Introduction Highly active antiretroviral therapy (HAART) has been demonstrated to be effective in decreasing the incidence of opportunistic infections, AIDS, and death among patients with HIV infection. Herpes zoster (HZ) infection is common among patients with HIV infection. However, the impact of HAART on the incidence of HZ infection is not well understood.

Methods This nationwide, population-based, retrospective cohort study was conducted using Taiwan National Health Insurance Research Database (NHIRD) from 2000–2010. The NHIRD identified 15,112 patients with HIV infection. The incidence rates were standardised according to age based on the 2000 WHO standard population. Cox proportional hazards models were used to assess the effect of HAART on the incidence of HZ infection among patients with HIV infection.

Results The average incidence of the first episode of HZ after the diagnosis of HIV infection was estimated at 5.07 people/100 person-years. Multivariate Cox proportional hazards model showed that history of HZ infection (adjusted hazard ratio [aHR]: 6.76; 95% CI, 5.19–8.82) and administration of HAART (aHR: 1.53; 95% CI, 1.28–1.84) were the factors associated with HZ. Life-table method was used to divide the duration of HIV from diagnosis into two phases, namely, ≤4 years and >4 years. Initiation of HAART within 4 years of HIV diagnosis was associated with an increased risk of HZ (HR: 1.79, 95% CI: 1.48–2.16, p < 0.0001) and after 4 years of HIV diagnosis was associated with a decreased risk of HZ infection (HR: 0.60, 95% CI: 0.47–0.78, p < 0.0001). Among HIV-infected patients on HAART, with ≥85% adherence was showed significantly lower risk of developing HZ (HR: 0.40, 95% CI: 0.19–0.85,  p < 0.001).

Conclusion With high level of HAART adherence, it had significantly lowered HZ infection risk. Therefore, we suggested emphasising the importance of early treatment and HAART adherence.