OR I G I N A L AR T I C L E High prevalence of impaired glucose metabolism among children and adolescents living with HIV in Ghana Ruth Ayanful-Torgby1,2 | Veronika Shabanova1,3 | Akosua A. Essuman2 | Emmanuel Boafo2 | Frank Aboagye4 | Yusuf Al-Mahroof4 | Jones Amponsah2 | John K. A. Tetteh2 | Linda E. Amoah2 | Elijah Paintsil1 1Department of Pediatrics, Yale School of Medicine, New Haven, Connecticut, USA 2Department of Immunology, Noguchi Memorial Institute for Medical Research, University of Ghana, Accra, Ghana 3Department of Biostatistics, Yale School of Public Health, New Haven, Connecticut, USA 4Biomedical & Public Research Unit, Council for Scientific & Industrial Research-Water Research Institute, Accra, Ghana Correspondence Elijah Paintsil, Department of Pediatrics, Yale School of Medicine, New Haven, CA, USA. Email: elijah.paintsil@yale.edu Funding information Fogarty International Center and the National Institute of Alcohol Abuse and Alcoholism of the National Institutes of Health, Grant/Award Number: D43 TW011526 Abstract Background: Antiretroviral therapy (ART)-associated metabolic abnormali- ties, including impairment of glucose metabolism, are prevalent in adults liv- ing with HIV. However, the prevalence and pathogenesis of impaired glucose metabolism in children and adolescents living with HIV, particularly in sub- Saharan Africa, are not well characterized. We investigated the prevalence of impaired glucose metabolism among children and adolescents living with peri- natally infected HIV in Ghana. Methods: In this multicentre, cross-sectional study, we recruited participants from 10 paediatric antiretroviral treatment clinics from January to June 2022 in 10 facilities in Greater Accra and Eastern regions of Ghana. We determined impaired glucose metabolism in the study sample by assessing fasting blood sugar (FBS), insulin resistance as defined by the homeostatic model assessment for insulin resistance (HOMA-IR) index and glycated haemoglobin (HbA1c) levels. The prevalence of impaired glucose metabolism using each criterion was stratified by age and sex. The phenotypic correlates of glucose metabolism markers were also assessed among age, sex, body mass index (BMI) and waist- to-hip ratio (WHR). Results: We analysed data from 393 children and adolescents living with HIV aged 6–18 years. A little over half (205/393 or 52.25%) of the children were female. The mean age of the participants was 11.60 years (SD = 3.50), with 122/393 (31.00%) aged 6–9 years, 207/393 (52.67%) aged 10–15 years, and 62/393 (15.78%) aged 16–18 years. The prevalence rates of glucose impairment in the study population were 15.52% [95% confidence interval (CI): 12.26– 19.45], 22.39% (95% CI: 18.54–26.78), and 26.21% (95% CI: 22.10–30.78) using HbA1c, HOMA-IR, and FBS criteria, respectively. Impaired glucose metabo- lism detected by FBS and HOMA-IR was higher in the older age group, whereas the prevalence of abnormal HbA1c levels was highest among the youngest age group. Age and BMI were positively associated with FBS and HOMA-IR (p < 0.001). However, there was negative correlation of WHR with HOMA-IR (p < 0.01) and HbA1c (p = 0.01). Received: 12 October 2023 Accepted: 8 January 2024 DOI: 10.1111/hiv.13614 HIV Med. 2024;1–10. wileyonlinelibrary.com/journal/hiv © 2024 British HIV Association. 1 https://orcid.org/0000-0003-1066-3345 https://orcid.org/0000-0002-3881-9789 https://orcid.org/0009-0005-3803-3013 https://orcid.org/0000-0002-8283-4564 mailto:elijah.paintsil@yale.edu http://wileyonlinelibrary.com/journal/hiv http://crossmark.crossref.org/dialog/?doi=10.1111%2Fhiv.13614&domain=pdf&date_stamp=2024-01-19 Conclusion: The high prevalence of impaired glucose metabolism observed among the children and adolescents living with HIV in sub-Saharan Africa is of concern as this could contribute to the development of metabolic syndrome in adulthood. KEYWORD S adolescents, children, impaired glucose metabolism, paediatric HIV INTRODUCTION Antiretroviral therapy (ART) has significantly improved the life expectancy of people living with HIV (PLWH) [1]. However, long-term use of ART has been associated with increased susceptibility to metabolic-related morbidities [2]. In this population, metabolic-related disorders are linked with persistent defects in the cellular processes, including impaired glucose metabolism. Impaired glucose metabolism is a recognized precursor of various cardiometabolic sequelae and may serve as an predictive marker for the increasing incidence of metabolic disorders among PLWH [3]. Recent studies have observed a higher occur- rence of impaired glucose metabolism among PLWH on ART compared with those not on ART or the HIV- uninfected general population [3, 4]. The prevalence of impaired glucose metabolism in PLWH is estimated to be as high as 32%, compared with a prevalence of 11% in the HIV-uninfected general population [4, 5]. While age is considered a risk factor in the general pop- ulation, the duration of ART usage has been identified as a significant contributor to metabolic abnormalities in PLWH [6]. Therefore, children who acquired HIV mainly through mother-to-child transmission are at a higher risk of developing metabolic disorder due to lifelong exposure to ART from childhood. Studies assessing impaired glucose metabolism in children and adolescents living with HIV (CALWH) have reported varying prevalence rates, ranging from as low as 1.6% to as high as 40%, depending on the population studied [7–9]. However, although impaired glu- cose metabolism in children, including insulin resistance, could be a risk or predictor of cardiovascular disease in adulthood [10, 11], diagnosing glucose impairment in chil- dren is challenging as it often presents asymptomati- cally [12]. Likewise, differences in glucose metabolism between children and adults, attributed to factors such as developmental variations, ageing, and physical inactivity, contribute to subclinical glucose abnormalities in children, which poses diagnostic challenge [13, 14]. A frequently observed impairment in glucose metabo- lism is hyperglycaemia, yet it remains a challenge for the paediatric population living with HIV, particularly among adolescents, where it is highly prevalent [15]. Hyperglycaemia in children living with HIV may impact the prevalence of insulin resistance and the risk of devel- oping diabetes in adulthood. Therefore, in populations with high HIV infections, the prevalence of impaired glu- cose metabolism in children could be high [16]. Although only a few studies have monitored glucose tolerance or sensitivity in children living with HIV in sub-Saharan Africa, where 90% of the paediatric HIV population resides, the reported prevalence of impaired glucose is high compared with the paediatric population without HIV [4, 7, 17–19]. Interestingly, the majority of these individuals with impaired glucose levels are non-obese and primarily exhibit subclinical manifestations [20]. Recent reports have indicated that the prevalence of impaired glucose metabolism in children living with HIV is on the increase [7, 21]. To effectively manage these dis- orders, it is crucial to identify impaired glucose metabo- lism in children and adolescents and develop appropriate management strategies. However, there is limited infor- mation and lack of harmonized criteria to assess the bur- den of impaired glucose metabolism in this population, especially in sub-Saharan Africa. This limitation may result in underreported prevalence of impaired glucose metabolism in this population, including those living in Ghana. Currently, Ghana lacks a comprehensive pro- gramme to monitor or assess the burden of impaired glu- cose in children, and few studies have examined impaired glucose levels using multiple assessment parameters in the same cohort of CALWH. Therefore, the aim of this study was to estimate the prevalence of impaired glucose metabolism using different criteria and investigate potential differences in glucose levels based on age and sex. Additionally, this study sought to describe the phenotypic correlates of glucose metabolism markers in CALWH in Ghana. METHODS Study population This cross-sectional, multi-centre study was conducted from January to June 2022 among children and 2 AYANFUL-TORGBY ET AL. 14681293, 0, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/hiv.13614 by U niversity of G hana - A ccra, W iley O nline L ibrary on [01/02/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense adolescents living with perinatally acquired HIV, aged 6– 18 years. The participants were recruited from 10 paediat- ric ART treatment clinics at 10 health facilities in Greater Accra and the eastern region of Ghana – Ashiaman Poly- clinic, Greater Accra Regional Hospital, Amasaman Gov- ernment Hospital, Weija-Gbawe Municipal Hospital, LEKMA Hospital, Princess Marie Louise Children Hospi- tal, Tema General Hospital, Tema Polyclinic, and Maa- mobi Hospital, and Atua Government Hospital in the eastern region of Ghana. All the study sites are known to have similar socioeconomic indices [22]. The study's inclusion criteria were children and adolescents aged 6– 18 years who had perinatal HIV and were currently on ART. During the sampling period, national ART treatment for children and adolescents prioritized nucleoside reverse transcriptase inhibitor (NRTIs) and non-NRTIs (NNRTIs) for first-line regimens, NRTIs, NNRTIs and protease inhib- itors (PIs) for second-line regimens, and NRTIs, NNRTIs, and integrase strand transfer inhibitors (INSTIs) for third- line regimens. Additionally, individuals with viral hepatitis and tuberculosis followed distinct regimens involving other antivirals or antibiotics [23]. Exclusion criteria included CALWH with coinfections of tuberculosis and viral hepatitis, as these infections and their treatment could introduce potential metabolic cofounding factors such as therapies other than ART. Additionally, individ- uals who were clinically unstable or unwilling to partici- pate were also excluded. Individuals who met the study's criteria and were willing to participate were recruited. Par- ticipants were given a clear explanation of the study's objectives and procedures before consenting. Those who agreed to participate selected a suitable time for the required sampling procedures at the clinic. Study ethical approvals The study received approval from the Institutional Review Boards of Ghana Health Service (protocol no. GHS-ERC: 025/09/21) and Noguchi Memorial Insti- tute for Medical Research, University of Ghana (protocol no. CPN 015/21–22). Written parental informed consent for children aged 6–12 years, parental informed consent and assent for those aged 12–17 years, and consent for 18-year-old participants were obtained before their enrol- ment. Study participants were given a guarantee of ano- nymity and data protection. Data collection Questionnaires were administered to each participant or parent/guardian to extract information regarding demographics, educational background and socioeco- nomic background of their parents/guardian by trained research assistants. Anthropometric parameters Anthropometric measurements were taken by a trained study team member or health staff to record the follow- ing: upper arm circumference, waist circumference, hip circumference and neck circumference taken with the children in light clothing using an inelastic tape measure. Weight and height were measured with a vertical stadi- ometer. Body mass index (BMI) was calculated from weight (kg) and height (cm), and obesity was defined as a BMI ≥ 28 kg/m2. Waist-to-hip ratio (WHR) was calcu- lated from waist and hip circumferences (cm). Measure- ments were recorded to the nearest 1.0 mm, 0.5 cm, and 0.1 kg from an average of two readings. Biochemical parameters All participants underwent overnight fasting (≥10 h) before the early morning blood draw. Beverages and snacks were provided immediately after specimen collection prior to anthropometric readings and administering of the question- naire. We collected 2.5 mL of venous blood by sterile tech- niques from each participant. Whole blood samples were aliquoted in a gel tube (2 mL) for the serum and micro- EDTA (0.5 mL) tube. Fasting blood sugar (FBS) was assessed instantly with a blood drop using a glucometer strip (True Metrix, Trividia Health Inc., Fort Lauderdale, FL, USA). Serum Insulin levels were measured using sand- wich ELISA (R&D Systems, USA) following the manufac- turer's instructions. In brief, we coated the primary antibody (2 μg/mL) onto a 96-well plate and incubated overnight at 4�C (Nunc MaxiSorp; Fisher Scientific, Leices- tershire, UK). 1% bovine serum albumin in phosphate-buff- ered saline (PBS) was used for blocking and 0.05% Tween 20 in PBS was used for the washing steps. Standard solu- tions and serum samples were added to the wells and incu- bated for 2 h. Detection was performed with a secondary antibody (200 ng/mL) and streptavidin-horseradish peroxi- dase B. Substrate solution was added to bind antibodies, and colour development was stopped with 0.2 M H2SO4. The absorbance was measured at 450 nm using a micro- plate reader (Multiskan FC; Thermo Scientific, USA), and the optical density (OD) values were converted into concen- tration from the mean value of duplicate measurements. Glycated haemoglobin (HbA1c) level was determined using an immunofluorescence assay (Med Source, Haryana, India) at an absorbance of 450 nm on a plate reader HIV MEDICINE 3 14681293, 0, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/hiv.13614 by U niversity of G hana - A ccra, W iley O nline L ibrary on [01/02/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense (Varioskan Lux; ThermoFisher Scientific, Waltham, MA). We also assessed individuals with insulin resistance using the Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) index derived from the formula: fasting insulin (μIU/ml) � fasting glucose (mg/dL)/405 formula [24]. Impaired glucose metabolism characterization Primary outcome of impaired glucose metabolism in the study population was evaluated using three assessment criteria: FBS to determine blood sugar levels after fasting, HbA1c to estimate blood sugar levels within 2–3 months, and HOMA-IR, a surrogate parameter to assess insulin- mediated response to glucose levels. The following cut-off values to classify individuals with impaired glucose metabolism were used for each type of assessment: (1) FBS level ≥ 100 mg/dL, (2) HbA1c ≥6.5%, or (3) HOMA-IR ≥3.0 [25–28]. Statistical analysis Categorical data were expressed as count (n) and percent- age (%), and numeric variables were expressed as means and standard deviation (SD). Analysis were not stratified by site as the sites had similar demographic and social indices, patients frequently sought care outside their catchment area due to HIV-related stigma, and the sam- ple size was not large enough for stratification. Unad- justed prevalence of the primary outcome using each assessment type was summarized as a proportion with 95% confidence intervals (95% CIs). The difference by sex (biological male vs. biological female) and age group (6–9 years, 10–15 years, and 16–18 years) was assessed using the χ2 test. Means of FBS, HOMA-IR and HbA1c were compared by sex and age using the analysis of vari- ance (ANOVA), followed by pairwise comparisons between groups using Student's t-test. Separate unad- justed and adjusted linear regression models were created for FBS, HOMA-IR and HbA1c, with BMI, WHR, age and sex as predictors. Findings were summarized as esti- mated regression parameters (slopes) with 95% CIs, with the unadjusted slopes presented within the scatterplot figures and the adjusted slopes. Hypothesis tests were conducted at the two-sided alpha level of 0.05. As these were independent individual hypotheses of interest (three glucose metabolism criteria, sex and age groups) we did not adjust the alpha level (type 1 error rate) [29]. Statisti- cal analyses were performed with Prism version 9.01 (GraphPad Software) and R version 4.3.2 (R Statistical Software). RESULTS Characteristics of study participants We enrolled 400 perinatally acquired HIV-infected children and adolescents into the study, and the analysis included data from 393 participants who had complete results for all three glucose assessment parameters. Seven excluded partici- pants were not systematically different from the analytical sample with respect to demographic and anthropometric variables. About 52.25% of study participants were females. The mean age of the participants was 11.60 years (SD = 3.50), with 31.00% aged 6–9 years, 52.67% aged 10– 15 years, and 15.78% aged 16–18 years (Table 1). The age dis- tribution was comparable between females (mean age 11.43, SD = 3.5) and males (mean age 11.72, SD = 3.45) (p = 0.28) (Tables 1, S1). Almost all the participants attended school (99.25%), and over three-quarters (76.59%) lived with a bio- logical parent. As expected, there was an increasing trend in mean age, BMI and educational levels from the younger to the older age group. By contrast, a decreasing trend was observed for mean WHR from the younger to the older age group (p < 0.001) (Table 1). There were no differences in the mean values for BMI (p = 0.19) and WHR (p = 0.51) between females (BMI = 17.23 kg/m2, SD 3.54; WHR = 0.82, SD = 0.07) and males (BMI = 16.66 kg/m2, SD 3.29; WHR = 0.83; SD = 0.07) (Table S1). Impairment of glucose metabolism by age The assessment of the prevalence of glucose impairment using the three criteria showed that FBS (26.21%, 95% CI: 22.10–30.78) and HOMA-IR (22.39%, 95% CI: 18.54– 26.78) yielded a higher prevalence of glucose impairment compared with HbA1c (15.50%, 95% CI: 12.26–19.45) (Table 2). Mean FBS and HOMA-IR increased, while mean HbA1c was comparable across the age groups (Figure 1a–c).We observed an increase in the prevalence of glucose impairment using either FBS (p < 0.001) or HOMA-IR (p < 0.001) with increasing age, while glucose impairment by HbA1c (p = 0.02) was more prevalent in the two younger age groups than in those aged 16 years or older (Table 2). Impairment of glucose metabolism by sex Females had on average higher values of HOMA-IR, while mean FBS and HbA1c levels were similar by sex (Figure 1d–f). Prevalence of glucose impairment was similar between females and males when assessed either by FBS or HbA1c, but glucose impairment characterized by higher 4 AYANFUL-TORGBY ET AL. 14681293, 0, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/hiv.13614 by U niversity of G hana - A ccra, W iley O nline L ibrary on [01/02/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense HOMA-IR values was more prevalent in females than in males (Table 3). Markers associated with impaired glucose metabolism In the unadjusted analyses, our results showed that age and BMI were positively associated with FBS and HOMA-IR (p < 0.001), while WHR was negatively asso- ciated with HOMA-IR (p < 0.006) and HbA1c (p < 0.01) (Figure 2). For every 1-year increase in age, FBS increased by 0.61 (95% CI: 0.36–0.85) and HOMA-IR increased by 0.17 (95% CI: 0.10–0.25). For every incremental unit of BMI, FBS increased by 0.40 (95% CI: 0.18–0.63), and HOMA-IR increased by 0.20 (95% CI: 0.14–0.27). As WHR increased, HOMA-IR decreased by an average of 5.01 (95% CI: �8.56 to �1.48) and HbA1c TABLE 1 Demographic and anthropometric characteristics of the study participants. Variable Age groups Total (N = 393)6–9 years (n = 124) 10–15 years (n = 207) 16–18 years (n = 62) Age (years) [mean (SD)] 7.44 (1.09) 12.56 (1.71) 16.79 (0.77) 11.60 (3.50) BMI (kg/m2) [mean (SD)] 14.34 (2.22) 17.58 (3.48) 20.15 (4.36) 17.00 (3.85) WHR [mean (SD)] 0.87 (0.07) 0.82 (0.06) 0.80 (0.07) 0.83 (0.08) Sex [n (%)] Female 72 (58.10) 103 (49.80) 30 (48.40) 205 (52.25) Education [n (%)] Basic 124 (100) 196 (94.69) 32 (51.61) 352 (89.57) Secondary 0 (0.00) 10 (4.83) 29 (46.77) 39 (9.92) None 0 (0.00) 1 (0.48) 1 (1.61) 2 (0.51) Guardian [n (%)] Parenta 104 (83.87) 148 (71.49) 49 (79.03) 301 (76.59) Sibling 1 (0.81) 9 (4.34) 2 (3.22) 12 (3.05) Grandparent 9 (7.25) 20 (9.66) 2 (3.22) 31 (7.89) Other relative 6 (4.84) 15 (7.25) 3 (4.84) 24 (6.11) Othersb 4 (3.23) 14 (6.76) 6 (9.67) 24 (6.11) Abbreviations: BMI, body mass index; WHR, waist-to-hip ratio. aLiving with mother or father. bLiving alone or with a non-relative (including orphanage). TABLE 2 Prevalence of impaired glucose metabolism stratified by age groups. Variable Age groups p-value* Total (N = 393)6–9 years (n = 124) 10–15 years (n = 207) 16–18 years (n = 62) FBS Impaired [n (%)] 17 (13.71) 59 (28.50) 27 (43.55) <0.001 103 (26.21) 95% CI 8.68–20.98 22.75–35.05 31.80–56.07 22.10–30.78 HOMA-IR Impaired [n (%)] 10 (8.06) 50 (24.15) 28 (45.16) <0.0001 88 (22.39) 95% CI 4.39–14.37 18.79–30.47 33.27–57.07 18.54–26.78 HbA1c Impaired [n (%)] 20 (16.13) 36 (17.39) 5 (8.06) 0.02 61 (15.52) 95% CI 10.63–23.71 12.80–23.19 3.39–18.00 12.26–19.45 Abbreviations: CI, confidence interval; FBS, fasting blood sugar; HbA1c, glycated haemoglobin; HOMA-IR, homeostatic model assessment for insulin resistance. *χ2 test or analysis of variance (ANOVA) test. HIV MEDICINE 5 14681293, 0, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/hiv.13614 by U niversity of G hana - A ccra, W iley O nline L ibrary on [01/02/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense decreased by 4.74 (95% CI: �8.35 to �1.26). In multivari- able linear regression, age was significantly associated with FBS (slope = 0.55, 95% CI: 0.23–0.87; p < 0.001). In the adjusted model, BMI emerged as the strongest predic- tor (slope = 0.16, 95% CI: 0.08–0.24; p < 0.001) for HOMA-IR, whereas only WHR remained a predictor in the model (slope = �6.38, 95% CI: �10.48 to �2.28; p = 0.002) for HbA1c. DISCUSSION Perturbation in glucose metabolism in childhood is linked to the onset of cardiovascular diseases in adult- hood [30]. In this multicentre cross-sectional study, we assessed impaired glucose metabolism among CALWH. The study explored the prevalence of impaired glucose metabolism among the study population using three dif- ferent assessment parameters, i.e. HbA1c, FBS, and HOMA-IR. We found a high prevalence of impaired glu- cose metabolism in 15–26% of study population, depend- ing on the assessment parameter used. Overall, we found that impaired glucose metabolism is prevalent among CALWH in Ghana, suggesting that routine assessment of glucose metabolism could aid in early detection and man- agement of glucose-related conditions in this population. Using the three assessment parameters, the study observed that HbA1c levels to estimate impaired glucose metabolism yielded the lowest prevalence of 15.5% of the population. However, HbA1c prevalence was slightly higher than the other parameters in 6–9 years age group, possibly because individuals in this group consume higher amounts of carbohydrates such as sugary diets or snacks compared with other age groups [31, 32]. None- theless, there are limited available data on HbA1c levels in children and adolescents, posing a challenge in draw- ing direct comparisons with our study's findings, FIGURE 1 Boxplot of age- and sex-stratified glucose parameters (N = 393). (a–c) Age-associated fasting blood sugar (FBS) (a), homeostatic model assessment for insulin resistance (HOMA-IR) (b), and glycated haemoglobin (HbA1c) (c). (d–f) Sex-associated FBS (d), HOMA-IR (e), and HbA1c glucose parameters (f). TABLE 3 Prevalence of impaired glucose metabolism stratified by sex. Variable Female (n = 205) Male (n = 188) p-value* FBS Impaired [n (%)] 50 (24.39) 53 (28.19) 0.46 95% CI 19.00–30.72 22.23–35.02 HOMA-IR Impaired [n (%)] 53 (25.85) 35 (18.62) 0.11 95% CI 20.31–32.30 16.00–24.84 HbA1c Impaired [n (%)] 28 (13.66) 33 (17.55) 0.29 95% CI 9.59–19.09 12.74–23.60 Abbreviations: CI, confidence interval; FBS, fasting blood sugar; HbA1c, glycated haemoglobin; HOMA-IR, homeostatic model assessment for insulin resistance. *χ2 test or Student's t-test. 6 AYANFUL-TORGBY ET AL. 14681293, 0, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/hiv.13614 by U niversity of G hana - A ccra, W iley O nline L ibrary on [01/02/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense particularly among those living with HIV in sub-Saharan Africa. The few studies that have examined HbA1c have reported higher mean values, including non-obese Asian adolescents living with HIV who showed elevated HbA1c with mean values of 5.2% compared with our estimated population's mean value of 4.1% [33]. However, this Asian population was significantly older, with a mean age of 16.7 years [33]. Most studies that have evaluated HbA1c levels in children have focused primarily on mon- itoring glycaemic levels in diabetics or screening for type 2 diabetes in individuals with obesity [34]. Our study has demonstrated that CALWH in Ghana exhibit elevated HbA1c levels. We observed a prevalence of insulin resis- tance of 22% based on the HOMA index, which is slightly higher than in previous studies conducted among chil- dren in sub-Saharan Africa (20%) [35]. However, a much higher prevalence has been reported in European chil- dren (52%) and adults in sub-Saharan Africa (47%) [21, 36]. The prevalence of insulin resistance among chil- dren living with HIV varies depending on factors such as age, sex or Tanner stage. Despite being comparable in age, females had higher mean HOMA-IR values than the males in this study. Generally, HOMA-IR levels markedly increase during puberty, the onset of puberty is earlier in females, and females may be more mature than males of the same age. Studies have also reported that the preva- lence of insulin resistance tends to be higher among female adolescents than among males of the same age group [37, 38]. We observed the highest prevalence of impaired glucose metabolism using FBS, with 25% of study participants having elevated FBS levels. Wide variability in the prevalence of abnormal glucose levels by FBS have been reported in the literature [28, 39]. Vari- ables such as age, type of ART regimen and BMI have been found to influence the prevalence of abnormal glu- cose levels assessed by FBS [40]. The prevalence of FIGURE 2 Unadjusted associations of age (a), body mass index (BMI) (b), and waist-to-hip ratio (WHR) (c) with glucose metabolism parameters in study population. FBS, fasting blood sugar; HOMA-IR, homeostatic model assessment for insulin resistance; HbA1c, glycated haemoglobin A1c. HIV MEDICINE 7 14681293, 0, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/hiv.13614 by U niversity of G hana - A ccra, W iley O nline L ibrary on [01/02/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense elevated FBS levels in our study is lower than in a recent study conducted among adolescents in Tanzania (46%). This could be due to the higher levels of overweight or obese individuals in the Tanzania study [7]. In our study cohort, central obesity, as defined by waist circumference to height ratio > 0.5 [41], was a rare occurrence among our participants and not a risk factor for impaired glucose metabolism. Instead, we observed abnormal waist circumference, as determined by elevated waist-to-hip ratio ≥1 [42], to be more indicative of a potential risk factor for impaired glucose metabolism in our cohort. Participants with abnormal waist circumfer- ence were predominately below 10 years old, yet charac- terized by low BMI, and were referred for nutritional assessment. Consistent with findings from others, the BMI in our study participants was comparable to other CALWH in different regions [18, 43]. BMI tends to increase substantially with age and is more pronounced from early adolescence to adulthood. Despite the cohort's relatively low BMI, a positive trend with glucose values was observed. It is worth noting that most of these chil- dren reside in economically disadvantaged households and are more prone to experience nutritional inadequa- cies with resultant stunted growth [44]. Thus, unsurpris- ingly, the rate of obesity among CALWH in sub-Saharan Africa is low. Thus, we found that relying solely on cen- tral or general obesity as a criterion for monitoring impaired glucose metabolism can underestimate the prevalence of impaired glucose among CALWH or lean children and adolescents in the general population. Taken together, we observed a wide variability in the proportion of individuals with impaired glucose metabo- lism in the study population using the three assessment criteria. While the proportion of impaired FBS aligns with insulin resistance, as HOMA-IR is derived from FBS, there is a lower prevalence of individuals with ele- vated HbA1c. This discrepancy can be attributed to sev- eral factors, including wide variations in HbA1c values corresponding to FBS levels, as HbA1c 6.0–7.0% can be equivalent to 100–185 mg/dL mean FBS [45]. Addition- ally, the presence of hemoglobinopathies, anaemia and nutritional deficiencies within the study population, as highlighted by Wegmüller et al. [46], can affect the hae- moglobin glycation gap and index. These factors may lead to clinical inaccuracies in HbA1c values and contrib- ute to the discordance observed between the proportion of individuals with high FBS levels but normal HbA1c. Considering prevailing haemoglobin disorders and nutri- tional deficiencies in our population, it becomes evident that HbA1c may not be the preferred parameter for asses- sing impaired glucose metabolism in this specific context. When considering the sensitivity of assessment parame- ters, FBS provided the highest prevalence, which makes it potentially the most sensitive tool. Previous studies have also reported FBS values as having the best predic- tive power to identify impaired glucose metabolism in children and adolescents, particularly in healthy or non- obese populations [47]. Although the prevalence of glu- cose impairment by FBS and HOMA-IR were compara- ble, factors such as accessibility and robustness of FBS are more practical to a setting in the sub-Sahara African region. Therefore, using FBS as a parameter for assessing impaired glucose metabolism not only offers high sensi- tivity but is also more practical in resource-constrained settings. Evaluating glucose levels using the FBS parame- ter may be easier, faster, more affordable and acceptable to individuals due to its less invasive sampling method and lower requirement of blood volume [48]. Our study has several limitations, which need to be considered in the interpretation of the findings. First we relied solely on age as grouping variable instead of also assessing puberty, as age may not fully capture the varia- tions in developmental maturity levels among individ- uals. Nonetheless, while age may not be a perfect proxy for developmental maturity, it is a solid proxy for cumu- lative exposure to ART among CALWH with perinatal HIV infection, and we observed associations of age with FBS and HOMA-IR levels. Second, the cross-sectional nature of our study only allowed us to examine the preva- lence of glucose impairment at one point in time, which could affect the estimates. The dynamic changes in glu- cose parameters with respect to age and sex should be further examined in a longitudinal study. However, our findings are consistent with previous research indicating high frequency of impaired glucose metabolism in CALWH, which may serve as a proxy for metabolic dis- eases in adulthood [17, 49, 50]. CONCLUSION In summary, this cross-sectional study reveals a high prevalence of impaired glucose metabolism in CALWH. This may pose a potential risk for an increase in cardio- metabolic diseases within perinatally infected HIV patients residing in sub-Saharan Africa. Yet, questions remain as to whether the observed impairment in glucose metabolism is transient or permanent. Could it be critical in the causal pathways of metabolic syndrome in early adulthood? Prospective monitoring is crucial to answer these questions, to gain better understanding of the risk predictors for cardiometabolic diseases in individuals living with HIV. Our findings also underscore the importance of the type of assessment used to identify impaired glucose metabolism among children and adoles- cents. Considering the clinical diagnostic effectiveness, 8 AYANFUL-TORGBY ET AL. 14681293, 0, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/hiv.13614 by U niversity of G hana - A ccra, W iley O nline L ibrary on [01/02/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense integrating simple and readily available FBS measure- ments may aid in predicting early glucose abnormalities, such as hyperglycaemia, and facilitate management. AUTHOR CONTRIBUTIONS Conception and design of experiments: EP, RAT, LEA, JKAT. Data collection: RAT, JKAT, AAE, FA, YA-M. Per- formed experiments: RAT, AAE, EB, JA. Analysed data: RAT, VS. Original draft: RAT, VS. ACKNOWLEDGEMENTS Our gratitude goes to the study participants and their par- ents or guardians, the staff of ART clinics in the study facilities and the Ghana National AIDS control pro- gramme. This study was supported by Fogarty Interna- tional Center and the National Institute of Alcohol Abuse and Alcoholism of the National Institutes of Health under award number D43 TW011526. CONFLICTS OF INTEREST The authors have no conflicts of interest. ORCID Ruth Ayanful-Torgby https://orcid.org/0000-0003-1066- 3345 Frank Aboagye https://orcid.org/0000-0002-3881-9789 Yusuf Al-Mahroof https://orcid.org/0009-0005-3803- 3013 Elijah Paintsil https://orcid.org/0000-0002-8283-4564 REFERENCES 1. Feinstein L, Edmonds A, Chalachala JL, et al. Temporal changes in the outcomes of HIV-exposed infants in Kinshasa, Democratic Republic of Congo during a period of rapidly evolv- ing guidelines for care (2007–2013). Aids. 2014;28(Suppl 3): S301-S311. 2. Anuurad E, Semrad A, Berglund L. Human immunodeficiency virus and highly active antiretroviral therapy–associated meta- bolic disorders and risk factors for cardiovascular disease. Metab Syndr Relat Disord. 2009;7(5):401-409. 3. Jeric�o C, Knobel H, Montero M, et al. Metabolic syndrome among HIV-infected patients. Diabetes Care. 2005;28(1): 132-137. 4. Maganga E, Smart LR, Kalluvya S, et al. Glucose metabolism disorders, HIV and antiretroviral therapy among Tanzanian adults. PLOS One. 2015;10(8):e0134410. 5. Satlin MJ, Hoover DR, Glesby MJ. Glycemic control in HIV- infected patients with diabetes mellitus and rates of meeting American Diabetes Association management guidelines. AIDS Patient Care STDS. 2011;25(1):5-12. 6. Ryom L, Lundgren JD, El-Sadr W, et al. Cardiovascular disease and use of contemporary protease inhibitors: the D:a:D inter- national prospective multicohort study. Lancet HIV. 2018;5(6): e291-e300. 7. Nkinda L, Buberwa E, Memiah P, et al. Impaired fasting glu- cose levels among perinatally HIV-infected adolescents and youths in Dar es Salaam, Tanzania. Front Endocrinol. 2022;13: 1045628. 8. Reng SR, Uloko EA, Puepet HF, Onwuegbuzie AG, Ramalan AM. Prevalence and determinants of glucose intoler- ance among HIV/AIDS patients in north-central Nigeria. Niger J Med J Natl Assoc Resid Dr Niger. 2016;25(2):128-132. 9. Innes S, Abdullah KL, Haubrich R, Cotton MF, Browne SH. High prevalence of dyslipidemia and insulin resistance in HIV- infected Prepubertal African children on antiretroviral therapy. Pediatr Infect Dis J. 2016;35(1):e1-e7. 10. Gast KB, Tjeerdema N, Stijnen T, Smit JWA, Dekkers OM. Insulin resistance and risk of incident cardiovascular events in adults without diabetes: meta-analysis. PLoS ONE. 2012;7(12): e52036. 11. Yajnik CS, Katre PA, Joshi SM, et al. Higher glucose, insulin and insulin resistance (HOMA-IR) in childhood predict adverse cardiovascular risk in early adulthood: the Pune Children's study. Diabetologia. 2015;58(7):1626-1636. 12. Assunção SNFD, Boa Sorte NCA, Alves CDAD, Mendes PSA, Alves CRB, Silva LR. Glucose alteration and insulin resistance in asymptomatic obese children and adolescents. J Pediatr (Rio J). 2018;94(3):268-272. 13. Al-Daghri NM, Al-Attas OS, Alokail M, et al. Does visceral adi- posity index signify early metabolic risk in children and adoles- cents?: association with insulin resistance, adipokines, and subclinical inflammation. Pediatr Res. 2014;75(3):459-463. 14. Cruz ML, Goran MI. The metabolic syndrome in children and adolescents. Curr Diab Rep. 2004;4(1):53-62. 15. Reisi M, Keivanfar M, Mostofizadeh N, Madihi Y, Esmaeilian Y, Hashemi E. Correlation between stable hyper- glycemia and mortality in children admitted to the pediatric intensive care unit of imam Hossein hospital. Adv Biomed Res. 2021;10(1):2. 16. Rivera DM, Frye RE. Pediatric HIV Infection. Medscape Drug Reference [Internet]. 2020 Available from: https://emedicine. medscape.com/article/965086-overview#a1 17. Paganella MP, Cohen RA, Harris DR, et al. Association of Dys- lipidemia and Glucose Abnormalities with Antiretroviral Treat- ment in a cohort of HIV-infected Latin American children. JAIDS J Acquir Immune Defic Syndr. 2017;74(1):e1-e8. 18. Salako AO, Gbaja-Biamila TA, Ezemelue PN, et al. Metabolic syndrome among adolescents and young adults living with HIV in Lagos: a cross-sectional study. Glob Pediatr. 2021;1:100001. 19. Das RR, Mangaraj M, Panigrahi SK, Satapathy AK, Mahapatro S, Ray PS. Metabolic syndrome and insulin resis- tance in schoolchildren from a developing country. Front Nutr. 2020;7:31. 20. Chantry CJ, Hughes MD, Alvero C, et al. Insulin-like growth Factor-1 and lean body mass in HIV-infected children. JAIDS J Acquir Immune Defic Syndr. 2008;48(4):437-443. 21. Rosso R, Parodi A, d'Annunzio G, et al. Evaluation of insulin resistance in a cohort of HIV-infected youth. Eur J Endocrinol. 2007;157(5):655-659. 22. Ghana demographics and health survey. Ghana demographics and health survey, Final Report. GSS and GHS. 2014 https:// dhsprogram.com/methodology/survey/survey-display-437.cfm 23. National AIDS/STI Control Programme. Ghana Health Service. Consolidated Guidelines for HIV Care in Ghana Test, Treat & Track. 2019 https://ghs.gov.gh/wp-content/uploads/2022/10/ Consolidated-HIV-Guidelines.pdf HIV MEDICINE 9 14681293, 0, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/hiv.13614 by U niversity of G hana - A ccra, W iley O nline L ibrary on [01/02/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://orcid.org/0000-0003-1066-3345 https://orcid.org/0000-0003-1066-3345 https://orcid.org/0000-0003-1066-3345 https://orcid.org/0000-0002-3881-9789 https://orcid.org/0000-0002-3881-9789 https://orcid.org/0009-0005-3803-3013 https://orcid.org/0009-0005-3803-3013 https://orcid.org/0009-0005-3803-3013 https://orcid.org/0000-0002-8283-4564 https://orcid.org/0000-0002-8283-4564 https://emedicine.medscape.com/article/965086-overview#a1 https://emedicine.medscape.com/article/965086-overview#a1 https://dhsprogram.com/methodology/survey/survey-display-437.cfm https://dhsprogram.com/methodology/survey/survey-display-437.cfm https://ghs.gov.gh/wp-content/uploads/2022/10/Consolidated-HIV-Guidelines.pdf https://ghs.gov.gh/wp-content/uploads/2022/10/Consolidated-HIV-Guidelines.pdf 24. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insu- lin resistance and?-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7): 412-419. 25. World Health Organization. Definition and diagnosis of diabe- tes mellitus and intermediate hyperglycemia: report of a WHO/IDF consultation. Production. 2006 https://apps.who. int/iris/handle/10665/43588 26. Gillett MJ. International expert committee report on the role of the A1c assay in the diagnosis of diabetes: diabetes. Clin Bio- chem Rev. 2009;30(4):197-200. 27. Lee JM, Gebremariam A, Wu EL, LaRose J, Gurney JG. Evalu- ation of nonfasting tests to screen for childhood and adolescent Dysglycemia. Diabetes Care. 2011;34(12):2597-2602. 28. Center for Disease Control and Prevention. Diabetes Tests. https://www.cdc.gov/diabetes/basics/getting-tested.html 29. Rubin M. When to Adjust Alpha during Multiple Testing: A Consideration of Disjunction, Conjunction, and Individual Testing. 2023 https://arxiv.org/abs/2107.02947 30. Tricò D, Natali A, Arslanian S, Mari A, Ferrannini E. Identifi- cation, pathophysiology, and clinical implications of primary insulin hypersecretion in nondiabetic adults and adolescents. JCI Insight. 2018;3(24):e124912. 31. Ogum Alangea D, Aryeetey RN, Gray HL, Laar AK, Adanu RMK. Dietary patterns and associated risk factors among school age children in urban Ghana. BMC Nutr. 2018; 4(1):22. 32. Wang J, Light K, Henderson M, et al. Consumption of added sugars from liquid but not solid sources predicts impaired glu- cose homeostasis and insulin resistance among youth at risk of obesity. J Nutr. 2014;144(1):81-86. 33. Santiprabhob J, Tanchaweng S, Maturapat S, et al. Metabolic disorders in HIV-infected adolescents receiving protease inhibi- tors. Biomed Res Int. 2017;2017:1-14. 34. Vijayakumar P, Nelson RG, Hanson RL, Knowler WC, Sinha M. HbA1c and the prediction of type 2 diabetes in chil- dren and adults. Diabetes Care. 2017;40(1):16-21. 35. Ketha F, Beatriz P, Sandra T, et al. Insulin resistance and glu- cose intolerance in HIV infected children on antiretroviral therapy at Lubango pediatric hospital - Angola. Int J Virol AIDS. 2020;7:3. 36. Noumegni SRN, Nansseu JR, Ama VJM, et al. Insulin resis- tance and associated factors among HIV-infected patients in sub-Saharan Africa: a cross sectional study from Cameroon. Lipids Health Dis. 2017;16(1):148. 37. Geffner ME, Patel K, Miller TL, et al. Factors associated with insulin resistance among children and adolescents perinatally infected with HIV-1 in the pediatric HIV/AIDS cohort study. Horm Res Paediatr. 2011;76(6):386-391. 38. Geffner ME, Patel K, Jacobson DL, et al. Changes in insulin sensitivity over time and associated factors in HIV-infected adolescents. Aids. 2018;32(5):613-622. 39. Taheri F, Bijari B, Kazemi T, Namakin K, Zardast M, Chahkandi T. Prevalence of high normal FBS and prediabetes among adolescents in Birjand, east of Iran, 2012. J Educ Health Promot. 2015;4(1):68. 40. Levitt NS, Peer N, Steyn K, et al. Increased risk of dysglycaemia in south Africans with HIV; especially those on protease inhibi- tors. Diabetes Res Clin Pract. 2016;119:41-47. 41. McCarthy HD, Ashwell M. A study of central fatness using waist-to-height ratios in UK children and adolescents over two decades supports the simple message–‘keep your waist circum- ference to less than half your height’. Int J Obes (Lond). 2006; 30(6):988-992. 42. Kułaga Z, Świąder-Le�sniak A, Kotowska A, Litwin M. Popula- tion-based references for waist and hip circumferences, waist- to-hip and waist-to-height ratios for children and adolescents, and evaluation of their predictive ability. Eur J Pediatr. 2023; 182(7):3217-3229. 43. Tzur F, Chowers M, Agmon-Levin N, Mekori YA, Hershko AY. Increased prevalence of diabetes mellitus in a non-obese adult population: HIV-infected Ethiopians. Isr Med Assoc J IMAJ. 2015;17(10):620-623. 44. Fox AM. The HIV–poverty thesis Re-examined: poverty, wealth or inequality as a social determinant of HIV infection in sub- Saharan Africa? J Biosoc Sci. 2012;44(4):459-480. 45. Nathan DM, Kuenen J, Borg R, et al. Translating the A1C assay into estimated average glucose values. Diabetes Care. 2008; 31(8):1473-1478. 46. Wegmüller R, Bentil H, Wirth JP, et al. Anemia, micronutrient deficiencies, malaria, hemoglobinopathies and malnutrition in young children and non-pregnant women in Ghana: findings from a national survey. PLOS One. 2020;15(1):e0228258. 47. Guerrero-Romero F, Violante R, Rodríguez-Mor�an M. Distribu- tion of fasting plasma glucose and prevalence of impaired fast- ing glucose, impaired glucose tolerance and type 2 diabetes in the Mexican paediatric population. Paediatr Perinat Epidemiol. 2009;23(4):363-369. 48. Sinha R, Fisch G, Teague B, et al. Prevalence of impaired glu- cose tolerance among children and adolescents with marked obesity. N Engl J Med. 2002;346(11):802-810. 49. Dimock D, Thomas V, Cushing A, et al. Longitudinal assess- ment of metabolic abnormalities in adolescents and young adults with HIV-infection acquired perinatally or in early childhood. Metabolism. 2011;60(6):874-880. 50. Frigati LJ, Jao J, Mahtab S, et al. Insulin resistance in south Afri- can youth living with perinatally acquired HIV receiving antire- troviral therapy. AIDS Res Hum Retroviruses. 2019;35(1):56-62. SUPPORTING INFORMATION Additional supporting information can be found online in the Supporting Information section at the end of this article. How to cite this article: Ayanful-Torgby R, Shabanova V, Essuman AA, et al. High prevalence of impaired glucose metabolism among children and adolescents living with HIV in Ghana. HIV Med. 2024;1‐10. doi:10.1111/hiv.13614 10 AYANFUL-TORGBY ET AL. 14681293, 0, D ow nloaded from https://onlinelibrary.w iley.com /doi/10.1111/hiv.13614 by U niversity of G hana - A ccra, W iley O nline L ibrary on [01/02/2024]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://apps.who.int/iris/handle/10665/43588 https://apps.who.int/iris/handle/10665/43588 https://www.cdc.gov/diabetes/basics/getting-tested.html https://arxiv.org/abs/2107.02947 info:doi/10.1111/hiv.13614 High prevalence of impaired glucose metabolism among children and adolescents living with HIV in Ghana INTRODUCTION METHODS Study population Study ethical approvals Data collection Anthropometric parameters Biochemical parameters Impaired glucose metabolism characterization Statistical analysis RESULTS Characteristics of study participants Impairment of glucose metabolism by age Impairment of glucose metabolism by sex Markers associated with impaired glucose metabolism DISCUSSION CONCLUSION AUTHOR CONTRIBUTIONS ACKNOWLEDGEMENTS CONFLICTS OF INTEREST REFERENCES