Contrary to our hypothesis, this study provides evidence that more extensive, long-term cannabis use among HIV+ individuals is not associated with a higher likelihood of balance disturbances. While we did not find any research study in the literature to compare with our findings, one prior report found that the occurrence of balance disturbances was associated with a 13-fold higher odds of recurrent falls among HIV+ individuals (Nakatsukasa et al. 2014). It is not clear why the more frequent use of cannabis was not associated with a higher likelihood of balance disturbances in the HIV+ group, but a plausible explanation is that any deleterious effects of cannabis are counteracted by its effect of reducing inflammation (Juknat et al. 2012; Manuzak et al. 2018). Yet, the difference between HIV+ and HIV− individuals in cannabis-associated balance disturbances as both acute effect and chronic effect suggest that neuroinflammatory differences alone may not explain these results. In the brain, CB2 receptor expression is associated with inflammation and it is primarily localized to microglia (resident macrophages of the brain). This selective localization together with the modulatory effect of the CB2 receptor on microglia function (Kendall and Yudowski 2017) is particularly relevant since microglial cells have a significant role in neuroinflammation in HIV infection. In fact, HIV-infected monocytes not only infect brain resident cells (microglia) upon migration into the CNS but also produce proinflammatory cytokines, which in turn, further activate microglia. These activated microglia, along with perivascular macrophages, are the main contributors to neuroinflammation in HIV infection, resulting in neuronal dysfunction and death (Hong and Banks 2015).

In contrast, more prolonged chronic cannabis use was related to more severe imbalance among the HIV− individuals. One potential mechanism for this is adverse effects on the cerebellum and basal ganglia, both of which express high levels of CB1 receptors. Prior research (Bolbecker et al., 2018) found that chronic cannabis use in HIV− individuals was associated with increased postural sway in individuals who were not acutely intoxicated. Our results are similar to those of Bidwell et al. (2020) who found that balance function was impaired after immediate cannabis use and different from those of Pearson-Dennett et al. (2017) who found that the effect of long-term cannabis use was associated with long-lasting changes in open-chain elements of walking gait, but the magnitude of change was not clinically detectable.

Those studies assessed balance impairment after immediate cannabis use in small samples. In contrast, our study focused on prolonged use of cannabis and had more power due to the large population size. In Pearson-Dennett et al.’s (2017) study the average population age was between 24 and 25 years; however, in our study the average population was 45.4 years and any subtle walking gait changes observed in younger individuals may become more significant with aging due to age-related changes in postural control, leading to balance disturbances.

We also found that HIV+ individuals reported more balance disturbances and were more likely to have polyneuropathy signs and to use specific medications (opiates and sedatives), both of which are strongly associated with balance disturbances. The interaction between HIV serostatus and chronic cannabis use persisted after adjustment for these potential confounding factors. Our findings are consistent with previous research showing that higher rates of falls among HIV+ individuals are related to multiple comorbidities and the use of opiates, sedatives, and antihypertensive medications (Erlandson et al. 2017). Those using cannabis were more likely to use these balance-affecting medications, increasing the risk of drug-drug interactions or synergistic toxicities. A meta-analysis looking at the association between medication classes and fall risk showed the respective odd ratios: 1.24 [1.01–1.50] for antihypertensive agents, 1.47 [1.35–1.62] for sedatives and hypnotics, and 0.96 [0.78–1.18] for narcotics (Article 2009). Participants with history of diabetes, along with a higher prevalence of chronic distal polyneuropathy were more likely to have balance disturbances as well. Diabetes and cDSPN were controlled for in our statistical model and the effect of long-term cannabis use persisted. Therefore, our findings might indicate that HIV+ individuals might be less susceptible to cannabis’ effect on balance. The proportion of participants who ever used cannabis did not differ among HIV+ and HIV− individuals. The proportion of virologic suppression in HIV+ individuals on treatment was low in our population; however, it did not affect the primary outcome (balance disturbance). A positive urine toxicology for THC was not significantly more common in participants with balance disturbances. Urine toxicology is not a good proxy for recency of use or intoxication, since the test may remain positive for up to 30 days.

There are limitations to this study. First, data were based on a retrospectively self-reported data. Second, the cross-sectional design cannot show a causal relationship between cannabis use and balance disturbances. Nevertheless, a recent comparison of fall recall and prospective fall collection among 600 older adults in Germany found nearly identical incidence rate of falls, supporting the validity of retrospective fall history (Rapp et al. 2014). Also, we did not collect any information about cannabis route of administration or whether cannabis use was recreational or prescribed, which factors may moderate cannabis-balance association.