Factors contributing to the sexual behavioural patterns and increased risk of HIV infection amongst migrant construction workers in Botswana

Ashby, Clive Norman
Journal Title
Journal ISSN
Volume Title
Background: HIV/AIDS has been one of the major crises to affect Southern Africa, particularly Botswana where prevalence rates have reached 37.3% of the adult population (2003). Due to the difficult working environment and long periods of separation from their partner, migrant workers have been highly susceptible to HIV infection and one of the main vehicles through which the virus has been transmitted. While much research has been carried out with mine workers, truck drivers, and other migrant groups, few studies have investigated the risk factors of construction workers, which form one of the largest employment groups in Botswana and Southern Africa. The purpose of this study was to determine which factors contributed to construction workers’ engagement in sexual risk behaviours, which have placed them at greater risk of HIV infection. Methods: A cross-sectional analytic study design was used. Structured interviews were carried out with individual migrant workers using a standardised questionnaire. 171 male migrant workers were interviewed, involved in skilled, semi-skilled, and supervisory professions. Interviews took place at three construction sites across Botswana, in the capital city of Gaborone and the rural village of Serowe. Bivariate and multivariate logistic regression analysis were carried out to determine which contributing factors were significantly associated with respondents’ sexual risk behaviours. Results: Unprotected sex was the most prevalent sexual risk behaviour, practised by 68.1% of construction workers. Significantly more workers engaged in unprotected sex with their long-term partner (70.3%) than with their casual girlfriend (35.1%). The second most common sexual risk behaviour was having multiple sexual partners. 57% of migrant construction workers reported having an extra girlfriend in addition to their long-term partner during the last five years. Transactional sex, involving the exchange of sex and material support, was a key part of most workers’ (76.6%) sexual relations with casual girlfriends. Commercial sex, though, was rarely reported and accounted for only 1.8% of workers. Men having sex with men (MSM) was strongly denied by construction workers, although 9.5% reported its occurrence in the workplace. In the multivariate analysis, migration was one of the most significant factors associated with respondents’ sexual risk behaviours. Compared to those who remained in one location during the last year, workers who migrated between work locations were 3.01 times more likely to have had transactional sex (p=0.013) and 4.42 times more likely to have had an extra girlfriend over the last five years (p=0.005). Workers who were separated from their main partner for a month or more at a time were 3.74 times more likely to have had an extra girlfriend in the last year (p=0.009) and 4.57 times more likely to have had transactional sex in the last five years (p=0.001). Workers who stayed in the construction on-site accommodation when away from home were 3.00 times (p=0.023) more likely to have multiple partners compared to those who stayed in private accommodation, where their partner had more opportunity to visit them. A second major contributing factor was respondents’ gender attitudes, particularly the perception that ‘one woman is not enough to sexually satisfy me as man’. Workers with this attitude were 6.21 times more likely to have currently multiple partners (p<0.001), 9.05 times more likely to have had an extra girlfriend in the last five years (p=0.015), and 3.35 times likely to have had transactional sex (p=0.031). A number of socio-demographic factors were significantly associated with sexual risk behaviours including respondents’ age, number of children, employment position, salary, workplace location, and education level. It is important to note, however, that respondents’ alcohol consumption and level of HIV/AIDS awareness did not significantly influence their sexual risk behaviours. Conclusion: These findings indicate that labour migration plays a central role in determining whether workers engage in sexual risk behaviours. Steps taken by employers to: (i) increase the frequency with which workers can visit their partner, (ii) provide facilities for long-term partners to visit the workplace, and (iii) reduce the frequency with which workers are transferred between sites - could significantly reduce workers’ susceptibility to HIV infection. Alongside migration, though, gender attitudes played a major role, pointing to the need for more education which focuses on gender attitudes and behaviour change rather than solely HIV/AIDS awareness.
HIV, sexual risk behavior, migrant workers, construction, Botswana