Human Immunodeficiency Virus (HIV) patients are now living longer than before [1, 2]. This is mainly attributed to access to free combination antiretroviral therapy (cART) [2] or combinations consisting of a minimum of two active drugs from two classes of antiretroviral agents; usually they contain more different active drugs and are referred to as HAART. The improvement in life expectancy and quality of life observed among cART recipients could have been even better had these drugs not been associated with illnesses such as diabetes and diabetes-related comorbidities (DRC) [3, 4] or any comorbidity associated with Type 2 diabetes [5].

Association between cART and DRCs has been long suspected [4, 6, 7]. Studies have provided evidence showing that this association may depend not only on the duration of exposure to cART [8] but also the type of cART regimen taken [4]. In other words, while cART is considered a providential life-saving treatment for persons living with HIV (PLWH), it also has the potential of taking away lives by exposing recipients to an array of unwanted illnesses like diabetes mellitus and related complications [3]. This is a disturbing situation that makes the task of controlling and preventing morbidity and mortality among HIV patients more difficult, especially in settings with a high prevalence of HIV like in Botswana [9].

The initiative to provide free cART in Botswana started in 2002 [9]. Recipients of cART have access to standard treatment as defined by the Botswana national HIV & AIDS treatment guidelines [10] and the handbook of the Botswana integrated HIV clinical care guidelines [11] in use between 2002 and 2015. Components of cART regimens in Botswana were as presented in Appendix and described by Rankgoane-Pono and colleagues’ [12]. All these regimens have been to some extent, in different settings, associated with the development of DRCs after some time of exposure [6, 7, 13, 14]. Some experts have reported these associations only with regimens containing protease inhibitors [13], while others have described DRCs association with all types cART regimens [6, 7, 14]. The average time duration from cART initiation to the development of DRCs also significantly differed from one study to another [8, 14]. However, these results are perhaps due to differences in patient characteristics. For example, the patients’ underlying biomedical and demographic risk factors of DRC, such as being overweight or having CD4 cell count below 200 cells/mm³ have the potential to alter outcomes among patients if not well controlled [14]. Attributes associated with the risk of developing DRCs need to be identified earlier before patients’ enrolment in cART programs. This will allow for suitable care which minimizes cART adverse effects and delays or completely prevents DRC development. The availability of models highlighting DRCs’ significant prognostic factors is important to guide clinical care, policy and management of recipients of cARTs in a setting.

The study aimed to: (i) examine cART relationship with DRCs among patients attending HIV clinics in Gaborone, Botswana (which cART regimens are associated with shorter/longer time to the event), (ii) characterize patients’ underlying biomedical and demographic risk factors of DRC and identify the most important (those that significantly affect time to the event), (iii) investigate survival of patients on different cART regimens in the presence of these risk factors.

Nine of the 540 records that were reviewed did not meet the inclusion criteria and were excluded. Thus, 531 patients’ records were included in the analysis. Of these, 368 (69.3%) were females. The mean age [± standard error of the mean (SEM)] of participants was 41.4 ± 8.9 years. The youngest was 19 years old while the oldest aged 82.0 years. The mean (± SEM) weight of patients at cART initiation was 60.6 ± 11.8 kg with the lowest record of 17.5 kg and 101.0 kg as the highest. After cART initiation the mean (± SEM) weight documented was 67.9 ± 14.50 kg, with the minimum of 32.5 kg and 117.7 kg as the maximum. The mean CD4 count (± SEM) at cART initiation was 139.5 ± 5.11 cells/mm³, the minimum count was 0.00 cells/mm³ and the maximum was 889.0 cells/mm³. After cART initiation, the mean CD4 count was 536.0 ± 10.16 cells/mm³ with the minimum count of 25.0 cells/mm³ and the maximum of 1441.0 cells/mm³. At cART initiation, 408 (76.6%) participants had a CD4 count of ≤200 cells /mm³compared to 34 (6.4%) after the initiation of cART. Four hundred and forty two patients or 83.2% did not develop any DRC, whereas 89 (16.8%) of them developed DRCs.

Patients on first-line cART were 2.03 times more at risk of developing DRCs earlier than those who were not on the first-line cART [UHR = 2.03; (95% CI: 1.26–3.27)]. Patients on second-line cART had a 52% less risk of developing DRCs earlier than patients who were not on the second-line cART [UHR = 0.48; (95% CI: 0.30–0.79)]. Those on third-line cART had 5% less risk of developing a DRC earlier than patients not on the third-line cART but the relationship was not statistically significant [UHR = 0.95; (95% CI: 0.13–6.83) and p = 0.9]. When patients on second-line cART were aggregated with those on third-line (cART line2/3), recipients showed 51% less risk of developing DRCs earlier than patients who were not on cART line2/3 [UHR 0.49, (95% CI 0.31–0.79).

UHR for CD4–1 and CD4–2 count as continuous variables are not shown here as both variables did not achieve significance in the exploratory two independent samples t-test analysis. Age as a continuous variable indicated that the hazard of developing DRCs increased by 3% [UHR = 1.03; (95% CI: 1.01–1.05)] for every one year increase in recipients age. Of all the characteristics investigated, seven showed significant crude associations with DRCs (p < 0.05) while others achieved p ≤ 0.9 and were considered for multivariate analysis, details on these results are given in Table 2.

HIV patients on second- and third-line cART (cART line2/3) were 47% less likely to develop DRCs earlier than those on first-line cART [AHR = 0.53; (95% CI: 0.33–0.87)]. Recipients of cART who had a CD4–1 count equal to or less than 200 cells/mm³ at cART initiation were 68% more likely to develop DRCs earlier than those who had a CD4–1 count greater than 200 cells/mm³ at initiation of cART [AHR = 1.68; (95% CI: 1.03–2.73)]. Participants who reported adherence to cART had a 3.1 times higher risk of developing DRCs earlier than those who did not adhere [AHR = 3.1; (95% CI: 0.92–10.0), p = 0.06]. Overweight patients at initiation of cART, had a 63% higher risk of developing DRCs earlier than recipients who had normal BMI at cART initiation [AHR = 1.63; (95% CI: 1.07–2.64)]. Males had a 51% lower hazard of the outcome than females [AHR = 0.49; (95% CI: 0.32–0.77)]. Age as a continuous variable did not enter the multivariate model as it did not achieve a p < 0.05.

The risk of new onset DRCs increased each year with cumulative exposure to each cART regimen under study. However, the most significant risk was documented among recipients of the first-line cART regimen while the least risk of new DRC cases per year of cumulative exposure to cART was recorded among recipients of the second-line cART regimen (Fig. 1).

The survival function of patients on second- and third-line regimens almost overlapped, with both functions disassociated with the first-line function. Data presented in Fig. 2 show survival functions of aggregated recipients of second- and third-line cART regimens in comparison with those of the first-line cART regimen.

The log rank (Mantel Cox) test confirmed that recipients of second-, third-line cART had significantly longer survival to DRC development than recipients of first-line cART (X² = 8.97, p = 0.003).

Findings from this study bring new insights to the fight against HIV and related comorbidities. The risk of new DRC onset per year of cumulative exposure to cART was higher among recipients of first-line cART than in recipients of cART second-line and cART third-line regimens. The longest survival/time duration to the development of DRC per year of cumulative exposure to cART was documented among recipients of second- and third-line cART regimens. In other words recipients of the second-, and third-line cART were less likely to develop DRCs earlier than recipients of the first-line cART. This lends support to the claim that the survival of cART recipients to DRC development is not solely a function of the duration that patients are subjected to these drugs but also a function of the specific regimen they are exposed to [19, 20] and the recipients underlying biomedical and demographic characteristics [21]. While the overlapping of the second-, third-line regimen DRC hazard/survival functions may suggest comparable effects between the two cART regimens, they also indicate that recipients of both the second and third-line regimen had a longer survival probability compared to their counterparts on first-line regimen. This implies that had some patients been directly initiated on second- or third-line cART regimen, the outcome might have been much better for all the recipients.

In short, important findings in this study were that recipients of the second-, third-line cART regimen were less likely to develop DRCs earlier than first-line cART regimen recipients. This is good news that may indicate that different cART regimens are differently associated with DRCs and different groups of recipients. Thus, depending on circumstances, the current second- and third-line regimens may be considered for use as first-line regimen while taking into account factors such as recipients’ pre-existing DRC risk factors, drug availability, cost-effectiveness, etc. For instance, patients with characteristics such as overweight, female gender or CD4–1 equal or less than 200 cells/mm³ were identified in this study as at more risk of developing DRCs earlier than those without and should have been put directly on second- or third-line regimen rather than starting with the first-line treatment. Recipients aged 35 years or older also need particular attention compared to those aged less than 35 years before being assigned to a specific cART regimen.

However, the challenge lies in the cost difference between the first-line, the second- and third-line regimens and possible adverse effects of the third-line regimen. While WHO states that treatment programs using Tenofovir or AZT are promising, they may not be the ultimate solution to alleviate the situation since both drugs are known to be expensive [22] and yet to some extent associated with DRCs [6, 7, 23–25]. The fact that despite their high cost, these drugs are still associated to a certain degree with DRCs [23–25] calls for a meticulous drug selection that takes into account biomedical and demographic characteristics of recipients so as to improve their survival probability to the event.

As with any other study, this study had some limitations. It was a retrospective review of data and may be missing some important DRC risk factors like physical activity, cigarette smoking, alcohol consumption and family history of DRCs. Prospective studies are needed to address this particular issue. Even so, the study has provided useful evidence to inform policy and decision making on alternative approaches for the management and care of HIV patients. The study has also indicated that patients with underlying DRC risk factors such as overweight, female gender, aged 35 years or older or those with CD4 ≤ 200 cells/mm³ should be selectively assigned to treatment regimens for possible improvement of their survival.

The study has shown that the risk of new DRC onset among cART recipients is not only a function of the type of cART regimen or the duration of exposure to these drugs but also a function of patients’ underlying biomedical and demographic DRC risk factors. The study has also presented a survival model highlighting DRCs’ significant prognostic factors to guide clinical care, policy and management of recipients of cARTs. Further studies in the same direction will likely improve the survival to the development of DRC of every cART recipient in this community.