Prevalence of Unsafe Sexual Behavior Among HIV-Infected Individuals: The
Swiss HIV Cohort Study

AIDS Journal of Acquired Immune Deficiency Syndromes August 1, 2003;
33(4):494-499

NATAP - www.natap.org

*†Katja Wolf; *James Young; ‡Martin Rickenbach; §Pietro Vernazza; ||Markus
Flepp; ¶Hansjakob Furrer; #Enos Bernasconi; **Bernard Hirschel; ††Amalio
Telenti; ||Rainer Weber; *†Heiner C. Bucher; Swiss HIV Cohort Study

Summary:

Sexual contact is the major mode of HIV transmission. Increased sexual risk
taking has been described in HIV-infected individuals receiving potent
antiretroviral therapy. A new questionnaire on sexual behavior was introduced into the
Swiss HIV Cohort Study on April 1, 2000. We evaluated sexual behavior in all
individuals who completed the questionnaire for the first time within 1 year
after its introduction. Our primary hypothesis was that self-reported unsafe
sexual behavior would be more prevalent among individuals with optimal viral
suppression. On April 1, 2000, 4948 individuals were registered in the study, and
4723 (95%) completed the questionnaire. Of these individuals, 12% reported
unsafe sex, 78% received antiretroviral therapy, and 25% had optimal viral
suppression (HIV RNA level always <50 copies/mL during the preceding 12 months).
During the preceding 6 months, 55% of individuals had stable and 19% had
occasional partners, and 6% had both types of partners. Sexual intercourse was
reported by 82% of individuals with stable and 87% of individuals with occasional
partners, and of those reporting sexual intercourse in each group, 76% and 86%,
respectively, said that they always used condoms. After adjustment for
covariates, reported unsafe sex was not associated with optimal viral suppression
(odds ratio, 1.04; 95% confidence interval, 0.81-1.33) or antiretroviral therapy
(odds ratio, 0.83; 95% confidence interval, 0.65-1.07), but it was associated
with gender, age, ethnicity, HIV transmission group, HIV status of partner,
having occasional partners, and living alone. There is no evidence that
self-reported unsafe sexual behavior is more prevalent among HIV-infected individuals
with optimal viral suppression. However, unsafe sex is associated with other
factors.

 

RESULTS

On April 1, 2000, 4948 individuals were registered and not known to have left
the SHCS. Of these individuals, 4767 (96%) had at least one follow-up visit
between April 1, 2000, and March 31, 2001, and 4723 (95%) responded to the
sexual behavior questionnaire. The percentage of females, intravenous drug users,
and individuals with only basic education was higher among those individuals
who did not respond than among those who did respond. Of those individuals who
responded, 55% had a stable partnership and 19% had occasional partners during
the preceding 6 months, and 6% had both types of partners. Of those
individuals with stable partners, 82% reported sexual intercourse, and of those
reporting sexual intercourse, 76% said that they always used condoms. Of those
individuals with occasional partners, 87% reported sexual intercourse, and of those
reporting sexual intercourse, 86% said that they always used condoms. Overall,
12% of the individuals reported unsafe sex, 81% denied unsafe sex, and the rem
aining 7% neither reported nor denied unsafe sex. Of those individuals who
responded, 78% received antiretroviral therapy, and 25% had optimal viral
suppression with viral loads of <50 copies/mL during the preceding 12 months.

In multivariate analysis, reported unsafe sex was not associated (P > 0.05)
with optimal viral suppression, antiretroviral therapy, diagnosis of an
AIDS-defining disease, or education . However, reported unsafe sex was associated
with gender, age, ethnicity, HIV transmission group, HIV status of the stable
partner, having occasional partners, and living alone. After adjusting for all
other covariates, the OR for reported unsafe sex in individuals with optimal
viral suppression was 1.04 (95% CI, 0.81-1.33). Males (OR, 0.58; 95% CI,
0.45-0.75), individuals aged 41 years or older (OR, 0.64; 95% CI, 0.50-0.80), and
individuals living alone (OR, 0.50; 95% CI, 0.39-0.64) were less likely to report
unsafe sex. Individuals from ethnic groups other than white (OR, 1.50; 95% CI,
1.07-2.08), intravenous drug users (OR, 1.73; 95% CI, 1.33-2.26), individuals
with HIV-infected partners (OR, 15.2; 95% CI, 12.1-19.1), and those with
occasional partners (OR, 4.04; 95% CI, 3.07-5.31) were more likely to report
unsafe sex.

With unsafe sex not denied as the response, there was less evidence of
associations with age and occasional partners and more evidence of associations with
education and antiretroviral therapy. Most ORs suggest that the nature of any
association was similar for both responses. For both reported and not denied
unsafe sex, ORs were lower for individuals receiving antiretroviral therapy
and for individuals with higher education. However, ORs differed between the two
responses for men having sex with men and for those with occasional partners.
Compared with other HIV transmission groups, men having sex with men were no
more likely to report unsafe sex (OR, 0.96; 95% CI, 0.71-1.31) but were more
likely to not deny unsafe sex (OR, 1.66; 95% CI, 1.32-2.10). Individuals with
occasional partners were more likely to report unsafe sex (OR, 4.04; 95% CI,
3.07-5.31) but were no more likely to not deny unsafe sex (OR, 1.18; 95% CI,
0.95-1.47).

We evaluated the interaction of gender and drug use because female drug users
may sell unsafe sex for drugs. As a replacement for gender in the
multivariate analysis, female drug users were more likely to report unsafe sex (OR, 2.12;
95% CI, 1.49-3.01) and not to deny unsafe sex (OR, 1.71; 95% CI, 1.29-2.29).
With this interaction included, gender was then not associated with either
response.

DISCUSSION

In this study of a large, well-described HIV-infected population, there was
no evidence of an association between unsafe sexual behavior and optimal viral
suppression. This is in contrast to other studies reporting increased rates of
unprotected sex among individuals taking potent antiretroviral therapy13 and
among those with suppressed HIV RNA.

This study showed that unsafe sexual behavior is relatively uncommon in
individuals of the SHCS compared with other studies of HIV-positive or HIV-negative
individuals. There is no evidence from this cohort study to support the
hypothesis that individuals taking antiretroviral therapy and those with optimal
viral suppression are more likely to have unsafe sex. Other researchers have
found that safer sexual behavior is related to adherence to antiretroviral
therapy, and this suggests that individuals who take care of themselves by adhering
to potent antiretroviral therapy are also more likely to take care of others
and protect them from infection.

We identified, however, additional factors that were associated with unsafe
sexual behavior. Individuals with HIV-infected stable partners were more likely
to report unsafe sex. A number of experts still recommend safer sexual
behavior if both partners are HIV positive to avoid transmission of a resistant
virus, but as yet there is little evidence to support this recommendation.18
Intravenous drug users and females were also more likely to report unsafe sexual
behavior. Female drug users were more likely to report unsafe sex and not to
deny unsafe sex, and when this factor was added into the multivariate analysis,
gender was then not associated with either response. It could be that female
drug users have difficulty negotiating condom use or sell unsafe sex for drugs.

Individuals of ethnic groups other than white were more likely to report
unsafe sex. In our data, the percentage of individuals with unknown ethnicity was
high (41%), but of those with unknown ethnicity, 93% gave their nationality as
a country in south or northwest Europe. This suggests that most individuals
in the category "white or unknown" were in fact white. Other investigators have
found that ethnicity is associated with unprotected sex in individuals with
HIV infection and that ethnic groups other than white may be at a higher risk
of acquiring HIV infection. There is evidence that in high-income countries,
HIV infection is moving into poorer and more deprived communities, including
ethnic minorities.

Men having sex with men were no more likely to report unsafe sex but were
more likely to not deny unsafe sex. This suggests a reporting bias. A second
reporting bias is likely for questions on occasional partners. Most of those
individuals who neither reported nor denied unsafe sex did not report occasional
partners. Unsafe sex with occasional partners is a concern since this behavior
may contribute to more rapid transmission of HIV infection. In several
countries, the incidence of sexually transmitted diseases is increasing.24, 25 In
Switzerland, the number of new cases of gonorrhea and chlamydial infection
reported each year increased between 2000 and 2002, from 5.8 to 7.4 and 32.2 to 43.3
per 100,000, respectively. Some countries that have a stable or decreasing
rate of HIV infection could therefore soon be facing a new increase in the rate
of HIV infection.

 

Our study has several limitations. First, information about sexual behavior
was self-reported, and patients were interviewed by their physician or study
nurse. Patients may have underreported unsafe sexual behavior because they are
expected to practice safe sex. Second, we have no information on the number of
partners, an additional risk factor for unsafe sexual behavior. Third,
individuals who responded to the questionnaire were different from those who did not
respond, which suggests that this study may underestimate the prevalence of
unsafe sexual behavior. Fourth, participants in the SHCS are intensively
followed by the study centers and may therefore not be representative of all
HIV-infected patients in Switzerland or elsewhere. Again this suggests that the study
may underestimate the prevalence of unsafe sexual behavior in the wider
community.

However, the study has several strengths. First, we considered not just
reported unsafe sex but whether unsafe sex was denied. This allowed us to explore
to some extent the possibility of reporting bias. Similar ORs were seen for
both responses, except for individuals with occasional partners and for men
having sex with men. Second, we included in our modeling a total of 10 confounding
variables. Since these variables tend to be correlated to some degree, missing
variables such as the number of partners are unlikely to affect estimates of
the relationship between unsafe sex and optimal viral suppression. Third,
although this study may underestimate the prevalence of unsafe sex, differences
between those individuals who respond and those who do not and between those in
the cohort and those outside will not necessarily affect estimates of the
relationship between unsafe sex and optimal viral suppression. We achieved a very
high response rate (95%), which makes this cross-sectional study highly
representative for those in the SHCS. Fourth, with such a large cohort, the power to
detect relevant differences is high.

CD4 cell count was not used in our analysis although it is a measure of the
success of antiretroviral therapy. Sexual behavior is a concern because of its
implications for HIV transmission, and hence plasma HIV load is a better
measure of successful antiretroviral therapy because it is directly linked to HIV
transmission and probably linked to the perceived risk of infectiousness.

In conclusion, the present study underlines the importance of epidemiologic
data on sexual behavior in HIV-infected populations. In the SHCS, 4 of 5
HIV-infected individuals report safer sexual behavior with their partners.
Individuals receiving potent antiretroviral therapy and those with optimal viral
suppression do not seem more likely to engage in unsafe sex. However, unsafe sex is
more likely in some subgroups of individuals with HIV infection. Sexual health
programs targeting these subgroups should complement programs aimed at the
general population.