Prevalence, Risk Factors, and Pathophysiology of Dysglycemia among People Living with HIV in Sub-Saharan Africa

Table 1

Summary of studies on prevalence and risk factors of diabetes mellitus and prediabetes among PLHIV.

Author(s) and country (reference number)	Study design and population	Dysglycemia definition	Prevalence	Identified independent risk factors	Comments
Noumegni et al., Cameroon [35]	Cross-sectional: 452 adults age 30–74 years of whom 400 were on ART	DM: FPG ≥7.0 mmol/l on two separate occasions at least 48 hours apart or self-report of taking antidiabetic medicine	DM: 2.0%	BMI ≥ 30 kg/m^2 associated with insulin resistance: OR 2.28	Patients on ART had significantly higher BMI, waist circumference, waist-hip ratio, obesity, and abdominal obesity compared to those not on ART
Chimbetete et al., Zimbabwe [23]	Retrospective: 4110 PLHIV aged ≥ 16 years starting ART	DM: at baseline, an RBS > 11.0 mmol/l in the presence of DM symptoms or FPG > 7.0 mmol/l or known diagnosis of DM prior to ART initiation	DM: 0.77%	Male gender: aHR 2.31; Age > 40 years: aHR 2.32; BMI > 30 kg/m^2: aHR 3.1; (all associated with incident dysglycemia)	While this was an incidence study of 4110 PLHIV starting ART, 42 of the 5467 PLHIV in the initial cohort were excluded due to prevalent DM defined as a known diagnosis of DM or DM diagnosed at the baseline visit
Magodoro et al., Zimbabwe [29]	Retrospective: 1033 PLHIV aged ≥ 18 years on ART	Known diagnosis of DM as per patient records	DM: 2.1%	Associations with dysglycemia not reported	Median duration on ART was 5.3 years; Case ascertainment was not possible as details on how DM diagnosis had been made was not available
Levitt et al., South Africa [27]	Cross-sectional: PLHIV aged ≥ 18 years in three groups: 393 ART-naive PLHIV, 439 PLHIV on 1st line ART, and 108 PLHIV on 2nd line ART	DM: FPG ≥ 7.0 mmol/l or 2 hr glucose ≥ 11.1 mmol/l; IFG: FPG 6.1 mmol/l and <7.0 mmol/l with normal 2 hr glucose; IGT: 2 hr glucose ≥ 7.8–11 mmol/l with FBS < 7.0 mmol/l	DM: On 1st line ART: 2.3%; On 2nd line ART: 5.6%; ART-naive: 3.1%; Pre-DM^a: On 1st line ART: 23.7%; On 2nd line ART: 31.4%; ART-naive: 18.6%	Age (years): 35–44 (OR 1.82), 45–54 (3.27), and 55–64 (OR 4.75); BMI > 30 kg/m^2: OR 1.92; Female gender: OR 2.17; 1st line ART use: OR 2.47; 2nd line ART use: OR 4.1; (all associated with prevalent dysglycemia)	1st line ART regimens comprised dual NRTI plus one NNRTI while 2nd line ART regimens comprised dual NRTI plus a boosted PI; A community-based sample group was also included of 880 participants who were not on ART. Dysglycemia prevalence was lower in this group compared to PLHIV groups; however, their HIV status was not known
Isa et al., Nigeria [25]	Retrospective: 2632 ART-naive PLHIV aged ≥ 18 years	DM: RBS ≥ 11.1 mmol/l or FPG ≥ 7.0 mmol/l or self-reported use of antidiabetic drugs	DM: 2.3%	Age > 40 years associated with prevalent dysglycemia: aOR 3.5; BMI ≥ 25 kg/m^2 associated with incident dysglycemia: aOR 7.5	At one year follow-up after initiating ART, an additional 5.3% of the cohort developed diabetes driving up prevalence to 7.6%
Mohammed et al., Ethiopia [31]	Cross-sectional: 393 PLHIV aged ≥ 21 years of whom 285 were on ART and 109 were ART-naive	DM: FPG ≥ 7.0 mmol/l; IFG: FPG ≥ 6.2 mmol/l and <7.0 mmol/l	DM: 6.4%; IFG: 19.6%	Age ≥ 40 years: aOR 4.8; ART use ≥ 5 years: aOR 26.93; Hypertension: aOR 4.78; LDL-C ≥ 130 mg/dL: aOR 5.67; (all associated with prevalent dysglycemia)	Lack of OGTT may have underestimated the prevalence of DM and pre-DM
Maganga et al., Tanzania [28]	Cross-sectional: Adults aged > 18 years in three groups: 150 PLHIV on ART for ≥2 years, 151 recently diagnosed ART-naive PLHIV, and 153 HIV-negative	DM: FPG ≥ 7.0 mmol/l or 2 hr glucose ≥ 11.1 mmol/l; IFG: FPG 6.1–6.9 mmol/l with normal 2 hr glucose; IGT: 2 hr glucose ≥ 7.8–11 mmol/l with FBS < 7.0 mmol/l	DM: On ART: 18%; ART-naive: 0.7%; HIV (−): 5.2%; Pre-DM^a: On ART: 14.7%; ART-naive: 7.3%; HIV (−): 2%	ART use ≥ 2 years: aOR 5.72 associated with prevalent dysglycemia	HIV-negative participants were not aged- or sex-matched
Oni et al., South Africa [36]	Retrospective: electronic prescription refill records for 32,474 receiving ≥ 1 prescription for HIV, TB, DM, or/and HTN medications	DM: prescription refill for either metformin, glibenclamide, or insulin	DM: 17%	Associations with dysglycemia not reported	Case ascertainment was not possible as details on how DM diagnosis had been made was not available
Kagaruki et al., Tanzania [26]	Cross-sectional: 671 PLHIV aged ≥ 18 years of whom 354 were on ART and 317 were ART-naive	DM: FPG ≥ 6.1 mmol/l or prior known diagnosis	DM: On ART: 3.7%; ART-naive: 4.7%	Associations with dysglycemia not reported	Overall cases of DM were too low to assess between-group difference or associated risk factor relationships; Lack of OGTT may have underestimated DM prevalence
Ngatchou et al., Cameroon [34]	Cross-sectional: 108 ART-naive PLHIV and 96 HIV-negative aged-matched controls	IFG: FPG ≥ 5.6–6.9 mmol/lDM: FPG > 6.9 mmol/l	DM: ART-naive: 26%; HIV (−): 1%; IFG: ART-naive: 47%; HIV (−): 27%	Associations with dysglycemia not reported	Dysglycemia prevalence may have been underestimated due to lack of OGTT and exclusion of patients with known, or on treatment for, DM, hypertension or dyslipidemia, cigarette smokers or alcohol users, and patients with a first-degree familial history of DM
Negin et al., South Africa [33]	Survey: 194 PLHIV and 2864 HIV (−) adults aged ≥ 18 years	Self- report of known DM	DM: PLHIV: 4.1%; HIV (−): 9.7%	Associations with dysglycemia not reported	Case ascertainment was not possible as DM diagnosis based on self-report; Information unavailable for ART use
Dave et al., South Africa [24]	Cross-sectional: 443 PLHIV on ART for ≥6 months and 406 ART-naive PLHIV	DM: FPG ≥ 7.0 mmol/L or 2 hr glucose ≥ 11.0 mmol/l; Pre-DM: FPG ≥ 5.6-7.0 mmol/l or 2 hr glucose ≥ 7.8 mmol/l-11.1 mmol/l	DM: On ART: 2.2%; ART-naive: 3.4%; Pre-DM: On ART: 23.5%; ART-naive: 18.5%	Male gender: OR 1.96; Efavirenz use: OR 1.7; All associated with prevalent dysglycemia	Dysglycemia prevalence difference was not statistically significant between on ART and ART-naive group and may be underestimated by the exclusion of known history of DM or IGT; ART regimen in use was stavudine or zidovudine with lamivudine and nevirapine or efavirenz; Older age (OR 1.04) and CD4 count (OR 1.001) also associated with prevalent dysglycemia but cutoffs not specified
Anastos et al., Rwanda [22]	Cross-sectional: women aged ≥ 25 years divided into two groups: 606 ART-naive PLHIV and 218 HIV-negative	DM: FPG > 6.9 mmol/l or self-reported history of DM	DM: ART-naive PLHIV: 0.5%; HIV (−): 0.5%	Associations with dysglycemia not reported	This analysis was based on the Rwanda Women’s Inter-association Study and Assessment and inclusion was based on the availability of fasting lipoprotein levels and not glucose levels
Manuthu et al., Kenya [30]	Cross-sectional: 134 PLHIV on ART for ≥4 weeks and 161 ART-naive PLHIV	DM: FPG ≥ 7.0 mmol/l or 2 hr glucose ≥ 11.0 mmol/l; IFG: FPG ≥ 6.1 to 6.9 mmol/l; IGT: 2 hr glucose ≥ 7.8 mmol/l–11.1 mmol/l	DM: 1.5%; Pre-DM^a: 20.4%	No significant associations with dysglycemia reported	Excluded patients with known DM status thus may underestimate prevalence
Mutimura et al., Rwanda [32]	Cross-sectional: 150 PLHIV on ART for ≥6 months and 50 HIV (−) controls	Dysglycemia: IFG > 5.6 mmol/l	PLHIV: With LDS: 18%; Without LDS: 16%; HIV (−): 2%	Associations with dysglycemia not reported	Distinction was not made between DM and prediabetes

ADA: American Diabetes Association; aHR: adjusted hazard ratio; aOR: adjusted odds ratio; ART: antiretroviral therapy; FPG: fasting blood glucose; HIV: human immunodeficiency virus; HTN: hypertension; IGT: impaired glucose tolerance; LDL: low-density lipoprotein; LDS: lipodystrophy; NNRTI: nonnucleoside reverse transcriptase inhibitor; NRTI: nucleoside reverse transcriptase inhibitor; OGTT: oral glucose tolerance test; OR: odds ratio; PI: protease inhibitor; PLHIV: people living with HIV; PY: person-years; DM: diabetes mellitus; TB: tuberculosis. Only statistically significant risk factors are reported. ^aPrediabetes definition: impaired fasting glucose or impaired glucose tolerance.