire.¹¹ Overall, 12% of our sexually active participants reported casual partners; Bateganya (2005) report a much higher proportion (around 35% among ART-experienced and ART-naïve respondents).⁹ Our findings are also consistent with those reported from Uganda⁹ and Cote d'Ivoire¹¹ with regard to a higher proportion of participants on HAART reporting condom use with regular partners compared with those not receiving HAART.

Although sexual risk behaviour was lower among patients receiving HAART compared with those receiving PT, it is important to emphasize that risk of HIV-transmission remains. More than 40% of respondents in both groups reported regular partners of unknown status and between 13% and 24% of respondents reported HIV-negative regular partners. Overall, 60% of respondents reported UPS with regular partners, a third with casual partners and all with sex workers in the last six months. Importantly, almost half the respondents receiving HAART (14/32) and more than four-fifths (20/24) receiving PT who had regular partners of unknown status did not use condoms consistently and around two-thirds of respondents receiving HAART (7/10) and those receiving PT (8/14) who had HIV-negative regular partners did not use condoms consistently. Although disclosure rates to regular partners were slightly higher in patients receiving HAART, almost a fifth of the sexually active respondents did not disclose their HIV-status to regular partners, and more than two-thirds to casual and sex-worker partners. Lack of knowledge of partner's sero-status and low levels of disclosure of HIV-status, coupled with inconsistent condom use sets the stage for HIV-transmission to sero-discordant partners, especially in regular partner relationships. Bunnell (2006) reports from Uganda that 85% of risky sexual acts occurred within married couples.¹⁰ In such a setting the risk of HIV-transmission of resistant viral strains and re-infection with new strains poses a serious public health risk.^6–7 UPS can also carry the risk of unwanted pregnancy and subsequent HIV-transmission to the child.

STIs are often used as epidemiological markers of UPS. In our study, patients receiving PT were more likely to report STI symptoms in the last six months than patients on HAART. We also found that women were more likely to report STI symptoms than men. In the absence of confirmatory laboratory results, it is difficult to assess how many of the reported GD/GU were actual STIs.

Although prevention is stressed during counselling around HIV-testing and at the time of initiating ART, most counselling in HIV-care services is directed toward treatment adherence. Patients on HAART in this study received at least three preparatory counselling sessions on adherence followed by ongoing support. This emphasis on adherence may have contributed to the differences between groups. In addition, patients receiving HAART may perceive the seriousness of their illness differently, which could modify sexual behaviour.

Traditionally, the focus of HIV prevention programmes has been on high-risk groups. For HIV-positive persons’ counselling on prevention occurs mostly around HIV-testing, at the time of initiating ART and prevention of mother-to-child transmission (PMTCT) services for women. Prevention messages emphasizing sero-status disclosure, partner testing and consistent condom use with all partners are needed on an ongoing basis in HIV-care services.

This study has limitations. A cross-sectional study does not address the change in sexual behaviour over time. Although well-trained research staff, unlinked to the health facility, interviewed patients, recall and social desirability biases may have occurred. Our study used self-reports to elicit information on sexual behaviours; however, we feel that this does not unduly influence our results as almost 60% of sexually active respondents reported UPS in the last six months. Reviews of validity and reliability of HIV research have found that sexual behaviour data are fairly consistent and self-reported data on sexual acts and condom use are reasonably congruent especially for infrequent acts and relatively short recall periods.¹² We did not distinguish between types of sexual intercourse (vaginal or anal), but limited the enquiry to penetrative sexual intercourse.

The study design is strengthened by the fact that we had a comparison group and that the patients in that group (receiving PT) were exposed to health workers, received prevention messages and some form of treatment making them comparable with the HAART group and thereby minimizing biases.

In conclusion, we found that sexual risk behaviour is a concern among HIV-positive persons accessing HIV-care services although treatment with HAART is not associated with higher sexual risk behaviour. Prevention interventions targeting HIV-positive persons are needed on an ongoing basis in HIV-care services.

Table 3

Multivariate analysis of predictors of unprotected sex with regular partners (n = 135)

Variable	Univariate OR (95% CI)	P value	Multivariate OR (95% CI)	P value
Age (years)*
≤30 (n = 82)	1.0
31–40 (n = 145)	0.997 (0.455–2.204)	0.997
≥41 (n = 95)	0.892 (0.343–2.321)	0.892
Sex
Female (n = 80)	1.0		1.0	0.111
Male (n = 55)	2.097 (1.012–4.349)	0.047	1.934 (0.860–4.349)	0.111
Marital status
Single/widowed/divorced (n = 32)	1.0	0.003
Married/cohabiting (n = 103)	3.535 (1.547–8.081)	0.003	3.067 (1.251–7.519)	0.014
Education (years)*
≤7 (n = 68)	1.0	0.915
>7 (n = 67)	1.038 (0.520–2.072)	0.915
Number of sexual partners*
One (n = 127)	1.0	0.404
≥2 (n = 8)	2.009 (0.390–10.337)	0.404
Partner's HIV-status*
HIV-positive (n = 55)
HIV-negative (n = 24)	1.111 (0.414–2.981)	0.834
Unknown (n = 56)	1.030 (0.482–2.205)	0.939
Disclosure of own HIV-status*
Yes (n = 105)	1.0	0.348
No (n = 30)	0.675 (0.298–1.532)	0.348
STI*
No (n = 101)	1.0	0.585
Yes (n = 34)	1.253 (0.559–2.810)	0.585
HIV treatment
HAART (n = 76)	1.0	0.000	1.0	0.005
PT (n = 59)	3.932 (1.834–8.430)	0.000	4.269 (1.902–9.581)	0.005

*Variables not included in multivariate model

HAART = highly active antiretroviral therapy; PT = preventive therapy; STI = sexually transmitted infection

Footnotes

ACKNOWLEDGEMENTS

We thank the Ministry of Health, Government of Kenya for their support. We acknowledge the staff at Coast Province General Hospital, Port Reitz District Hospital, Bomu Medical Centre and Magongo Health Centre for their active support and participation. Lastly, we thank all study participants for their invaluable contribution.

FUNDING

Financial support for this study was provided by the President's Emergency Plan for AIDS Relief through the Office of HIV/AIDS, Bureau of Global Health, US Agency for International Development (USAID), through the Population Council's Horizons Program cooperative agreement of Award No. HRN-A-00-97-00012-00. The opinions expressed herein are those of the authors and do not necessarily reflect the views of USAID.

References

Crumm

, Riffenburgh

, Wegner

, Comparisons of causes of death and mortality rates among HIV-infected persons: analysis of the pre, early and late HAART eras. J Acquir Immune Defic Syndr 2006;41:194–200

Koech

. Current status of ART in Kenya. Concretizing pediatric, adolescent and adult ART scale up. Proceedings of the 4th National Consultative Conference on Antiretroviral treatment. September 2006. Nairobi, Kenya

Kozal

, Amico

, Chiarella

, Antiretroviral resistance and high risk transmission behaviour among HIV-positive patients in clinical care. AIDS 2004;18:2185–9

Chen

, Gibson

, Katz

, Continuing increases in sexual risk behaviour and sexually transmitted diseases among MSM. Am J Public Health 2002;92:1387–8

Stolte

, Dukers

, deWit

, Fennema

, Coutinho

. Increase in sexually transmitted infections among homosexual men in Amsterdam in relation to HAART. Sex Transm Infect 2001;77:184–6

Dukers

, Goudsmit

, deWit

JBF

, Prins

, Weverling

, Coutinho

. Sexual risk behaviour relates to virological and immunological improvements during highly active antiretroviral therapy in HIV–1 infection. AIDS 2001;15:369–78

Scheer

, Chu

, Klausner

, Katz

, Schwarcz

. Effect of highly-active antiretroviral therapy on diagnoses of sexually transmitted diseases in people with AIDS. Lancet 2001;357:432–5

Crepaz

, Hart

, Marks

. Highly-active antiretroviral therapy and sexual risk behaviour: A Meta-analytic review. JAMA 2004;292:224–36

Bateganya

, Colfax

, Shafer

, Antiretroviral therapy and sexual behaviour: a comparative study between ART-naïve and experienced patients at an urban HIV/AIDS care and research center in Kampala, Uganda. AIDS Patient Care STDs 2005;19:760–8

10.

Bunnell

, Ekwaru

, Solberg

, Changes in sexual behaviour and risk of HIV-transmission after antiretroviral therapy and prevention interventions in rural Uganda. AIDS 2006;20:85–90

11.

Moatti

, Prudhomme

, Traore

, Access to antiretroviral therapy and sexual behaviours of HIV-infected persons aware of their HIV-status in Cote d'Ivoire. AIDS 2003;17:S69–77

12.

Obermeyer

. Reframing research on sexual behaviour and HIV. Stud Fam Plann 2005;36:1–12