Nutrition indicators as potential predictors of AIDS-defining illnesses among ARV-naïve HIV-positive adults in Kapiri Mposhi, Zambia 2008-2009

Yi No Chen; Kristin M. Wall; Kadija Fofana; Carlos Navarro-Colorado

doi:10.1371/journal.pone.0219111

Abstract

Early changes in nutritional status may be predictive of subsequent HIV disease progression in people living with HIV (PLHIV). In addition to conventional anthropometric assessment using body mass index (BMI) and mid-upper arm circumferences (MUAC), measures of strength and fatigability may detect earlier changes in nutrition status which predict HIV disease progression. This study aims to examine the association between various nutritional metrics relevant in resource-scarce setting and HIV disease progression. The HIV disease progression outcome was defined as any occurrence of an incident AIDS-defining illnesses (ADI) among antiretroviral treatment (ART)-naïve PLHIV. From 2008–2009, HIV+ Zambian adult men and non-pregnant women were followed for 9 months at a Doctors without Borders (Medecins Sans Frontiers, MSF) HIV clinic in Kapiri Mposhi, Zambia. Since the study was conducted in the time period when former WHO recommendations on ART (i.e., ≤200 CD4 cell count as opposed to treating all individuals regardless of CD4 cell count or disease stage) were followed, caution should be applied when considering the implications from this study’s results to improve HIV case management under current clinical guidelines, or when comparing findings from this study with studies conducted in recent years. Bivariable and multivariable logistic regression was used to assess the associations between baseline nutritional measurements and the outcome of incident ADI. Self-reported loss of appetite study (AOR 1.90, 95% CI 1.04, 3.45, P = 0.036) and moderate wasting based on MUAC classification (AOR 2.40, 95% CI 1.13, 5.10, P = 0.022) were independently associated with increased odds of developing incident ADI within 9 months, while continuous increments (in psi) of median handgrip strength (AOR 0.74, 95%CI 0.60, 0.91, P = 0.004) was independently associated with decreased odds of incident ADI only among women. The association between low BMI and the short-term outcome of ADI was attenuated after controlling for these nutritional indicators. These findings warrant further research to validate the consistency of these observed associations among larger ART-naïve HIV-infected populations, as well as to develop nutritional assessment tools for identifying disease progression risk among ART-naïve PLHIV.

Citation: Chen YN, Wall KM, Fofana K, Navarro-Colorado C (2019) Nutrition indicators as potential predictors of AIDS-defining illnesses among ARV-naïve HIV-positive adults in Kapiri Mposhi, Zambia 2008-2009. PLoS ONE 14(7): e0219111. https://doi.org/10.1371/journal.pone.0219111

Editor: Sphiwe Madiba, Sefako Makgatho Health Sciences University, SOUTH AFRICA

Received: August 30, 2017; Accepted: June 17, 2019; Published: July 2, 2019

This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

Data Availability: The minimal anonymized dataset necessary to replicate our study has been uploaded: Wall, Kristin, 2019, "Replication Data for: Nutrition indicators as potential predictors of AIDS-defining illnesses among ARV-naïve HIV-positive adults in Kapiri Mposhi, Zambia 2008-2009", https://doi.org/10.7910/DVN/LQWVRQ, Harvard Dataverse.

Funding: This work was supported by Doctors Without Borders/Médecins Sans Frontières-OCBA.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The relationship between malnutrition and HIV infection has been well documented in the existing literature [1, 2]. Because untreated HIV infection may impose a negative nutritional balance by increasing the static energy requirement and compromising nutrient intake and absorption, antiretroviral treatment (ART)-naïve PLHIV are an especially vulnerable group for acute malnutrition, which may leave patients with a greater risk of mortality or other serious health conditions (e.g., organ dysfunction) over a short period of time [3]. Although there has been tremendous progress in scaling up ART coverage worldwide in the last five years, a significant portion of PLHIV continues to remain ART naïve (40% in the WHO African region and 25% or about 275,000 PLHIV in Zambia as of 2017)[4, 5]. As food insecurity and subsequent malnutrition, prominent barriers to ART initiation and adherence [6], continue to plague many resource-scarce communities, understanding the prognostic relationship between nutritional indicators and short-term HIV disease progression among ART-naïve populations in those communities is critical.

Body Mass Index (BMI) and Mid-Upper Arm Circumference (MUAC) are often studied as indicators of malnutrition in adult patients, and have been associated previously with adverse clinical outcomes among PLHIV [7–15]. Several studies have shown a relationship between malnutrition related muscle fatigability and strength and adverse disease outcomes in HIV-free patients [16–19]. Furthermore, changes in muscular performance become observable before changes in weight and muscle mass when responding to both nutritional deprivation and nutritional repletion [20, 21]. This may suggest the potential utility of muscular strength and fatigability in predicting health outcomes among PLHIV who are at risk of HIV-associated malnutrition.

In light of the existing literature on the conventional nutritional predictors of poor HIV prognosis [7–14] and the growing evidence of other anthropometric measurements in describing the effect of malnutrition on health in various populations [16–19], the purpose of this 9-month prospective study was to assess the performance of a wide range of baseline nutritional indicators in predicting short-term HIV disease progression among ART-naïve adults in a severely resource-scarce setting. Since the study was conducted in 2008, the ART was recommended for PLHIV whose CD4 cell count drop below 200 cells/mm³ as opposed to all PLHIV regardless of CD4 and disease stage as per current WHO guidelines. Despite the potentially decreased relevance to current clinical management of HIV, we felt that the analysis may inform understanding of the prognostic capacities of common nutritional indicators among ARV-naïve PLHIV in certain resource-scarce areas where patients may not have immediate access to ART or patients delayed ART care. This may contribute to the development of field-friendly nutrition screening tools for the early identification of malnourished PLHIV at risk of hastened disease progression.

Materials and methods

Study location, participants, and design

Kapiri Mposhi is a primarily rural district in Central Province, Zambia. The town of Kapiri Mposhi lies on a major transportation route for Zambians and travellers from neighbouring countries. In 2006, the district had an estimated HIV prevalence of 17.4% [22].

From December 12, 2007 to January 30, 2009, HIV-positive adults aged ≥ 18 years who attended HIV clinics run by MSF-Operational Centre Barcelona and Athens (MSF-OCBA) in the town of Kapiri Mposhi, Zambia were enrolled in a 9-month prospective cohort study entitled “Early detection of nutritional status and prediction of survival in HIV patients: analysis of anthropometry and functional indicators” (IRB00001131). This study and the secondary analysis were approved by the Research Ethic Committee at University of Zambia and by the Office for Human Research Protections-registered Institutional Review Boards at Emory University, respectively.

Nutritional information was collected from enrolled patients using a questionnaire administered at study admission and at quarterly nutritional follow-up visits (3 and 6 month) (S1 Fig). MSF-OCBA HIV clinics used Follow-Up of Clinical HIV Infection and AIDS (FUCHIA v. 1.5.1, Epicentre, Paris) software to collect and store demographic and clinical data from patients at their clinical visits. Patients from the clinics were actively referred to enroll in ART treatment programs if the 2006 World Health Organization (WHO) treatment criteria were met.

This analysis includes patients who were ART-naïve prior to and upon admission into the study who were strong enough to stand and be measured. Women who were pregnant and/or lactating were excluded, because they have meaningfully different metabolic characteristics that may require different nutritional assessment schemes than majority of our target adult population. The follow-up period was defined as the time between the study admission date and the date of ART initiation or the end of the 9-month study end date. Upon recruitment, patients who participated in the study would provide written consents. Since very few participants received any form of nutritional support at (n = 5) or after baseline (n = 6), the analysis aimed to target resource-scarce populations with no substantial external nutritional support programs in place.

Outcome of interest

Incident ADI was the outcome of interest in this study. According to the WHO disease staging system for HIV infection and disease in adults and adolescents, the following clinical conditions are considered ADIs (i.e., severe HIV-associated symptoms, clinical stage 4 of HIV infection, and requiring ART initiation regardless of the patient’s immune status): HIV wasting syndrome, Pneumocystis pneumonia, Recurrent severe bacterial pneumonia, Chronic herpes simplex infection (orolabial, genital or anorectal of more than one month’s duration or visceral at any site), Oesophageal candidiasis (or candidiasis of trachea, bronchi or lungs), Extrapulmonary and pulmonary tuberculosis (TB), Kaposi’s sarcoma, Cytomegalovirus infection (retinitis or infection of other organs), Central nervous system toxoplasmosis, HIV encephalopathy, Extrapulmonary cryptococcosis including meningitis, Disseminated non-tuberculous mycobacterial infection, Progressive multifocal leukoencephalopathy, Chronic cryptosporidiosis (with diarrhoea), Chronic isosporiasis, Disseminated mycosis (coccidiomycosis or histoplasmosis), Recurrent non-typhoidal Salmonella bacteraemia, Lymphoma (cerebral or B-cell non-Hodgkin) or other solid HIV-associated tumours, Invasive cervical carcinoma, Atypical disseminated leishmaniasis, Symptomatic HIV-associated nephropathy or symptomatic HIV-associated cardiomyopathy [23]. Regardless of participants’ status for these symptoms prior to the study, those newly diagnosed with at least one of these conditions during the follow-up period were considered to have had the outcome of interest.

Non-nutritional covariates

Baseline age, sex, CD4 cell count, diagnoses of non-severe HIV-associated symptoms, and diagnoses of ADIs were recorded at the time of entry into the study. Baseline age was trichotomized based on sex-specific tertile cut-points: 1) Male: age < 31 years/ Female: age years < 28 years; 2) Male: 37 years > age ≥ 31 years/ Females: 34 years > age ≥ 28 years; 3) Male: age ≥ 37 years/ Female: age ≥ 34 years. A dichotomous variable of baseline immunosuppression was made to indicate CD4 cell count lower than 200 cells/mm³.

Non-severe HIV-associated symptoms were coded as a composite binary variable describing a diagnosis of at least one of the following conditions: weight loss, minor mucocutaneous manifestations, herpes zoster, recurrent upper respiratory tract infections, bedridden during the last month, oral hairy leukoplakia, unexplained chronic diarrhea > 1 month, unexplained prolonged fever > 1 month, oral candidiasis, vulvovaginal candidiasis > 1 month [23]. The diagnosis of ADI at baseline was also a binary variable created to describe the diagnosis of at least one of the clinical stage 4 HIV-associated symptoms.

Nutritional covariates

Loss of appetite and loss of weight were self-reported. Food security at the household level was assessed by asking whether participants experienced any of the following in the last three months prior to the visit: 1) being unable to eat the preferred types of food due to lack of resources; 2) having to eat a smaller or fewer meals than usual; 3) having no food stored in the participant’s household, and 4) sleeping on an empty stomach.

The techniques for performing MUAC, height, and weight measurements followed the standard Médecins Sans Frontières/Doctors Without Borders protocols. Briefly, MUAC was measured on the left arm, when the arm is relaxed at the side of the body, at the mid-point between the elbow and the shoulder. A non-stretchable measuring tape was placed around the arm and the measurement was read from the window of the tape without pinching the arm or leaving the tape loose. The MUAC was recorded to the nearest millimeter. Height was measured using an estadiometer, to the closest millimeter. Weight was measured using standing scales, to the closest 100 grams. Scales were calibrated daily using a known weight. BMI, MUAC, and height were measured by staff.

BMI was categorized into normal (BMI≥ 18.5), mild thinness (18.5 >BMI≥ 17.0), moderate thinness (17 >BMI≥ 16.0), and severe thinness (BMI < 16.0) using a grading system derived for adult chronic energy deficiency. Moderate and severe thinness were collapsed during regression analysis to obtain stable association estimates. MUAC was trichotomized according to sex-specific categories used in conditions of food scarcity: normal (Male: ≥ 230 mm; Female: ≥ 220 mm), moderate wasting (Male: 200–229 mm; Female: 190–219 mm), and severe wasting (Male: <200 mm; Female: <190 mm)[24]. Continuous BMI was excluded from the bivariable and multivariable modeling process, due to severely skewed distribution.

Muscular strength and fatigability were measured using two separate tests–the handgrip strength test and the “sphygmomanometer test”. The handgrip strength test required participants to squeeze a hand dynamometer 10 times consecutively with maximum force. Grip force (measured in psi) was recorded at each squeeze, and the following summary statistics were obtained from the ten handgrip force readings (in psi): median, slope, and percentage strength lost between the maximum reading and final reading. The two latter indicators were considered as measures of muscular fatigability and were calculated as follows:

We devised the”sphygmomanometer test” to capture a composite measure of muscular strength and fatigability, by recording the cumulative pressure readings (mmHg) resulting from the repeated squeezes on the rubber bulb of mercury sphygmomanometer with maximum force for 60 seconds or until failure due to muscle fatigue. Each sphygmomanometer was emptied prior to conducting a new test. Cumulative pressure reading (mmHg) and the total number of hand squeezes obtained during the test were used to compute the average squeeze strength (mmHg/squeeze). A dichotomous variable was created to indicate if a participant failed at completing the 60-second “sphygmomanometer test”.

For all continuous measurements created from the two muscular strength and fatigability tests, skewness was obtained to assess distribution normality. Continuous measurements that had significantly skewed distribution (i.e., |skewness| > 0.50) were dichotomized using sex-specific median cut-point such that the characteristics could be modeled in logistic regression: percent strength loss (handgrip strength/fatigue test), slope of handgrip measures (handgrip strength/fatigue test), cumulative grip strength (“sphygmomanometer test”), and average grip strength (“sphygmomanometer test”).

Data management and statistical analysis

Data cleaning and analyses were conducted using MS Excel, SAS v9.3 and SAS v9.4 (Cary, NC). The distributions of nutritional and non-nutritional characteristics were described by the outcome of incident ADI diagnosis using frequencies and percentages for categorical covariates or means and standard deviations for continuous covariates. Chi-square tests (for categorical variables), two-sample equal variance t-tests for normally distributed data (for continuous variables), and Mann-Whitney U for non-parametric data (for continuous variables) were used to compare the distribution of characteristics by the outcome incident ADI diagnosis. The characteristics that were originally continuous measurement but categorized based on cutoff percentile were treated only as continuous variables for descriptive analysis. The distributions of incident ADI by illness type and the sum of ADI incidence and related visits per person were described.

Logistic regression models were used to assess bivariate and multivariate associations between covariates and incident AIDS-defining illnesses. The nutritional covariates that had significant crude odds ratios (P<0.05) in the bivariate analyses were candidate exposure variables for inclusion in a multivariable logistic regression model, while all significant non-nutritional covariates were included as potential confounders. Some nutritional characteristics that were represented by both categorical and continuous variables, and if both were significant in the bivariable model, only the categorical variable would be included in a given multivariable model in order to maximize the number of covariates we can keep in the model.

To build the final multivariable logistic regression model, covariates that were candidates for the final multivariable logistic models were assessed for collinearity. When at least two covariates were associated with a Conditional Index greater than 30 and variance decomposition proportion greater than or equal to 0.5, one of the covariates was dropped. The decision about which covariate to drop was based on the statistical significance and clinical meaningfulness of the variable, as well as the overall fitness of the model. Dichotomous baseline immunosuppression was excluded from the final primary multivariable model, as its inclusion would have reduced the sample size (by 24%) and event count (by 29%) considerably. All exposure variables were also assessed for effect-modification by sex using likelihood ratio test. Adjusted odds ratio and 95% CI were obtained for each exposure in a final model. For the purpose of sensitivity analysis, two separate multivariable logistic models were run: 1) a multivariable model (adjusting for the same set of variables as in the final model) excluding participants who were diagnosed with ADI at baseline; 2) a multivariable model adjusting for the variables included in the final model, and baseline CD4, a variable that was excluded from the primary model due to missingness (24% of baseline CD4 results were missing in the analysis dataset). Due to insufficient sample size (n = 43), no additional multivariable model was performed for participants with baseline ADI.

Results

Baseline non-nutritional and nutritional covariates

The cohort was comprised of 238 men (42%) and 334 women (58%). During the nine-month follow up of the study, 28 men (12%) and 47 women (14%) were diagnosed with an incident ADI. Among the 75 participants who developed incident ADI, there were overall 79 ADI events (75 new illness and 4 recurrent illness). Throughout the study period, all cases had only 1 follow-up clinical visit where at least an incident ADI event was recorded. 71 cases had only 1 incident ADI event in that visit, while other 4 cases had 2 concurrent incident ADI events. On average, cases developed 1.05 incident ADI event over the 9-month follow-up. The two most prominent ADI in this cohort included pulmonary (58%) and extra-pulmonary (14%) TB, which together accounted for 72% of all incident ADI events.

Table 1 describes unstratified and stratified distribution of baseline characteristics by the outcome of interest. More than one-third of participants reported some degrees of food insecurity, ranging from inability to eat preferred good (42%) to sleeping with empty stomach (32%). About 13% and 7% of participants experienced moderate or severe thinness (i.e., BMI<17) and severe wasting according to MUAC at baseline, respectively. Participants experiencing an incident ADI during follow-up were more likely to have baseline immunosuppression (69% v. 45%, P = 0.002), self-report loss of appetite (72% v. 49%, P<0.001), have lower BMI (18.6 v. 20.0 kg/m², P<0.001), and have lower MUAC (220.1 v. 243.8 mm, P<0.001) than those who did not experience an incident ADI. Additionally, those experiencing an incident ADI during follow-up were more likely to have BMI lower than 18.5 kg/m² (49% v. 28%, P<0.001) and have MUAC lower than 230 mm for men or 220 mm for women (55% v. 26%, P<0.001). Compared to the sub-population that did not develop incident ADI, greater proportion of participants experienced moderate or severe thinness based on baseline BMI in those that developed incident ADI (26% v. 11%). Similarly, greater proportion of participants experienced severe wasting based on the baseline measurement of MUAC in the subpopulation that developed incident ADI, compared to those that did not (19% v. 5%).

Download:

Table 1. Baseline characteristics of HIV-positive adults in Kapiri Mposhi, Zambia, by the incidence of AIDS-defining illnesses (ADI), 2008–2009.

https://doi.org/10.1371/journal.pone.0219111.t001

Participants experiencing an incident ADI during follow-up had lower median (8.4 v. 10.1 psi, P<0.001) grip strength, had a higher percent handgrip strength loss (25.0% v. 22.2%, P = 0.001), produced smaller number of grips (60.2 v. 65.4 grips, P = 0.019), and had lower cumulative strength measures (140.0 v. 150.0 mmHg, P<0.001) on “sphygmomanometer test” kit.

Bivariate and multivariate analysis

Significant bivariate associations were found between the diagnosis of incident ADI and ten covariates (Table 2). The following variables were included in the final multivariable logistic regression model: loss of appetite, categorical BMI, categorical MUAC, median of measures in handgrip strength test, categorical percent strength loss, total number of grip and categorical cumulative grip strength in the sphygmomanometer grip test. We observed negative bivariate trends between BMI and the odds of incident ADI (OR _mild 2.11, 95% CI 1.14, 3.88, P = 0.017; OR _{moderate/severe} 3.21, 95% CI 1.73, 5.97, P<0.001), and between MUAC and odds of incident ADI (OR _moderate 2.78, 95% CI 1.60, 4.82, P<0.001; OR _severe 5.77, 95% CI 2.75, 12.07, P<0.001).

Download:

Table 2. Unadjusted logistic model of incident AIDS-defining illnesses among HIV-positive adults in Kapiri Mposhi, Zambia, 2008–2009.

https://doi.org/10.1371/journal.pone.0219111.t002

The absence of meaningful collinearity among covariates in the multivariable logistic model was confirmed (Conditional Index = 29.5). The model suggested three independent predictors of incident ADI among HIV-positive Zambian adults (Table 3). Controlling for all other covariates, the odds of incident ADI among participants that self-reported loss of appetite at baseline was 1.90 times greater than those who did not (AOR 1.90, 95% CI 1.04, 3.45, P = 0.036); the odds of incident ADI among participants who experienced moderate wasting based on baseline MUAC measurements was 2.40 times (AOR _moderate 2.40, 95% CI 1.13, 5.10, P = 0.022) greater than those classified as normal. In addition, the association between the median of handgrip strength measures and the incident ADI was different by sex (likelihood ratio test for the interaction term: P = 0.009): among female participants, each one psi decreased in median handgrip strength on average was significantly associated with increasing the odds of developing incident ADI by 1.35 times (AOR 0.74, 95%CI 0.60, 0.91, P = 0.004), controlling for all other covariates. However, the median value of handgrip strength measures was not independently associated with the outcome of incident ADI among male participants (AOR 1.03, 95% CI 0.88, 1.22, P = 0.687).

Download:

Table 3. Multivariable logistic model of incident AIDS-defining illnesses among HIV-positive adults in Kapiri Mposhi, Zambia, 2008–2009 (N = 545, Event = 66).

https://doi.org/10.1371/journal.pone.0219111.t003

Sensitivity Analysis

After excluding the participants with baseline ADI, the associations between two of the predictors from the final multivariable model and incident ADI became mildly attenuated: loss of appetite (AOR 1.71, 95% CI 0.93, 3.17, P = 0.085) and MUAC (AOR_moderate 2.08, 95% CI 0.95, 4.55, P = 0.069; AOR_severe 3.47, 95% CI 0.72, 16.88, P = 0.1226). On the other hands, the association between median handgrip strength and the study outcome remained relative the same in the effect size and continued to be statistically significant among women (AOR 0.73, 95%CI 0.60, 0.89, P = 0.003).

After adjusting for baseline CD4, the associations between all three former predictors and short-term ADI became mildly attenuated in effect size: median handgrip strength among women (AOR = 0.77, 95% CI 0.61, 0.95, P = 0.020), loss of appetite (AOR = 1.73, 95% CI 0.86, 3.49, P = 0.124), and MUAC (AOR_moderate 2.17, 95% CI 0.89, 5.27, P = 0.087; AOR_severe 2.59, 95% CI 0.49, 13.59, P = 0.262). To ensure the absence of multicollinearity in the CD4-adjusted model, we excluded cumulative handgrip strength (as per the “sphygmomanometer test”) from the controlling covariates.

Discussion

Our analysis sought to understand if common nutritional indicators measured in a resource-scarce clinical setting had meaningful prognostic value for HIV disease progression among ART-naïve PLHIV. This may inform and potentially improve case management of HIV patients in resource-poor or emergency contexts, where co-morbidity with malnutrition may be prevalent. Our primary multivariable model suggested the following three baseline nutritional indicators were independently associated with the short-term elevated odds of incident ADI: self-reported loss of appetite and low MUAC measurements among PLHIV of both sexes and median handgrip strength measures among women. The findings from sensitive analysis were overall confirmatory in the estimated effect sizes, however, with less statistical power to validate the true association between the three nutritional indicators and the study outcome. Since about ¾ of the incident ADI events were related to TB, our study may be more relevant to describing relationship between malnutrition and TB-related ADI. The TB-heavy distribution found in this study is consistent with the knowledge that TB represented more than half of the AIDS-defining conditions in many developing countries, as well as the epidemiologic context of high HIV/TB comorbidity in Sub-Saharan Africa [25, 26].

The observed association with self-reported appetite loss in this study (AOR 1.90, 95% CI 1.04, 3.45, P = 0.036) is consistent with the existing body of evidence in which reduced appetite has been linked with poor nutrition and worsened HIV disease progression and treatment outcomes [27–29]. Such an association independent of other well-recognized nutritional prognostic markers (e.g., BMI, MUAC) could imply the potential utility of appetite loss assessment in screening ART-naïve PLHIV at risk of hastened disease progression in areas where access to adequate clinical and nutritional assessment is limited. On the other hand, the researchers from the Nutritional Support for Africans Starting Antiretroviral Therapy (NUSTART) trial [30] only observed crude association between mortality and baseline appetite among Tanzanian and Zambian PLHIV during the follow-up period between ART referral and early treatment phase [31]. This could potentially imply that recent appetite loss, which reflects change in appetite to some degree, may be better predictor of acute disease progression than a one-time measurement of appetite. Nevertheless, caution should be applied when interpreting the observed association between appetite loss and incident ADI, considering the TB-heavy distribution in our study population. It is possible that the association could be confounded by the already-existing yet undiagnosed TB disease, which commonly induces loss of appetite [32]. NUSTART and another randomized control trial also suggest a sizable burden of undiagnosed TB among malnourished Zambian PLHIV [31, 33]. From this perspective, increasing TB interventions (e.g., diagnosis and treatment) could potentially benefit ARV-naïve PLHIV who experienced malnutrition in rural Zambia.

Although various measures of weight loss and food insecurity have been shown to predict worsened HIV disease progression in some studies [34–36], we did not observe crude nor adjusted associations between short-term incident ADI and these self-reported variables in our study. It is possible that the lack of association with food insecurity could be related to the use of non-standard food security questionnaire. It could also imply the presence of recall or response biases [37]. For example, patients may over-report the events of food insecurity if they perceive food aids dependent on their responses [38]. Self-reported weight loss, on the other hands, relies on patients’ memory of their usual weights, which could be inaccurate for people who are not accustomed to tracking their own weight and/or do not have regular access to proper equipment. Self-reported loss of appetite is subjective and may also be subjected to recall bias [39]. However, the observed difference in the predictive associations among these three self-reported variables potentially suggest that self-reported loss of appetite may have been a more robust or objective measure than weight loss and food security.

A cross-sectional study done by Kelly P et al. in Zambia has suggested that low MUAC and BMI measurements were associated with the presence of opportunistic infection, a study outcome that included many clinical manifestation of ADI among ART-naïve PLHIV [9]. Our findings further supplement the evidence on the linkage between MUAC and clinical markers of HIV/AIDS among ART-naïve PLHIV by demonstrating that moderate wasting (based on baseline MUAC measurement) independently predicted short-term incident ADI after controlling for various nutritional metrics. In addition, the finding of a survival analysis study done among ART-naïve HIV patients in National Simão Mendes Hospital (NSMH) of Guinea-Bissau suggests that low MUAC (i.e., ≤ 250 mm) independently predicts mortality within 2.5 years among ART-naïve PLHIV, after adjusting for sex, HIV type, CD4 and BMI [40]. This could potentially imply that MUAC is not only an independent predictor for short term adverse clinical outcome, indicated by incident ADI within 9 months, but also for relatively long term and more severe disease outcome among ART-naïve PLHIV.

On the other hand, the observed association between BMI and the short-term ADI was significantly attenuated in the multivariable model. This is rather interesting, because MUAC, which has been shown strongly correlated with BMI among various populations [41, 42] including our study participants (chi-square test between categorical MUAC and BMI, P<0.001; correlation coefficient between continuous MUAC and BMI = 0.851, P<0.001), remains independently predictive of the short-term ADI. This could imply that MUAC is a better prognostic marker for hastened disease progression among ART-naïve PLHIV. This nuance between MUAC and BMI may be consistent with the findings of a nutritional supplementation trial among 1561 emergency care patients, in which BMI failed to predict mortality adjusting for MUAC and weight loss while MUAC remained predictive [43]. Further studies with larger sample size may still be required to validate the attenuation of the association between BMI and hastened disease progression after controlling for MUAC, as past studies have shown BMI to be independently predict worsened HIV disease progression over longer follow-up period [8, 44, 45]. Nevertheless, the relatively easy measurement of MUAC in resource-scarce clinical setting and for physically disable patients renders it great potential for field use as possible prognosis tool [46, 47].

In addition, our multivariable model shows a statistically marginal association between severe wasting and incident ADI. The greater adjusted OR could potentially imply that the association with short-term incident ADI may be even stronger for those with extreme degree of wasting as indicated by MUAC. However, given the limited sample size of participants with severely low MUAC, our study may not statistically validate such an implication; further research with ample samples on extremely low MUAC population may be required.

The predictive association observed among female PLHIV in this study is consistent with the growing body of epidemiologic narratives which demonstrates handgrip strength, a common index for functional status, as a prognostic marker for short-term and long-term adverse health outcomes (e.g., hospitalization, premature death, physical frailty) in various populations [20]. A nutritional supplementation trial in Tanzania and Zambia similarly demonstrates that low handgrip strength is associated with elevated risk of mortality among ART-naïve PLHIV, independent of BMI [31]. The observed difference in the association between median handgrip strength and incident ADI by sex may imply possible effect measure modification by sex. However, the validity of such implication remains questionable due to lack physiological basis in the current literature. Limited studies that assess handgrip strength as a predictor of adverse health outcomes report on effect measure modification by sex. A study of 35–74 year-old Adult Health Study (AHS) cohort members in Japan shows significant difference in the association between low handgrip strength and mortality by sex—low handgrip strength (compared to the sex- and age-specific third quintile) was associated with higher risk of mortality in women aged between 35 and 54, but not in men [48]. However, the same study also reveals that the association between 5kg-increment of grip (which is similar to the continuous handgrip strength measurement used in this study) and all-cause mortality does not differ significantly by sex. It is possible that the observed difference in the effect measure by sex in this study may be due to the differential manifestations of malnutrition between men and women.

A primary limitation in this study is that we did not control for baseline immunosuppression (i.e., CD4 count < 200 cells/mm³), a common predictor for HIV advanced disease progression [46], in our primary analysis due to significantly reduced sample size and event count. In the sensitivity analysis where we adjusted for baseline CD4, the associations of incident ADI with MUAC and with loss of appetite no longer reached statistical significance. This was likely in part due to loss of study power when controlling for baseline CD4. Even if the associations between incident ADI and the three nutritional predictors in the final model were truly confounded by baseline CD4, the results of this study may still be of great relevance for informing the clinical practices of identifying HIV patients at risk of hastened disease progression in settings where clinicians may not have sufficient laboratory resources to measure CD4 cell counts. In addition, we did not routinely collect data on participants’ socioeconomic status, previous events of ADI prior to the study period, or time of HIV infection. The inability to control for such characteristics may allow for residual confounding. The small sample size of the participants with baseline ADI did also not allow us to conduct a complete stratified assessment by baseline ADI. This study was conducted when WHO guidelines recommended PLHIV whose CD4 cell counts drop below 200 cells/mm³ initiate ARVs. With the recent ARV guideline of “Test and Treat” issued by WHO since 2015 [49, 50], we may expect a substantial decrease in the number of tested PLHIV who remain ARV-naïve in rural sub-Saharan Africa. This trend may affect the clinical applicability of our study findings in non-resource scarce regions. Lastly, weight loss, appetite loss, and household food security were self-reported and therefore subject to information bias.

The programmatic implications of this analysis may include providing nutritional supplements to HIV-infected populations at risk of malnutrition, which may impact both weight gain and treatment adherence [51]. A similar suggestion has also been made in one of the NUSTART trial studies, where vitamin and mineral supplementation was shown to improve CD4 count among PLHIV starting ART [49]. Other implications may include conducting further research to develop cost-effective HIV prognostic tools using a set of nutritional measurements. With the 2015 “Test and Treat” ARV initiation recommendation, the size of the eligible populations is expected to significantly increase worldwide [49]. This increased demand in ARV could aggravate the issue of sporadic drug shortage. In Zambia, about 24% of the PLHIV reported to have experienced ARV supply shortage from time to time according to a cross-sectional questionnaire conducted in Southern province in 2017 [52]. Moreover, about 39% of HIV-infected Zambians still were not retained in ART care 12 months after initiating the universal testing and treatment intervention based on the 2015 WHO guideline [53]. As a result, ensuring uninterrupted supply of ARV becomes a primary programmatic focus particularly in resource-scarce regions where sporadic drug supply and malnutrition may be more prevalent. Developing such a prognostic tool for acute disease progression could potentially encourage efficient allocation of medical resources and attention in those regions.

Conclusion

The results of our analysis contribute to the growing body of knowledge that emphasizes the clinical importance of nutritional maintenance among PLHIV living in resource-scarce setting. Based on our findings, loss of handgrip strength for women, self-reported loss of appetite, and low MUAC independently predicted short-term incident ADI in ART-naïve adults in rural Zambia. Given the minimal requirement for measurement equipment, screening for low MUAC and self-reported loss of appetite may have practical values for potential field application in resource-poor clinics. One limitation of this study was that the power to statistically validate the association between the three nutritional indicators and incident ADI decreased after adjusting for baseline CD4 or excluding participants with baseline ADI. As a result, further research should be conducted on these nutritional indicators with larger sample size to validate their prognostic relationship with disease progression, as well as examine the presence of effect measure modification by sex.

Supporting information

S1 Fig. Original questionnaire used by MSF to assess nutritional status among HIV Zambians in Kapiri Mposhi, 2008–2009.

https://doi.org/10.1371/journal.pone.0219111.s001

(PDF)

Acknowledgments

The authors thank the clinical staff residing in MSF HIV clinic at Kapiri Mposhi for their great efforts in implementing the collection of nutrition and clinical assessment data from the study participants in the 9 months of study period, and their firm dedication to the care of HIV patients. In addition, we would like to express our gratitude to Prof. Mike Golden for inspiring the concept of the study and the Research Ethic Committee at University of Zambia for providing logistical support to ensure full protection of the rights and welfare for our study participants. We also thank the Rollins School of Public Health at Emory University, and the Georgia State University’s School of Public Health for providing technological capacity necessary for conducting statistical analyses.

References

1. Skurnick JH, Bogden JD, Baker H, Kemp FW, Sheffet A, Quattrone G, et al. Micronutrient profiles in HIV-1-infected heterosexual adults. Journal of acquired immune deficiency syndromes and human retrovirology: official publication of the International Retrovirology Association. 1996;12(1):75–83. Epub 1996/05/01. pmid:8624765.
- View Article
- PubMed/NCBI
- Google Scholar
2. Mehra D, De Pee S, Bloem M. Food insecurity and public health. Ivers L, editor. Boca Raton, FL: CRC Press; 2015.
3. Koethe JR, Heimburger DC. Nutritional aspects of HIV-associated wasting in sub-Saharan Africa. The American journal of clinical nutrition. 2010;91(4):1138s–42s. Epub 2010/02/12. pmid:20147470; PubMed Central PMCID: PMC2844686.
- View Article
- PubMed/NCBI
- Google Scholar
4. HIV and AIDS in Zambia Brighton, UK: Avert.org; 2018. Available from: https://www.avert.org/professionals/hiv-around-world/sub-saharan-africa/zambia.
5. Antiretroviral therapy (ART) coverage among all age groups: World Health Organization; 2017. Available from: https://www.who.int/gho/hiv/epidemic_response/ART_text/en/.
6. Young S, Wheeler AC, McCoy SI, Weiser SD. A review of the role of food insecurity in adherence to care and treatment among adult and pediatric populations living with HIV and AIDS. AIDS and behavior. 2014;18 Suppl 5:S505–15. Epub 2013/07/12. pmid:23842717; PubMed Central PMCID: PMC3888651.
- View Article
- PubMed/NCBI
- Google Scholar
7. Malvy E, Thiébaut R, Marimoutou C, Dabis F, Groupe d'Epidemiologie Clinique du Sida en A. Weight loss and body mass index as predictors of HIV disease progression to AIDS in adults. Aquitaine cohort, France, 1985–1997. J Am Coll Nutr. 2001;20(6):609–15. pmid:11771676.
- View Article
- PubMed/NCBI
- Google Scholar
8. van der Sande MA, Schim van der Loeff MF, Aveika AA, Sabally S, Togun T, Sarge-Njie R, et al. Body mass index at time of HIV diagnosis: a strong and independent predictor of survival. Journal of acquired immune deficiency syndromes (1999). 2004;37(2):1288–94. Epub 2004/09/24. pmid:15385737.
- View Article
- PubMed/NCBI
- Google Scholar
9. Kelly P, Zulu I, Amadi B, Munkanta M, Banda J, Rodrigues LC, et al. Morbidity and nutritional impairment in relation to CD4 count in a Zambian population with high HIV prevalence. Acta tropica. 2002;83(2):151–8. Epub 2002/06/29. pmid:12088856.
- View Article
- PubMed/NCBI
- Google Scholar
10. Liu E, Spiegelman D, Semu H, Hawkins C, Chalamilla G, Aveika A, et al. Nutritional status and mortality among HIV-infected patients receiving antiretroviral therapy in Tanzania. The Journal of infectious diseases. 2011;204(2):282–90. Epub 2011/06/16. pmid:21673040.
- View Article
- PubMed/NCBI
- Google Scholar
11. Gustafson P, Gomes VF, Vieira CS, Samb B, Naucler A, Aaby P, et al. Clinical predictors for death in HIV-positive and HIV-negative tuberculosis patients in Guinea-Bissau. Infection. 2007;35(2):69–80. Epub 2007/04/03. pmid:17401710.
- View Article
- PubMed/NCBI
- Google Scholar
12. Zachariah R, Fitzgerald M, Massaquoi M, Pasulani O, Arnould L, Makombe S, et al. Risk factors for high early mortality in patients on antiretroviral treatment in a rural district of Malawi. AIDS (London, England). 2006;20(18):2355–60. Epub 2006/11/23. pmid:17117022.
- View Article
- PubMed/NCBI
- Google Scholar
13. Koethe JR, Jenkins CA, Shepherd BE, Stinnette SE, Sterling TR. An optimal body mass index range associated with improved immune reconstitution among HIV-infected adults initiating antiretroviral therapy. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America. 2011;53(9):952–60. Epub 2011/09/29. pmid:21946189; PubMed Central PMCID: PMC3189168.
- View Article
- PubMed/NCBI
- Google Scholar
14. Anand D, Puri S. Anthropometric and Nutritional Profile of People Living with HIV and AIDS in India: an Assessment. Indian J Community Med. 2014;39(3):161–8. pmid:25136157.
- View Article
- PubMed/NCBI
- Google Scholar
15. Koethe JR, Lukusa A, Giganti MJ, Chi BH, Nyirenda CK, Limbada MI, et al. Association between weight gain and clinical outcomes among malnourished adults initiating antiretroviral therapy in Lusaka, Zambia. Journal of acquired immune deficiency syndromes (1999). 2010;53(4):507–13. Epub 2009/09/05. pmid:19730111; PubMed Central PMCID: PMC3749827.
- View Article
- PubMed/NCBI
- Google Scholar
16. Humphreys J, de la Maza P, Hirsch S, Barrera G, Gattas V, Bunout D. Muscle strength as a predictor of loss of functional status in hospitalized patients. Nutrition (Burbank, Los Angeles County, Calif). 2002;18(7–8):616–20. Epub 2002/07/03. pmid:12093441.
- View Article
- PubMed/NCBI
- Google Scholar
17. Hunt DR, Rowlands BJ, Johnston D. Hand grip strength—a simple prognostic indicator in surgical patients. JPEN Journal of parenteral and enteral nutrition. 1985;9(6):701–4. Epub 1985/11/01. pmid:4068194.
- View Article
- PubMed/NCBI
- Google Scholar
18. Koopman JJ, van Bodegom D, van Heemst D, Westendorp RG. Handgrip strength, ageing and mortality in rural Africa. Age and ageing. 2015;44(3):465–70. Epub 2014/10/22. pmid:25331975; PubMed Central PMCID: PMC4411221.
- View Article
- PubMed/NCBI
- Google Scholar
19. Xu L, Hao YT. Effect of handgrip on coronary artery disease and myocardial infarction: a Mendelian randomization study. Scientific reports. 2017;7(1):954. Epub 2017/04/21. pmid:28424468; PubMed Central PMCID: PMC5430422.
- View Article
- PubMed/NCBI
- Google Scholar
20. Norman K, Stobaus N, Gonzalez MC, Schulzke JD, Pirlich M. Hand grip strength: outcome predictor and marker of nutritional status. Clinical nutrition (Edinburgh, Scotland). 2011;30(2):135–42. Epub 2010/11/03. pmid:21035927.
- View Article
- PubMed/NCBI
- Google Scholar
21. Jeejeebhoy KN. Nutritional assessment. Gastroenterology clinics of North America. 1998;27(2):347–69. Epub 1998/07/03. pmid:9650021.
- View Article
- PubMed/NCBI
- Google Scholar
22. Ndubani P, Kamwanga J, Tate J, Buckner B. PLACE in Zambia: Identifying Gaps in HIV Prevention in Kapiri Mposhi, Central Province, 2005. 2006.
23. WHO case definitions of HIV for surveillance and revised clinical staging and immunological classification of HIV-related disease in adults and children. France: World Health Organization, 2007.
24. Ferro-Luzzi A, James WPT. Adult malnutrition: simple assessment techniques for use in emergencies. British Journal of Nutrition. 1996;75(1):3–10. Epub 2007/03/09. pmid:8785188
- View Article
- PubMed/NCBI
- Google Scholar
25. Giri PA, Deshpande JD, Phalke DB. Prevalence of Pulmonary Tuberculosis Among HIV Positive Patients Attending Antiretroviral Therapy Clinic. North American journal of medical sciences. 2013;5(6):367–70. Epub 2013/08/08. pmid:23923111; PubMed Central PMCID: PMC3731868.
- View Article
- PubMed/NCBI
- Google Scholar
26. Gupta RK, Lucas SB, Fielding KL, Lawn SD. Prevalence of tuberculosis in post-mortem studies of HIV-infected adults and children in resource-limited settings: a systematic review and meta-analysis. AIDS (London, England). 2015;29(15):1987–2002. Epub 2015/08/13. pmid:26266773; PubMed Central PMCID: PMC4568896.
- View Article
- PubMed/NCBI
- Google Scholar
27. Koethe JR, Blevins M, Bosire C, Nyirenda C, Kabagambe EK, Mwango A, et al. Self-reported dietary intake and appetite predict early treatment outcome among low-BMI adults initiating HIV treatment in sub-Saharan Africa. Public Health Nutr. 2013;16(3):549–58. Epub 2012/06/13. pmid:22691872.
- View Article
- PubMed/NCBI
- Google Scholar
28. Watson R. Health of HIV infected people: food, nutrition and lifestyle with antiretroviral drugs: Academic Press; 2015.
29. Duggal S, Chugh TD, Duggal AK. HIV and malnutrition: effects on immune system. Clinical & developmental immunology. 2012;2012:784740. Epub 2012/01/14. pmid:22242039; PubMed Central PMCID: PMC3254022.
- View Article
- PubMed/NCBI
- Google Scholar
30. Rehman AM, Woodd S, Chisenga M, Siame J, Sampson G, PrayGod G, et al. Appetite testing in HIV-infected African adults recovering from malnutrition and given antiretroviral therapy. Public Health Nutr. 2015;18(4):742–51. Epub 2014/05/03. pmid:24785906.
- View Article
- PubMed/NCBI
- Google Scholar
31. Woodd SL, Kelly P, Koethe JR, Praygod G, Rehman AM, Chisenga M, et al. Risk factors for mortality among malnourished HIV-infected adults eligible for antiretroviral therapy. BMC infectious diseases. 2016;16(1):562. Epub 2016/10/14. pmid:27733134; PubMed Central PMCID: PMC5062813.
- View Article
- PubMed/NCBI
- Google Scholar
32. Wood R, Middelkoop K, Myer L, Grant AD, Whitelaw A, Lawn SD, et al. Undiagnosed tuberculosis in a community with high HIV prevalence: implications for tuberculosis control. American journal of respiratory and critical care medicine. 2007;175(1):87–93. Epub 2006/09/16. pmid:16973982; PubMed Central PMCID: PMC1899262.
- View Article
- PubMed/NCBI
- Google Scholar
33. Mfinanga S, Chanda D, Kivuyo SL, Guinness L, Bottomley C, Simms V, et al. Cryptococcal meningitis screening and community-based early adherence support in people with advanced HIV infection starting antiretroviral therapy in Tanzania and Zambia: an open-label, randomised controlled trial. Lancet (London, England). 2015;385(9983):2173–82. Epub 2015/03/15. pmid:25765698.
- View Article
- PubMed/NCBI
- Google Scholar
34. Weiser SD, Young SL, Cohen CR, Kushel MB, Tsai AC, Tien PC, et al. Conceptual framework for understanding the bidirectional links between food insecurity and HIV/AIDS. The American journal of clinical nutrition. 2011;94(6):1729s–39s. Epub 2011/11/18. pmid:22089434; PubMed Central PMCID: PMC3226026.
- View Article
- PubMed/NCBI
- Google Scholar
35. Gorbach SL, Knox TA, Roubenoff R. Interactions between nutrition and infection with human immunodeficiency virus. Nutrition reviews. 1993;51(8):226–34. Epub 1993/08/01. pmid:8302492.
- View Article
- PubMed/NCBI
- Google Scholar
36. Wheeler D. Weight loss and disease progression in HIV infection. The AIDS reader. 1999;9(5):347–53. pmid:12737124
- View Article
- PubMed/NCBI
- Google Scholar
37. Donaldson S, Grant-Vallone E. Understanding self-report bias in organizational behavior research. Journal of business and Psychology. 2002;17(2):245–60.
- View Article
- Google Scholar
38. Pinstrup-Andersen P. Food security: definition and measurement. Food Secur 1: 5–7. 2009.
- View Article
- Google Scholar
39. Kikuchi H, Yoshiuchi K, Inada S, Ando T, Yamamoto Y. Development of an ecological momentary assessment scale for appetite. BioPsychoSocial Medicine. 2015;9(1):2. pmid:25614760
- View Article
- PubMed/NCBI
- Google Scholar
40. Oliveira I, Andersen A, Furtado A, Medina C, da Silva D, da Silva ZJ, et al. Assessment of simple risk markers for early mortality among HIV-infected patients in Guinea-Bissau: a cohort study. BMJ open. 2012;2(6). Epub 2012/11/16. pmid:23151393; PubMed Central PMCID: PMC3532999.
- View Article
- PubMed/NCBI
- Google Scholar
41. Brito NB, Llanos JPS, Ferrer MF, García JGO, Brito ID, Castro FP-G, et al. Relationship between mid-upper arm circumference and body mass index in inpatients. PLoS One. 2016;11(8):e0160480. pmid:27494612
- View Article
- PubMed/NCBI
- Google Scholar
42. Nekatebeb H, Taffese S, Sherefa Z. Body mass index and mid upper arm circumference measurement at the HIV/AIDS clinic in Saint Peter’s Hospital, Addis Ababa, Ethiopia. Federation of American Societies for Experimental Biology; 2010.
- View Article
- Google Scholar
43. Powell-Tuck J, Hennessy EM. A comparison of mid upper arm circumference, body mass index and weight loss as indices of undernutrition in acutely hospitalized patients. Clinical Nutrition. 2003;22(3):307–12. pmid:12765671
- View Article
- PubMed/NCBI
- Google Scholar
44. Sharma A, Hoover DR, Shi Q, Gustafson D, Plankey MW, Hershow RC, et al. Relationship between Body Mass Index and Mortality in HIV-Infected HAART Users in the Women's Interagency HIV Study. PLoS One. 2015;10(12):e0143740. Epub 2015/12/25. pmid:26699870; PubMed Central PMCID: PMC4689347.
- View Article
- PubMed/NCBI
- Google Scholar
45. Maro I, Lahey T, MacKenzie T, Mtei L, Bakari M, Matee M, et al. Low BMI and falling BMI predict HIV-associated tuberculosis: a prospective study in Tanzania. Int J Tuberc Lung Dis. 2010;14(11):1447–53. pmid:20937186.
- View Article
- PubMed/NCBI
- Google Scholar
46. Benitez Brito N, Suarez Llanos JP, Fuentes Ferrer M, Oliva Garcia JG, Delgado Brito I, Pereyra-Garcia Castro F, et al. Relationship between Mid-Upper Arm Circumference and Body Mass Index in Inpatients. PLoS One. 2016;11(8):e0160480. Epub 2016/08/06. pmid:27494612; PubMed Central PMCID: PMC4975446.
- View Article
- PubMed/NCBI
- Google Scholar
47. Sultana T, Karim MN, Ahmed T, Hossain MI. Assessment of under nutrition of Bangladeshi adults using anthropometry: can body mass index be replaced by mid-upper-arm-circumference? PLoS One. 2015;10(4):e0121456. Epub 2015/04/16. pmid:25875397; PubMed Central PMCID: PMC4397021.
- View Article
- PubMed/NCBI
- Google Scholar
48. Sasaki H, Kasagi F, Yamada M, Fujita S. Grip strength predicts cause-specific mortality in middle-aged and elderly persons. The American journal of medicine. 2007;120(4):337–42. Epub 2007/04/03. pmid:17398228.
- View Article
- PubMed/NCBI
- Google Scholar
49. Filteau S, PrayGod G, Kasonka L, Woodd S, Rehman AM, Chisenga M, et al. Effects on mortality of a nutritional intervention for malnourished HIV-infected adults referred for antiretroviral therapy: a randomised controlled trial. BMC Medicine. 2015;13(1):17. pmid:25630368
- View Article
- PubMed/NCBI
- Google Scholar
50. HIV/AIDS: Countries should take up new WHO ‘test and treat’ guidelines [Internet]. Medecins San Frontieres; 2015; 24 November [cited 22 August 2017]. Available from: http://www.msf.org/en/article/hivaids-countries-should-take-new-who-%E2%80%98test-and-treat%E2%80%99-guidelines.
51. Audain KA, Zotor FB, Amuna P, Ellahi B. Food supplementation among HIV-infected adults in Sub-Saharan Africa: impact on treatment adherence and weight gain. Proc Nutr Soc. 2015;74(4):517–25. Epub 2015/03/13. pmid:25761769.
- View Article
- PubMed/NCBI
- Google Scholar
52. Paul R, Panchal S, Zimba K. Description of cases of parasuicides reported at University Teaching Hospital, Lusaka. Zambia: preliminary findings. The Health Press.14.
53. Hayes R, Floyd S, Schaap A, Shanaube K, Bock P, Sabapathy K, et al. A universal testing and treatment intervention to improve HIV control: one-year results from intervention communities in Zambia in the HPTN 071 (PopART) cluster-randomised trial. PLoS medicine. 2017;14(5):e1002292. pmid:28464041
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Skurnick JH, Bogden JD, Baker H, Kemp FW, Sheffet A, Quattrone G, et al. Micronutrient profiles in HIV-1-infected heterosexual adults. Journal of acquired immune deficiency syndromes and human retrovirology: official publication of the International Retrovirology Association. 1996;12(1):75–83. Epub 1996/05/01. pmid:8624765.
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Mehra D, De Pee S, Bloem M. Food insecurity and public health. Ivers L, editor. Boca Raton, FL: CRC Press; 2015.

[ref3] 3. Koethe JR, Heimburger DC. Nutritional aspects of HIV-associated wasting in sub-Saharan Africa. The American journal of clinical nutrition. 2010;91(4):1138s–42s. Epub 2010/02/12. pmid:20147470; PubMed Central PMCID: PMC2844686.
View Article
PubMed/NCBI
Google Scholar

[7] View Article

[8] PubMed/NCBI

[9] Google Scholar

[ref4] 4. HIV and AIDS in Zambia Brighton, UK: Avert.org; 2018. Available from: https://www.avert.org/professionals/hiv-around-world/sub-saharan-africa/zambia.

[ref5] 5. Antiretroviral therapy (ART) coverage among all age groups: World Health Organization; 2017. Available from: https://www.who.int/gho/hiv/epidemic_response/ART_text/en/.

[ref6] 6. Young S, Wheeler AC, McCoy SI, Weiser SD. A review of the role of food insecurity in adherence to care and treatment among adult and pediatric populations living with HIV and AIDS. AIDS and behavior. 2014;18 Suppl 5:S505–15. Epub 2013/07/12. pmid:23842717; PubMed Central PMCID: PMC3888651.
View Article
PubMed/NCBI
Google Scholar

[13] View Article

[14] PubMed/NCBI

[15] Google Scholar

[ref7] 7. Malvy E, Thiébaut R, Marimoutou C, Dabis F, Groupe d'Epidemiologie Clinique du Sida en A. Weight loss and body mass index as predictors of HIV disease progression to AIDS in adults. Aquitaine cohort, France, 1985–1997. J Am Coll Nutr. 2001;20(6):609–15. pmid:11771676.
View Article
PubMed/NCBI
Google Scholar

[17] View Article

[18] PubMed/NCBI

[19] Google Scholar

[ref8] 8. van der Sande MA, Schim van der Loeff MF, Aveika AA, Sabally S, Togun T, Sarge-Njie R, et al. Body mass index at time of HIV diagnosis: a strong and independent predictor of survival. Journal of acquired immune deficiency syndromes (1999). 2004;37(2):1288–94. Epub 2004/09/24. pmid:15385737.
View Article
PubMed/NCBI
Google Scholar

[21] View Article

[22] PubMed/NCBI

[23] Google Scholar

[ref9] 9. Kelly P, Zulu I, Amadi B, Munkanta M, Banda J, Rodrigues LC, et al. Morbidity and nutritional impairment in relation to CD4 count in a Zambian population with high HIV prevalence. Acta tropica. 2002;83(2):151–8. Epub 2002/06/29. pmid:12088856.
View Article
PubMed/NCBI
Google Scholar

[25] View Article

[26] PubMed/NCBI

[27] Google Scholar

[ref10] 10. Liu E, Spiegelman D, Semu H, Hawkins C, Chalamilla G, Aveika A, et al. Nutritional status and mortality among HIV-infected patients receiving antiretroviral therapy in Tanzania. The Journal of infectious diseases. 2011;204(2):282–90. Epub 2011/06/16. pmid:21673040.
View Article
PubMed/NCBI
Google Scholar

[29] View Article

[30] PubMed/NCBI

[31] Google Scholar

[ref11] 11. Gustafson P, Gomes VF, Vieira CS, Samb B, Naucler A, Aaby P, et al. Clinical predictors for death in HIV-positive and HIV-negative tuberculosis patients in Guinea-Bissau. Infection. 2007;35(2):69–80. Epub 2007/04/03. pmid:17401710.
View Article
PubMed/NCBI
Google Scholar

[33] View Article

[34] PubMed/NCBI

[35] Google Scholar

[ref12] 12. Zachariah R, Fitzgerald M, Massaquoi M, Pasulani O, Arnould L, Makombe S, et al. Risk factors for high early mortality in patients on antiretroviral treatment in a rural district of Malawi. AIDS (London, England). 2006;20(18):2355–60. Epub 2006/11/23. pmid:17117022.
View Article
PubMed/NCBI
Google Scholar

[37] View Article

[38] PubMed/NCBI

[39] Google Scholar

[ref13] 13. Koethe JR, Jenkins CA, Shepherd BE, Stinnette SE, Sterling TR. An optimal body mass index range associated with improved immune reconstitution among HIV-infected adults initiating antiretroviral therapy. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America. 2011;53(9):952–60. Epub 2011/09/29. pmid:21946189; PubMed Central PMCID: PMC3189168.
View Article
PubMed/NCBI
Google Scholar

[41] View Article

[42] PubMed/NCBI

[43] Google Scholar

[ref14] 14. Anand D, Puri S. Anthropometric and Nutritional Profile of People Living with HIV and AIDS in India: an Assessment. Indian J Community Med. 2014;39(3):161–8. pmid:25136157.
View Article
PubMed/NCBI
Google Scholar

[45] View Article

[46] PubMed/NCBI

[47] Google Scholar

[ref15] 15. Koethe JR, Lukusa A, Giganti MJ, Chi BH, Nyirenda CK, Limbada MI, et al. Association between weight gain and clinical outcomes among malnourished adults initiating antiretroviral therapy in Lusaka, Zambia. Journal of acquired immune deficiency syndromes (1999). 2010;53(4):507–13. Epub 2009/09/05. pmid:19730111; PubMed Central PMCID: PMC3749827.
View Article
PubMed/NCBI
Google Scholar

[49] View Article

[50] PubMed/NCBI

[51] Google Scholar

[ref16] 16. Humphreys J, de la Maza P, Hirsch S, Barrera G, Gattas V, Bunout D. Muscle strength as a predictor of loss of functional status in hospitalized patients. Nutrition (Burbank, Los Angeles County, Calif). 2002;18(7–8):616–20. Epub 2002/07/03. pmid:12093441.
View Article
PubMed/NCBI
Google Scholar

[53] View Article

[54] PubMed/NCBI

[55] Google Scholar

[ref17] 17. Hunt DR, Rowlands BJ, Johnston D. Hand grip strength—a simple prognostic indicator in surgical patients. JPEN Journal of parenteral and enteral nutrition. 1985;9(6):701–4. Epub 1985/11/01. pmid:4068194.
View Article
PubMed/NCBI
Google Scholar

[57] View Article

[58] PubMed/NCBI

[59] Google Scholar

[ref18] 18. Koopman JJ, van Bodegom D, van Heemst D, Westendorp RG. Handgrip strength, ageing and mortality in rural Africa. Age and ageing. 2015;44(3):465–70. Epub 2014/10/22. pmid:25331975; PubMed Central PMCID: PMC4411221.
View Article
PubMed/NCBI
Google Scholar

[61] View Article

[62] PubMed/NCBI

[63] Google Scholar

[ref19] 19. Xu L, Hao YT. Effect of handgrip on coronary artery disease and myocardial infarction: a Mendelian randomization study. Scientific reports. 2017;7(1):954. Epub 2017/04/21. pmid:28424468; PubMed Central PMCID: PMC5430422.
View Article
PubMed/NCBI
Google Scholar

[65] View Article

[66] PubMed/NCBI

[67] Google Scholar

[ref20] 20. Norman K, Stobaus N, Gonzalez MC, Schulzke JD, Pirlich M. Hand grip strength: outcome predictor and marker of nutritional status. Clinical nutrition (Edinburgh, Scotland). 2011;30(2):135–42. Epub 2010/11/03. pmid:21035927.
View Article
PubMed/NCBI
Google Scholar

[69] View Article

[70] PubMed/NCBI

[71] Google Scholar

[ref21] 21. Jeejeebhoy KN. Nutritional assessment. Gastroenterology clinics of North America. 1998;27(2):347–69. Epub 1998/07/03. pmid:9650021.
View Article
PubMed/NCBI
Google Scholar

[73] View Article

[74] PubMed/NCBI

[75] Google Scholar

[ref22] 22. Ndubani P, Kamwanga J, Tate J, Buckner B. PLACE in Zambia: Identifying Gaps in HIV Prevention in Kapiri Mposhi, Central Province, 2005. 2006.

[ref23] 23. WHO case definitions of HIV for surveillance and revised clinical staging and immunological classification of HIV-related disease in adults and children. France: World Health Organization, 2007.

[ref24] 24. Ferro-Luzzi A, James WPT. Adult malnutrition: simple assessment techniques for use in emergencies. British Journal of Nutrition. 1996;75(1):3–10. Epub 2007/03/09. pmid:8785188
View Article
PubMed/NCBI
Google Scholar

[79] View Article

[80] PubMed/NCBI

[81] Google Scholar

[ref25] 25. Giri PA, Deshpande JD, Phalke DB. Prevalence of Pulmonary Tuberculosis Among HIV Positive Patients Attending Antiretroviral Therapy Clinic. North American journal of medical sciences. 2013;5(6):367–70. Epub 2013/08/08. pmid:23923111; PubMed Central PMCID: PMC3731868.
View Article
PubMed/NCBI
Google Scholar

[83] View Article

[84] PubMed/NCBI

[85] Google Scholar

[ref26] 26. Gupta RK, Lucas SB, Fielding KL, Lawn SD. Prevalence of tuberculosis in post-mortem studies of HIV-infected adults and children in resource-limited settings: a systematic review and meta-analysis. AIDS (London, England). 2015;29(15):1987–2002. Epub 2015/08/13. pmid:26266773; PubMed Central PMCID: PMC4568896.
View Article
PubMed/NCBI
Google Scholar

[87] View Article

[88] PubMed/NCBI

[89] Google Scholar

[ref27] 27. Koethe JR, Blevins M, Bosire C, Nyirenda C, Kabagambe EK, Mwango A, et al. Self-reported dietary intake and appetite predict early treatment outcome among low-BMI adults initiating HIV treatment in sub-Saharan Africa. Public Health Nutr. 2013;16(3):549–58. Epub 2012/06/13. pmid:22691872.
View Article
PubMed/NCBI
Google Scholar

[91] View Article

[92] PubMed/NCBI

[93] Google Scholar

[ref28] 28. Watson R. Health of HIV infected people: food, nutrition and lifestyle with antiretroviral drugs: Academic Press; 2015.

[ref29] 29. Duggal S, Chugh TD, Duggal AK. HIV and malnutrition: effects on immune system. Clinical & developmental immunology. 2012;2012:784740. Epub 2012/01/14. pmid:22242039; PubMed Central PMCID: PMC3254022.
View Article
PubMed/NCBI
Google Scholar

[96] View Article

[97] PubMed/NCBI

[98] Google Scholar

[ref30] 30. Rehman AM, Woodd S, Chisenga M, Siame J, Sampson G, PrayGod G, et al. Appetite testing in HIV-infected African adults recovering from malnutrition and given antiretroviral therapy. Public Health Nutr. 2015;18(4):742–51. Epub 2014/05/03. pmid:24785906.
View Article
PubMed/NCBI
Google Scholar

[100] View Article

[101] PubMed/NCBI

[102] Google Scholar

[ref31] 31. Woodd SL, Kelly P, Koethe JR, Praygod G, Rehman AM, Chisenga M, et al. Risk factors for mortality among malnourished HIV-infected adults eligible for antiretroviral therapy. BMC infectious diseases. 2016;16(1):562. Epub 2016/10/14. pmid:27733134; PubMed Central PMCID: PMC5062813.
View Article
PubMed/NCBI
Google Scholar

[104] View Article

[105] PubMed/NCBI

[106] Google Scholar

[ref32] 32. Wood R, Middelkoop K, Myer L, Grant AD, Whitelaw A, Lawn SD, et al. Undiagnosed tuberculosis in a community with high HIV prevalence: implications for tuberculosis control. American journal of respiratory and critical care medicine. 2007;175(1):87–93. Epub 2006/09/16. pmid:16973982; PubMed Central PMCID: PMC1899262.
View Article
PubMed/NCBI
Google Scholar

[108] View Article

[109] PubMed/NCBI

[110] Google Scholar

[ref33] 33. Mfinanga S, Chanda D, Kivuyo SL, Guinness L, Bottomley C, Simms V, et al. Cryptococcal meningitis screening and community-based early adherence support in people with advanced HIV infection starting antiretroviral therapy in Tanzania and Zambia: an open-label, randomised controlled trial. Lancet (London, England). 2015;385(9983):2173–82. Epub 2015/03/15. pmid:25765698.
View Article
PubMed/NCBI
Google Scholar

[112] View Article

[113] PubMed/NCBI

[114] Google Scholar

[ref34] 34. Weiser SD, Young SL, Cohen CR, Kushel MB, Tsai AC, Tien PC, et al. Conceptual framework for understanding the bidirectional links between food insecurity and HIV/AIDS. The American journal of clinical nutrition. 2011;94(6):1729s–39s. Epub 2011/11/18. pmid:22089434; PubMed Central PMCID: PMC3226026.
View Article
PubMed/NCBI
Google Scholar

[116] View Article

[117] PubMed/NCBI

[118] Google Scholar

[ref35] 35. Gorbach SL, Knox TA, Roubenoff R. Interactions between nutrition and infection with human immunodeficiency virus. Nutrition reviews. 1993;51(8):226–34. Epub 1993/08/01. pmid:8302492.
View Article
PubMed/NCBI
Google Scholar

[120] View Article

[121] PubMed/NCBI

[122] Google Scholar

[ref36] 36. Wheeler D. Weight loss and disease progression in HIV infection. The AIDS reader. 1999;9(5):347–53. pmid:12737124
View Article
PubMed/NCBI
Google Scholar

[124] View Article

[125] PubMed/NCBI

[126] Google Scholar

[ref37] 37. Donaldson S, Grant-Vallone E. Understanding self-report bias in organizational behavior research. Journal of business and Psychology. 2002;17(2):245–60.
View Article
Google Scholar

[128] View Article

[129] Google Scholar

[ref38] 38. Pinstrup-Andersen P. Food security: definition and measurement. Food Secur 1: 5–7. 2009.
View Article
Google Scholar

[131] View Article

[132] Google Scholar

[ref39] 39. Kikuchi H, Yoshiuchi K, Inada S, Ando T, Yamamoto Y. Development of an ecological momentary assessment scale for appetite. BioPsychoSocial Medicine. 2015;9(1):2. pmid:25614760
View Article
PubMed/NCBI
Google Scholar

[134] View Article

[135] PubMed/NCBI

[136] Google Scholar

[ref40] 40. Oliveira I, Andersen A, Furtado A, Medina C, da Silva D, da Silva ZJ, et al. Assessment of simple risk markers for early mortality among HIV-infected patients in Guinea-Bissau: a cohort study. BMJ open. 2012;2(6). Epub 2012/11/16. pmid:23151393; PubMed Central PMCID: PMC3532999.
View Article
PubMed/NCBI
Google Scholar

[138] View Article

[139] PubMed/NCBI

[140] Google Scholar

[ref41] 41. Brito NB, Llanos JPS, Ferrer MF, García JGO, Brito ID, Castro FP-G, et al. Relationship between mid-upper arm circumference and body mass index in inpatients. PLoS One. 2016;11(8):e0160480. pmid:27494612
View Article
PubMed/NCBI
Google Scholar

[142] View Article

[143] PubMed/NCBI

[144] Google Scholar

[ref42] 42. Nekatebeb H, Taffese S, Sherefa Z. Body mass index and mid upper arm circumference measurement at the HIV/AIDS clinic in Saint Peter’s Hospital, Addis Ababa, Ethiopia. Federation of American Societies for Experimental Biology; 2010.
View Article
Google Scholar

[146] View Article

[147] Google Scholar

[ref43] 43. Powell-Tuck J, Hennessy EM. A comparison of mid upper arm circumference, body mass index and weight loss as indices of undernutrition in acutely hospitalized patients. Clinical Nutrition. 2003;22(3):307–12. pmid:12765671
View Article
PubMed/NCBI
Google Scholar

[149] View Article

[150] PubMed/NCBI

[151] Google Scholar

[ref44] 44. Sharma A, Hoover DR, Shi Q, Gustafson D, Plankey MW, Hershow RC, et al. Relationship between Body Mass Index and Mortality in HIV-Infected HAART Users in the Women's Interagency HIV Study. PLoS One. 2015;10(12):e0143740. Epub 2015/12/25. pmid:26699870; PubMed Central PMCID: PMC4689347.
View Article
PubMed/NCBI
Google Scholar

[153] View Article

[154] PubMed/NCBI

[155] Google Scholar

[ref45] 45. Maro I, Lahey T, MacKenzie T, Mtei L, Bakari M, Matee M, et al. Low BMI and falling BMI predict HIV-associated tuberculosis: a prospective study in Tanzania. Int J Tuberc Lung Dis. 2010;14(11):1447–53. pmid:20937186.
View Article
PubMed/NCBI
Google Scholar

[157] View Article

[158] PubMed/NCBI

[159] Google Scholar

[ref46] 46. Benitez Brito N, Suarez Llanos JP, Fuentes Ferrer M, Oliva Garcia JG, Delgado Brito I, Pereyra-Garcia Castro F, et al. Relationship between Mid-Upper Arm Circumference and Body Mass Index in Inpatients. PLoS One. 2016;11(8):e0160480. Epub 2016/08/06. pmid:27494612; PubMed Central PMCID: PMC4975446.
View Article
PubMed/NCBI
Google Scholar

[161] View Article

[162] PubMed/NCBI

[163] Google Scholar

[ref47] 47. Sultana T, Karim MN, Ahmed T, Hossain MI. Assessment of under nutrition of Bangladeshi adults using anthropometry: can body mass index be replaced by mid-upper-arm-circumference? PLoS One. 2015;10(4):e0121456. Epub 2015/04/16. pmid:25875397; PubMed Central PMCID: PMC4397021.
View Article
PubMed/NCBI
Google Scholar

[165] View Article

[166] PubMed/NCBI

[167] Google Scholar

[ref48] 48. Sasaki H, Kasagi F, Yamada M, Fujita S. Grip strength predicts cause-specific mortality in middle-aged and elderly persons. The American journal of medicine. 2007;120(4):337–42. Epub 2007/04/03. pmid:17398228.
View Article
PubMed/NCBI
Google Scholar

[169] View Article

[170] PubMed/NCBI

[171] Google Scholar

[ref49] 49. Filteau S, PrayGod G, Kasonka L, Woodd S, Rehman AM, Chisenga M, et al. Effects on mortality of a nutritional intervention for malnourished HIV-infected adults referred for antiretroviral therapy: a randomised controlled trial. BMC Medicine. 2015;13(1):17. pmid:25630368
View Article
PubMed/NCBI
Google Scholar

[173] View Article

[174] PubMed/NCBI

[175] Google Scholar

[ref50] 50. HIV/AIDS: Countries should take up new WHO ‘test and treat’ guidelines [Internet]. Medecins San Frontieres; 2015; 24 November [cited 22 August 2017]. Available from: http://www.msf.org/en/article/hivaids-countries-should-take-new-who-%E2%80%98test-and-treat%E2%80%99-guidelines.

[ref51] 51. Audain KA, Zotor FB, Amuna P, Ellahi B. Food supplementation among HIV-infected adults in Sub-Saharan Africa: impact on treatment adherence and weight gain. Proc Nutr Soc. 2015;74(4):517–25. Epub 2015/03/13. pmid:25761769.
View Article
PubMed/NCBI
Google Scholar

[178] View Article

[179] PubMed/NCBI

[180] Google Scholar

[ref52] 52. Paul R, Panchal S, Zimba K. Description of cases of parasuicides reported at University Teaching Hospital, Lusaka. Zambia: preliminary findings. The Health Press.14.

[ref53] 53. Hayes R, Floyd S, Schaap A, Shanaube K, Bock P, Sabapathy K, et al. A universal testing and treatment intervention to improve HIV control: one-year results from intervention communities in Zambia in the HPTN 071 (PopART) cluster-randomised trial. PLoS medicine. 2017;14(5):e1002292. pmid:28464041
View Article
PubMed/NCBI
Google Scholar

[183] View Article

[184] PubMed/NCBI

[185] Google Scholar

Figures

Abstract

Introduction

Materials and methods

Study location, participants, and design

Outcome of interest

Non-nutritional covariates

Nutritional covariates

Data management and statistical analysis

Results

Baseline non-nutritional and nutritional covariates

Bivariate and multivariate analysis

Sensitivity Analysis

Discussion

Conclusion

Supporting information

S1 Fig. Original questionnaire used by MSF to assess nutritional status among HIV Zambians in Kapiri Mposhi, 2008–2009.

Acknowledgments

References