In this patient population of adults with perinatally acquired HIV we identified a high prevalence (30.6%) of HTN. This striking hypertension prevalence is concerning given the available normative data regarding hypertension in the young adult age group (18–29 years). The National Health and Nutrition Examination survey demonstrated a hypertension prevalence among 24- to 32-year-olds of only 4%, whereas the Add Health Cohort reported a higher prevalence (19%) in an identical age bracket, with highest prevalence among individuals at the older end of this age spectrum. In the Johns Hopkins Precursors Study, a prospective cohort of 1132 white males, the cumulative incidence of hypertension at 25 years of age was only 0.3%.
Hypertension is well described in older HIV-infected patients, with several cohort studies with mean ages of 40–50 years estimating a prevalence range of 8–43%.[6–10,17,18,29–32] By contrast, patterns of hypertension in younger adults with HIV are largely unknown. In a single study among patients in the age range 2–25 years, the prevalence of hypertension among those aged 18–25 years was 24%, although the diagnostic criteria were significantly limited as they were based only on a single blood pressure reading.
Contrary to our hypothesis, we did not identify an association between a majority of traditional hypertension risk factors, including tobacco use and family history of hypertension, and risk of systemic hypertension. In addition, we did not identify a statistically significant association between HIV-associated risk factors and risk of hypertension. Although CD4 nadir in general was not significantly associated with the odds of being a case, a CD4 nadir of ≤ 200 cells/μL did trend towards significance with regard to the point estimate and CI.
There are several possible explanations for these findings. Given the relatively young age of this cohort, traditional hypertension risk factors, which have previously been identified largely among older adults, may not yet play as important a role given shorter duration of exposure with these risk factors. For example, given the known effects of ageing on the anatomy and function of blood vessels and cardiovascular performance, it is reasonable to assume that a population of young adults may not show enough meaningful loss of arterial elasticity such as that seen in older adults and which contributes to the higher rates of hypertension in older-aged cohorts. This same concept may also apply to the long-term effects of tobacco and alcohol use. The cumulative exposure of tobacco and alcohol (both established risk factors for hypertension) would also be expected to be lower among this younger cohort.[34,35]
In contrast to these findings, this study noted an increased odds of hypertension diagnosis among those with CKD. While the number of patients with a concurrent CKD diagnosis was significantly higher in the case group than in the control group, the relationship between these two diseases is difficult to clarify in the context of this study. Notably, the temporal relationship between CKD and HIV was difficult to elucidate, as we were unable to identify the specific dates of diagnosis for both diseases for the majority of subjects in either the case or control group. This relationship is a dynamic one, as uncontrolled hypertension can lead to kidney damage while existing kidney disease can result in the development of hypertension. In a previous study that included a majority of the patients also examined in this study, a mediation analysis showed that approximately one-third of the odds of a patient having hypertension was mediated by CKD; however, excluding those patients with CKD, the prevalence odds ratio remained significant for the rate of hypertension seen in PHIV adults compared with that of uninfected patients. Nevertheless, the directional relationship between these two conditions must be further clarified in this and other cohorts of PHIV adults.
Although we aimed to explore the relationship between BMI and the development of hypertension, we did not include it in our final analysis. In our data abstraction, we captured BMI readings from all clinic visits available in the electronic medical record; however, as this was not a prospective cohort, reliably establishing the temporal relationship between these readings and the date of diagnosis of hypertension was compromised due to inconsistent height and weight measurements over time.
Interestingly, exposure to lopinavir/ritonavir was associated with significantly decreased odds of having hypertension. Previous studies have shown conflicting data regarding lopinavir/ritonavir exposure and hypertension, with two small studies demonstrating higher rates of hypertension in those with exposure to the drug, while another study showed lower rates of hypertension.[36–38] Lopinavir/ritonavir is associated with increased risk for hypercholesterolaemia, hypertriglyceridaemia (occasionally resulting in pancreatitis), blood glucose disorders including diabetes mellitus, and QT interval prolongation; however, other than the studies alluded to here, no significant association has been established with blood pressure.[39,40] There are a few plausible explanations for this significant difference. The first is the potential for channelling bias. In this scenario, providers may have preferentially chosen to avoid lopinavir/ritonavir due to a pre-existing comorbidity which itself may have increased their risk for hypertension. Second, it is plausible that there were HIV-associated risk factors, unidentified in this study, that increase the risk of hypertension in this population; if that were to be true, lopinavir/ritonavir may have more effectively treated HIV in this population and thus led to a decrease in HIV-associated risk for hypertension.
This study has several limitations. It was conducted at a single university, with a majority of the study population being of African American ethnicity from an urban environment; this may limit the generalizability of our findings to other populations of PHIV adults. Due to the retrospective nature of the study design, some covariates had to be measured as binary (e.g. exposure to tobacco and alcohol). It is possible that a relationship between these covariates and the development of hypertension could be found if they were classified by burden of exposure. Many patients had one or more parents who were deceased, increasing the risk of misclassification of family history of hypertension. Additionally, given the retrospective nature of this study, potential differences in the reliability of documentation could be expected. A final limitation is the definition chosen for systemic hypertension. We chose the definition of hypertension to align with similarly designed studies in the HIV literature; however, by utilizing this definition across a large period of time it is possible that patients who met criteria for hypertension had only transient elevated blood pressures which subsequently normalized. The most recent American Heart Association recommendation on blood pressure thresholds for hypertension uses a blood pressure reading > 130/90 mmHg, while we chose to use the previous definition of > 140/90 mmHg. It is possible that applying the new definition could provide more power to detect associations, as the number of patients meeting a case definition would probably increase.
While the incidence of perinatal HIV transmission has declined dramatically in the United States, the number of children infected perinatally in the developing world continues to be high. As increasing global numbers of PHIV age into adulthood, the risk for chronic non-AIDS disease burden may continue to grow. This study provides further context to emerging literature specifically regarding the risk of systemic hypertension among adults with perinatally acquired HIV adults. Further research should prospectively examine the interplay of chronic HIV disease and antiretroviral burden on systemic hypertension among PHIV.
IDSA Medical Scholars Program
The first author would like to thank the Infectious Diseases Society of America (IDSA) Medical Scholars Program for funding support for this project. There are no conflicts of interest to report.
HIV Medicine. 2022;23(5):457-464. © 2022 Blackwell Publishing