Disabled individuals in Sub-Saharan Africa may have an elevated risk of HIV infection, according to a recent meta-analysis.¹ When data from the 13 studies included in the analysis were pooled, the risk of HIV among persons with disabilities was elevated by about a third (relative risk, 1.3), and by about half when the analysis was restricted to men (1.5). However, the quality of studies was generally low, and associations were not consistently statistically significant for specific categories of disability.

Although one in seven people worldwide are disabled, such individuals are frequently overlooked by HIV programs. To examine whether disability is associated with HIV risk in Sub-Saharan Africa, Beaudrap and colleagues performed a systematic review and meta-analysis of studies measuring infection rates in disabled populations. In addition to searching a variety of databases for articles published between 2000 and 2013, the investigators looked for relevant data in Demographic and Health Surveys (DHS), AIDS Indicator Surveys and other sources; they also contacted experts and organizations that work on disability-related issues to identify other potential data sources, including unpublished work. Studies were eligible for inclusion in the analysis if they measured HIV prevalence or incidence among disabled adults in Sub-Saharan Africa; used a cross-sectional or cohort design; examined disability by type, disorder, activity limitation or need for support; and had a response rate of at least 50%. In additional to assessing eligibility, two reviewers assessed study quality, which was based on such factors as sampling method, use of objective criteria for disability, inclusion of a control group and use of an appropriate method of HIV testing.

Because few of the studies that met the eligibility criteria included a nondisabled control group, the investigators used contemporaneous data on HIV prevalence in the relevant country or region as a basis for comparison. They calculated risk ratios for HIV infection both for all disabled individuals and for those with specific categories of disabilities.

Of the 13 studies included in the analysis, eight had been published in peer-reviewed journals, two were DHS or national HIV studies, and three were unpublished survey reports. Most focused on mental or intellectual disabilities (six studies) or hearing disabilities (three studies); the remaining four examined all types of disabilities and in some cases reported data by disability category. Studies were conducted in South Africa, Cameroon, Uganda, Mali and Kenya. Six of the studies included fewer than 300 disabled individuals; only one included more than 700.

HIV prevalence among disabled individuals ranged from 1% to 29% in individual studies and was highly correlated with the prevalence in the local population. For example, prevalence was low (1%) in Senegal, where HIV infection in the general population is relatively rare, and highest in Uganda (11–18%) and South Africa (9–29%), where the epidemic is widespread. Overall, the estimated risk ratio for disabled individuals was 1.3; the risk ratio was slightly smaller (1.2) and nonsignficant when the analysis was restricted to the eight studies published in peer-reviewed journals. Analyses by sex revealed an elevated risk among males (1.5), but the risk ratio among females (1.3) fell short of statistical significance.

In four of the eight studies with data on HIV prevalence among individuals with mental illness or cognitive impairment, these disabilities were associated with an elevated risk of infection (relative risks, 1.6–2.2). However, in two other studies, persons with mental or intellectual impairments had a reduced risk of HIV infection (0.3–0.4), and the pooled data for the full set of studies did not yield a statistically significant association for individuals with these conditions. The four studies with data on hearing impairment and the four with data on disabilities of any kind yielded similar results: Data from one study in each category suggested that HIV risk is elevated among relevant individuals, but the pooled data did not reveal an association. In all of these categories, heterogeneity among studies was substantial.

All of the studies had methodological limitations. Only two used random sampling; just three used and described a control group; and many did not report confidence intervals or other indicators of precision. Overall, only two achieved a score of eight or better on the researchers’ nine-point quality scale. The meta-analysis itself also had limitations, the researchers note; these include the small number of eligible studies and the need to use data from DHS and other sources to create "control groups" for studies that lacked them.

Despite these issues, the authors conclude that, contrary to the assumptions of some stakeholders, "people with disabilities do not have a lower risk risk of HIV infection when compared to the general population." Because evidence from prior studies suggests that people with disabilities are less likely than others to obtain preventive services, testing and treatment for HIV, the authors recommend that HIV programs be "designed in collaboration with persons with disabilities," to ensure "effective inclusion of one of the world’s largest minorities."—P. Doskoch

REFERENCE

1. De Beaudrap P, Mac-Seing M and Pasquier E, Disability and HIV: a systematic review and a meta-analysis of the risk of HIV infection among adults with disabilities in Sub-Saharan Africa, AIDS Care, 2014, doi: 10.1080/09540121.2014.936820, accessed July 30, 2014.