Vestibular and oculomotor abnormalities among HIV-infected and HIV-uninfected men and women: A pilot study

1 Introduction

Human Immunodeficiency Virus (HIV) infection is associated with both peripheral and central vestibular dysfunction [12]. Post-mortem studies of HIV seropositive (HIV+) patients before the use of highly active antiretroviral therapy (HAART) suggested that direct viral damage occurs in the vestibular end-organs. Also, previous studies performed before the use of current antiretroviral protocols found neuronal loss in the central vestibular pathways [4, 19]. Other potential etiologies of vestibular impairments in the HIV+ population include normal disorders of aging and opportunistic infections such as otosyphilis or encephalitis [10, 21].

A systematic review identified 13 studies that examined the association between vestibular disorders and HIV [12]. The authors concluded that central vestibular impairments occur even at asymptomatic stages of HIV infection and are more prevalent at advanced stages [3, 23]. Peripheral vestibular impairments are less prevalent at the asymptomatic stage but have been found in up to 50% of advanced stage HIV+ patients [2, 23]. Heinze et al. concluded that patients with HIV infection should be routinely monitored for vestibular disorders [12].

The studies conducted by Heinze had some limitations [12–14]. For example, they had small sample sizes; they lacked CD4 cell count, HIV viral load, antiretroviral treatment characterization and appropriate statistical adjustment for these co-factors. These studies also varied in the availability of data from objective diagnostic tests of the vestibular system. Cohen et al. found no increased association between vestibular impairment indicated by vestibular screening tests and HIV in subjects who were well-controlled on HAART [5]. To gain a better understanding of the prevalence of vestibular impairment among community-dwelling virologically and immunologically controlled middle-aged HIV+ men and women, we conducted objective diagnostic tests of the vestibular system among participants from two longstanding observational cohorts of HIV+ and HIV– men and women in the US.

2 Methods

A total of 102 women (84 HIV+) from the Washington, DC site of the Women’s Interagency HIV Study (WIHS) and 138 men (80 HIV+) from the Baltimore-Washington, DC site of the Multicenter AIDS Cohort Study (MACS) were eligible for this study. The MACS and WIHS longitudinal study designs have been described elsewhere [1, 15]. Participants were excluded from the study if they had a history of lower extremity prostheses or used a cane, rollator or walker. Additional exclusion criteria included a history of head trauma or radiotherapy, cervical spine disease, weight over 300 pounds, or use of alcohol, narcotics, antihistamines or sedatives in the last 48 hours.

Participants were tested from August 2008 to September 2014. All participants underwent the following clinical testing: electronystagmography (ENG) including spontaneous nystagmus, gaze evoked nystagmus, saccades, Dix-Hallpike maneuvers (DHP) and bi-thermal caloric stimulation. The dependent measures included: saccade peak velocity, saccade accuracy, and saccade latency, all measured dichotomously as normal or abnormal; gaze testing: for all positions – center, up, down, right, and left, the presence of three or more beats of nystagmus were classified as abnormal. For Dix-Hallpike maneuvers the presence of three or more beats of nystagmus was considered positive, i.e., abnormal. Bi-thermal caloric tests with air were performed at 30 deg C and 44 deg C. Bi-thermal caloric weakness ≥20% or absent responses were considered abnormal and were combined together. All eye movements were recorded with infra-red video-oculography.

Subjects were also tested on cervical VEMP, using surface Ag/AgCl electrodes located at the center of the sternocleidomastoid body, on the sternum and on the forehead. Subjects laid supine and were asked to lift their heads against gravity during pure tone tests with insert electrodes. Initial stimuli were given at 90 dBHL decreasing in 5 dBHL steps.

Descriptive measures were sex; age at time of testing; race/ethnicity; CD4 + T cell count and log₁₀ plasma HIV RNA at the study visit closest to the date of the test; currently taking HAART at the time of testing; ever reporting AIDS; ever having used monotherapy, combination antiretroviral therapy (ART), or HAART; and undetectable HIV RNA was defined as < = 80 cps/mL.

For analysis, an abnormal result in any one of the three saccade tests was categorized as an abnormal saccade result; an abnormal result in any of the VEMP tests (threshold, R/L P13 or R/L N13) was categorized as an abnormal VEMP response; and any abnormal or absent weakness from the bi-thermic caloric stimulation testing was considered a response.

Logistic regression models were used to analyze data from the combined HIV+ and HIV– participants and HIV+ participants only. The models using the combined population were adjusted for HIV status, sex, and age at testing. Additional covariates for models including only the HIV+ participations were: ever monotherapy (yes or no); and undetectable viral load (yes or no).

The Johns Hopkins University School of Public Health, Georgetown University and Whitman Walker Health Institutional Review Boards approved this study. Subjects gave written informed consent prior to participation.

4 Discussion

HIV– men were less likely to have bi-thermal caloric weakness than HIV+ men or HIV+ or HIV– women. The finding that middle-aged HIV– men are less likely to have vestibular disorders than women is consistent with the epidemiological literature on vestibular disorders [17, 20]. We expected that ever using antiretroviral monotherapy and having an undetectable viral load with lessening neuro-inflammation would decrease the odds of vestibular abnormalities, yet our findings were mixed across gender.

Our findings are consistent with the earlier literature. For example, Castello et al. and Hausler et al. have shown that even asymptomatic HIV+ stage 1 patients may have oculomotor pathology [3, 11]. This literature must be viewed with caution as some studies were performed without controls or before the advent of modern HAART. Nevertheless, the findings are suggestive.

Heinze et al. described a small sample (N = 15) of HIV+ stage 1 subjects who were most comparable to our subjects although slightly younger [13]. They reported an overall prevalence of 19%, which is similar to our findings. Unfortunately, the investigators grouped the Stage 1 subjects together with HIV+ stages 2 to 4 subjects, used some discredited tests such as the Fukuda stepping tests and head impulse tests, and did not adjust the statistical analyses for the several confounding variables such as age, gender, use of antiretroviral therapy, and HIV viral load [6, 7].Therefore, these results are not comparable to the present study.

Impairments in saccades indicate abnormal function of specific brainstem pathways. Such impairments may be early harbingers of multiple sclerosis, Parkinson’s disease, or other central neurologic disorders. Therefore, saccadic impairments should be taken very seriously. An HIV+ patient who complains of blurred vision, but who passes vision testing in optometry or ophthalmology, should be referred for objective oculomotor tests and possibly to neurology or neuro-ophthalmology for work-up. Women are at a significantly greater risk than men and HIV+ patients may be at greater risk than HIV– patients.

The present study had several limitations: 1) relatively small sample size; 2) low prevalence of abnormalities; and 3) possible residual confounding of age and gender across the HIV+/HIV– groups. Despite these issues, the data suggest that women are at greater risk of vestibular and oculomotor abnormalities than men and HIV+ adults are at greater risk than HIV– adults. Thus, health care providers who care for HIV+ men and women should monitor the symptoms of vestibular and oculomotor impairment. Clinicians should keep in mind that an HIV+ adult who complains of general dizziness, vertigo specifically, balance problems, or a history of falls that might be associated with vestibular disorders should be worked up for a vestibular disorder, especially in the age range of these subjects. Referral to neurotology and neurology may be appropriate when patients are symptomatic.