There is increasing interest globally about malnutrition which directly affects one in three people living in the world [1], and refers to both undernutrition and over-nutrition [2]. Malnutrition and dietary factors are important risk factors for the global burden of diseases such as diabetes [3], cardiovascular disease [4], and tuberculosis [5, 6]. According to the 2016 Global Nutrition report, the economic consequences of malnutrition represent losses of 11% of Gross Domestic Product (GDP) every year in Africa and Asia, whereas preventing malnutrition delivers US$16 in returns on investment for every US$1 spent [1].

Sub-Saharan Africa in particular has the highest prevalence estimates of undernourishment in the world, with 23.2% of its population affected [7]. Likewise, the region has the highest burden of Human Immunodeficiency Virus (HIV) infection, constituting 69% of the estimated 36.7 million people living with HIV globally in 2015 [8]. HIV infection results in functionally defective metabolic ability at the individual level to absorb, store and utilize nutrients thus resulting in nutrient deficiencies, compromised immunity and increased risk of acquiring infectious diseases [9]. Insufficient food intake, together or with malabsorption, result in further progression of HIV-disease [10], and the subsequent weight loss and severe malnutrition that ensue are significant predictors of Acquired Immune Deficiency Syndrome (AIDS) related morbidity and mortality [11]. Despite the high global burden of HIV/AIDS, between 2010 and 2015 there has been more than a two-fold increase in the number of HIV-positive people receiving antiretroviral therapy (ART), which reached 10.3 million in eastern and southern Africa, the world’s most affected regions. The scale up of ART has resulted in AIDS-related deaths in the region decreasing by 36% since 2010. While this is good news, there are certain factors associated with poor outcomes. For example, in sub-Saharan Africa, malnutrition in the form of low body mass index (BMI) is common at ART initiation ranging from 10% to 33% [12–15] and this is associated with poor treatment outcomes and increased mortality [13, 16].

Zimbabwe is one of the sub Saharan countries worst affected by the HIV epidemic with an HIV prevalence of 14.7% (14.66–14.71%) among adults aged 15–49 years according to the 2015 national HIV estimates [17]. This translates to an estimated 1.4 million people aged 15 years and older living with HIV, although as of December 2015, only 788,000 (56%) were enrolled on ART (source = National ART Programme). Zimbabwe’s gross domestic product per capita in 2015 was US$924,10 compared to US$1,588.50 for the whole sub-Saharan Africa region [18]. Currently there is inadequate information on malnutrition prevalence among people living with HIV (PLHIV) in Zimbabwe and anecdotal evidence also suggests that nutritional assessment, care and support for PLHIV are weak. In line with Ministry of Health and Child Care in Zimbabwe (MoHCC) priorities, nutrition in people living HIV is a priority under the focal area of care, treatment and mitigation in the Zimbabwe National HIV/AIDS Strategic Plan (ZNASP) II 2011-2015 document. The MoHCC therefore commissioned a study to better understand the interactions between HIV and nutrition in the country. The study was aimed at determining i) prevalence of malnutrition and ii) factors associated with both undernutrition and over-nutrition among PLHIV enrolled at ART clinics in Zimbabwe.

As shown in Table 3, the majority of respondents (n = 1,152, 93%) were on ART and their median duration on treatment was 3 years (IQR, 1.1–4.3). Self-reported adherence to ART in the month preceding the survey was high with 1,056 (93%) reporting adherence ≥95%. Nearly a third of respondents reported a history of TB treatment whilst 165 (13%) reported experiencing diarrhoea in the two weeks prior to the survey. Slightly more than half of the study respondents (56%) were documented with an illness characterised by WHO clinical stage 3 or 4 at the most recent visit. Only 349 (28%) study participants had a CD4 cell count measurement taken within six months of the interview date, and of these patients, the median CD4 cell count was 348 cells/mL (IQR, 174–510).

The nutritional status according to BMI among enrolled study patients is shown in Fig. 1. The mean BMI for all HIV-positive patients was 23.5 kg / m² (standard deviation 9.4). The majority (n = 776, 63.6%) had a normal BMI, 122 (10%) had under nutrition which was severe in 17 (1.4%), moderate in 18 (1.5%) and mild in 87 (7.1%). There were 224 (18.4%) who were overweight and 98 (8%) with obesity: thus overall 26% of all respondents were classified with over-nutrition.

This is the first nationally representative study in Zimbabwe conducted among HIV-positive patients enrolled in HIV treatment and care settings and aimed at determining prevalence of malnutrition and associated factors. Our study findings show that there is a high burden of malnutrition with about two in five of all HIV-positive patients being affected. Of note, the problem of malnutrition is a double-edged sword with both over-nutrition and undernutrition being prevalent in this population. Over-nutrition affected approximately a quarter of all patients in comparison to under-nutrition which affected only a tenth of all HIV-positive patients. We also observed that undernutrition was more prevalent among males compared to females, among older adolescents and young adults, among those who encountered difficulty accessing food in the past month and those with advanced HIV disease. We also observed that over-nutrition was higher among females, in the older age groups and also increased with higher wealth quintile and increasing CD4 cell count.

The strengths of this study include a large sample and the use of PPS sampling which ensured that our findings are nationally representative of HIV-positive patients enrolled in HIV treatment and care settings in Zimbabwe. The weight of study participants was measured using standardized and calibrated weight scales whilst height was measured using standardized height boards thus ensuring the body mass indices for these patients were current, accurate and comparable. A standardized questionnaire was also administered by trained data collectors and this ensured that all study participants were comparable. Study limitations may have included the possibility of recall bias by study respondents on the type of foods consumed in their households in the week preceding the survey and the frequency of their consumption. This may have led to incorrect calculation and misclassification of the household food consumption score. This study may not be representative of those HIV-positive patients who were enrolled in HIV treatment and care settings but missed their scheduled clinic review visits because they were too sick to attend the clinic. Whilst data were collected for pregnant and lactating women and those aged <15 years old, these were excluded from the analysis of nutritional status since their numbers were too small to allow meaningful inferences and also because body mass index is not an appropriate measure of nutritional status in pregnant and lactating women. Lastly, the use of BMI may underestimate the prevalence of obesity since it is reported to have low sensitivity and can miss a proportion of adults with excess body adiposity [27, 28].

In this study we observed that undernutrition was less common when compared to over-nutrition among HIV-positive adults enrolled in ART clinics in Zimbabwe. This is in sharp contrast to other studies in the African region where malnutrition (BMI < 18.5 kg/m²) was more common and as high as 26.3% [14] and 36% [29]. In one meta-analysis published in 2015, malnutrition (BMI < 18.5 kg/m²) was reported as one of the leading causes of hospital admission among HIV-positive adults in the Africa region [30]. Weight loss which leads to severe and moderate malnutrition is common in HIV infection and can be caused by low dietary intake due to loss of appetite, mouth ulcers, food insecurity and also due to malabsorption of macronutrients and an altered metabolism [31]. Men were more likely to have undernutrition compared to women, and we attribute this to the established fact that they present late for HIV treatment and care with advanced HIV disease and low CD4 counts in Zimbabwe [32] and other African settings [33, 34]. Lower CD4 cell counts have been found to be associated with severe and moderate malnutrition [14, 29] and these are also associated with increased mortality [35]. Similar to other studies, household food insecurity which affects the frequency of meals and dietary diversity of meals consumed in a household has been commonly reported as a predictor of poor nutritional status [36]. There were no differences in levels of undernutrition between PLHIV in rural and urban residences and this could be explained by our findings that show that study participants were generally from poor households given that the majority had household incomes below the national poverty datum line and food poverty line. Whilst urban dwellers may be perceived to have better food security and subsequently better nutritional status due to high economic activity, high levels of food insecurity and consequently malnutrition have been reported in Southern Africa and are related to rapid urbanization resulting in informal settlements [37].

In our sample, there were significantly higher proportions of HIV-positive patients who were overweight or obese compared with those at the other end of the scale. Our data also shows that the prevalence of being overweight and obese was slightly lower among HIV-positive women compared to women in the general population, with this data taken from women in the 2010-11 Zimbabwe Demographic and Health Survey (ZDHS) [38]. Interestingly, the population prevalence of women who were overweight or obese from this ZDHS (31%) was higher than the pooled prevalence for the Sub-Saharan Africa region at 23% [38], and this could be related to the 2011 World Bank poverty headcount ratio at US$1.90 which was 21% for Zimbabwe compared to 44% for sub-Saharan Africa [39]. Increasing CD4 cell count was found to be associated with being overweight and obese and this concurs with findings from a follow-up of 17 ART cohorts in North America with sustained viral suppression [40, 41]. With women being more likely to be overweight or obese relative to men, this may again be related to their earlier entry into HIV care and ART initiation with relatively higher CD4 cell counts compared to men [40, 41]. In many African settings including Zimbabwe, women have better access than men to HIV testing and treatment services through integration of HIV testing services in antenatal care settings [42], which act as an entry point into prevention-of-mother-to- child HIV transmission (PMTCT) services. Given that the majority of HIV-positive individuals enrolled in this study were on ART, there may also be a high prevalence of HIV-related lipodystrophy which has been reported, in one systematic review study, to develop within the first six months of therapy particularly with the use of stavudine [43]. Lipodystropy can predispose HIV-positive individuals to cardiovascular disease [44], and hence the need to evaluate its prevalence, pathogenesis and prognosis in order to inform clinical management given the rapid national scale-up of ART and the subsequent longer survival as HIV-related mortality continues to decline. However, the use of stavudine based antiretroviral regimens has been phased out since 2013 due to these adverse events.

In addition, metabolic complications such as dyslipidemia, insulin resistance, and diabetes mellitus are also common among obese patients with well-controlled HIV-infection who are on long-term ART. These complications are also associated with increased risk of cardiovascular disease. Metabolic complications probably develop due to traditional risk factors such as obesity or genetic predisposition and HIV-specific and ART-specific contributions including chronic inflammation and immune activation [45]. Study participants who were of a higher wealth index were more likely to be overweight and obese but this trend was observed only among the urban participants. Unlike rural dwellers, where food insecurity is an issue, this trend is likely to be due to urban dwellers having better food security, having access to more refined and unhealthy food diets and also having more sedentary lifestyles [46]. Other studies have also found people of higher socioeconomic and wealth status [38] to have a higher prevalence of obesity and they are more likely to develop or have hypertension and diabetes [47].

This study has a number of implications. First, there is a double burden of undernutrition and over-nutrition amongst the country’s population of HIV-positive patients enrolled in ART clinics. It is therefore important that health workers routinely assess and document weight and height measurements in individual health facility records to enable BMI monitoring as a proxy measure of nutritional status. A previous nationally representative record review of ART treatment outcomes among HIV-positive patients enrolled in ART clinics in Zimbabwe showed that approximately a quarter of all patients did not have recorded weight at ART initiation and this must change and improve [40]. Low baseline weight at ART initiation has been shown to increase the risk of ART attrition from care and mortality [40, 48] and given its association with late presentation with advanced HIV, this points towards a need to encourage early seeking of HIV testing services and subsequent ART initiation. Currently, the national ART programme is shifting towards an HIV ‘treat-all’ approach which promotes earlier ART initiation with promising improved treatment outcomes.

Second, the addition of micronutrient supplements to ART may provide increased nutritional recovery among HIV-positive patients with undernutrition [49, 50] and also more importantly improve adherence to ART among food-insecure adults as has been shown in Zambia [51] though no significant gains were observed with respect to weight and CD4 cell count response. However, despite improved health and ART adherence, perceptions among food-insecure HIV-positive adults participating in one Kenyan food supplementation program highlighted the inevitability of sharing food disbursements with people outside their families thus depleting the food available for them [52]. Any adoption of food supplementation programmes in ART clinics will therefore need to take this into consideration so as to ensure that the targeted malnourished HIV-positive patients benefit from these interventions.

Third, with obesity becoming prevalent in the HIV population on ART, HIV may change from being a highly fatal infectious disease into a chronic manageable disease that is associated with an elevated risk of cardiovascular-related conditions such as diabetes mellitus and hypertension. This calls for an urgent need to integrate the screening and management of hypertension and diabetes mellitus into HIV clinics [53], and particularly among patients who are overweight and obese.

In conclusion, we have found that being overweight or obese is more prevalent than being underweight in HIV-positive patients enrolled in HIV treatment and care settings and on ART in Zimbabwe, and this is more common in females, in those of older age and in those with increasing wealth index. Given that over-nutrition predisposes patients to cardiovascular-related conditions, there is an urgent need to integrate into HIV treatment and care services the screening of hypertension and diabetes mellitus in order to enable these conditions to be identified and appropriately managed to prevent cardiovascular complications.