Mid-Upper Arm Circumference Is a Strong Predictor of Mortality Among Ugandan Adults With HIV-Associated Cryptococcal Meningitis: A Prospective Cohort Study

Abstract

Background: Mortality among adults diagnosed with HIV-associated cryptococcal meningitis remains high (24%-40%). We hypothesized that nutritional state, as measured by mid-upper arm circumference (MUAC), is a potentially modifiable risk factor for mortality.

Methods: Ugandan adults hospitalized with HIV-associated cryptococcal meningitis had MUAC measurements performed at baseline. We compared MUAC measurements with baseline clinical and demographic variables and investigated associations with survival using Cox regression.

Results: Of 433 participants enrolled, 41% were female, the median CD4 T-cell count (interquartile range [IQR]) was 15 (6-41) cells/μL, and 37% were antiretroviral therapy naïve. The median MUAC (IQR) was 24 (22-26) cm, the median weight (IQR) was 53 (50-60) kg, and MUAC correlated with weight (Pearson r = 0.6; P < .001). Overall, 46% (200/433) died during the 18-week follow-up. Participants in the lowest MUAC quartile (≤22 cm) had the highest mortality: 39% (46/118) at 2 weeks and 62% (73/118) at 18 weeks. A baseline MUAC ≤22 cm was associated with an 82% increased risk of 18-week mortality as compared with participants with an MUAC >22 cm (unadjusted hazard ratio, 1.82; 95% CI, 1.36-2.42; P < .001). Following adjustment for antiretroviral therapy status, CD4 count, hemoglobin, amphotericin dose, and tuberculosis status, the adjusted hazard ratio was 1.84 (95% CI, 1.27-2.65; P < .001). As a continuous variable, 18-week mortality was reduced by 10% for every 1-cm increase in MUAC. CSF Th17 immune responses were positively associated with MUAC quartile.

Conclusions: MUAC measurement is a simple bedside tool that can identify adults with HIV-associated cryptococcal meningitis at high risk for mortality for whom an enhanced bundle of care, including nutritional supplementation, should be further investigated.

Keywords: HIV; cryptococcal meningitis; malnutrition; mid–upper arm circumference; mortality.

Figure 1.

Eighteen-week mortality by mid–upper arm circumference group in persons with HIV and cryptococcal meningitis. In comparing the mortality for MUAC ≤22 cm vs MUAC >22 cm, the hazard ratio was 1.82 (95% CI, 1.36–2.42; P < .001). When adjusted for CD4 T-cell count, antiretroviral therapy status, baseline hemoglobin, amphotericin B deoxycholate dose per MUAC cm, and time-updated tuberculosis status, the adjusted hazard ratio was 1.84 (95% CI, 1.27–2.65; P < .001). Abbreviation: MUAC, mid–upper arm circumference.

Figure 2.

Final tuberculosis diagnosis by mid–upper arm circumference group in persons with HIV and cryptococcal meningitis. Tuberculosis status throughout the follow-up study period. Prior TB disease was defined as having received TB therapy >14 days before cryptococcal meningitis diagnosis. TB at meningitis diagnosis (prevalent TB) was defined as those who received TB therapy within 14 days of cryptococcal meningitis diagnosis. Incident TB was defined as those who received TB therapy >14 days after cryptococcal meningitis diagnosis. The distribution of TB classification was not significantly different for those in the lowest MUAC quartile compared with those in quartile 2 or above (P = .05). Abbreviations: MUAC, mid–upper arm circumference; TB, tuberculosis.