Association between food insecurity and key metabolic risk factors for diet-sensitive non-communicable diseases in sub-Saharan Africa: a systematic review and meta-analysis

Abstract

In previous studies, food insecurity has been hypothesised to promote the prevalence of metabolic risk factors on the causal pathway to diet-sensitive non-communicable diseases (NCDs). This systematic review and meta-analysis aimed to determine the associations between food insecurity and key metabolic risk factors on the causal pathway to diet-sensitive NCDs and estimate the prevalence of key metabolic risk factors among the food-insecure patients in sub-Saharan Africa. This study was guided by the Centre for Reviews and Dissemination (CRD) guidelines for undertaking systematic reviews in healthcare. The following databases were searched for relevant literature: PubMed, EBSCOhost (CINAHL with full text, Health Source - Nursing, MedLine). Epidemiological studies published between January 2015 and June 2019, assessing the associations between food insecurity and metabolic risk outcomes in sub-Saharan African populations, were selected for inclusion. Meta-analysis was performed with DerSimonian-Laird's random-effect model at 95% confidence intervals (CIs). The I² statistics reported the degree of heterogeneity between studies. Publication bias was assessed by visual inspection of the funnel plots for asymmetry, and sensitivity analyses were performed to assess the meta-analysis results' stability. The Mixed Methods Appraisal Tool (MMAT) - Version 2018 was used to appraise included studies critically. The initial searches yielded 11,803 articles, 22 cross-sectional studies were eligible for inclusion, presenting data from 26,609 (46.8% males) food-insecure participants, with 11,545 (42.1% males) reported prevalence of metabolic risk factors. Of the 22 included studies, we identified strong evidence of an adverse association between food insecurity and key metabolic risk factors for diet-sensitive NCDs, based on 20 studies. The meta-analysis showed a significantly high pooled prevalence estimate of key metabolic risk factors among food-insecure participants at 41.8% (95% CI: 33.2% to 50.8%, I² = 99.5% p-value < 0.00) derived from 14 studies. The most prevalent type of metabolic risk factors was dyslipidaemia 27.6% (95% CI: 6.5% to 54.9%), hypertension 24.7% (95% CI: 15.6% to 35.1%), and overweight 15.8% (95% CI: 10.6% to 21.7%). Notably, the prevalence estimates of these metabolic risk factors were considerably more frequent in females than males. In this systematic review and meta-analysis, exposure to food insecurity was adversely associated with a wide spectrum of key metabolic risk factors, such as obesity, dyslipidaemia, hypertension, underweight, and overweight. These findings highlight the need to address food insecurity as an integral part of diet-sensitive NCDs prevention programmes. Further, these findings should guide recommendations on the initiation of food insecurity status screening and treatment in clinical settings as a basic, cost-effective tool in the practice of preventive medicine in sub-Saharan Africa.PROSPERO registration number: PROSPERO 2019 CRD42019136638.

Figure 1

Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) flow diagram of literature search and study selection process. Source: Adopted from Ref.

Figure 2

Map of Africa indicating the regions where the included studies were situated and the number of pooled study participants [Figure generated by the first author using Canva version 2.93.0, Available from: http://about.canva.com/].

Figure 3

The metabolic risk factors prevalence derived from 22 studies by gender strata.

Figure 4

Forest plot showing study-specific and pooled estimates for being underweight and overweight prevalence in Sub-Saharan Africa, from studies published during the period 2015–2019, (A: The pooled prevalence estimates for being underweight; B: The pooled prevalence estimates for being overweight).

Figure 5

Forest plot showing study-specific and (A): The pooled prevalence estimates for hypertension; (B) The pooled prevalence estimates for obesity; (C) The pooled prevalence estimates for dyslipidaemia.

Figure 6

Pooled estimates for metabolic risk factors prevalence by year of publication.