Abstract
Introduction
Sub-Saharan African migrant populations are exposed to new environmental factors, both of which have been linked to increased rates of obesity and insulin resistance. Given the complex relationship between adipokines and cardiometabolic traits, we hypothesized that these associations may vary depending on geographical context. Our aim was to examine the influence of geographic location on the association between serum leptin and adiponectin and cardiometabolic traits.
Methods
This is a cross-sectional analysis among 2,640 participants from the RODAM study living in Amsterdam, the Netherlands, and urban and rural Ghana. Correlation and linear regression models were used to examine the relationship between adipokines and cardiometabolic traits, including body mass, insulin resistance, inflammation, and lipid metabolism per location.
Results
Body mass was the key determinant of serum leptin, less so for serum adiponectin. There was a significant (p < 0.001) interaction in the association between BMI and serum leptin according to geographic location in women and in the association between waist circumference and serum leptin in men, suggesting increased leptin resistance during rural to urban transition, but with similar slopes for urban Ghanaians living in tropical and temperate climates. There was no significant interaction with location in the association between body mass and adiponectin. Inflammation and lipid metabolism explained the least amount of variance in serum adipokines across the locations.
Conclusion
There was significant variability in the relationship between serum leptin and the cardiometabolic traits examined across locations. These findings suggest that rural to urban transition significantly affects this relationship. Future studies may help to further delineate the effects of environmental factors on adipokine production, obesity, and cardiometabolic disease.
Keywords: Leptin, Leptin resistance, Adiponectin, Body mass index, Waist circumference, Geographic location, RODAM
Introduction
There is a world-wide epidemic of obesity and obesity-related diseases and a need to better understand the underlying biology to curb this epidemic [1]. Recent advances in the underlying biology have shown the potential of peptides such as Ozempic (semaglutide) in the treatment of obesity and obesity-related diseases [2]. Similarly, leptin and adiponectin, collectively called adipokines/adipocytokines, are peptide hormones produced by adipose tissue that have potential as therapeutic targets or markers of disease progression because of their suspected physiological role in obesity-related diseases [3]. Both leptin and adiponectin play a key role in the pathophysiology of obesity-related diseases [4, 5] and have been associated with variables of body mass (such as body mass index [BMI]), chronic inflammation, insulin resistance, and triglyceride/lipid metabolism [6–8]. The strength of the association between the adipokines and body mass, insulin resistance, and inflammation differs, with leptin strongly associated with body mass and to a lesser extent to insulin resistance, whilst the reverse is true for adiponectin [9, 10]. Most studies on these adipokines have been conducted in vivo and in vitro and there is a need to understand their association with cardiometabolic traits in diverse human populations under different environmental factors.
Leptin resistance, prevalent in people living with obesity and associated with obesity-related disease, is a state in which the body’s normal response to the hormone leptin is blunted, despite high circulating levels of leptin [11]. Leptin resistance in experimental models have been well documented and has been demonstrated by showing lower leptin levels in diet-induced obese mice with an anti-leptin antibody improving energy homeostasis [12, 13]. It is hypothesized that high leptin levels may lead to leptin resistance, similar to what is observed with insulin resistance. Additionally, environmental factors including ambient temperature may influence leptin responsiveness and upregulate the number of leptin receptors [14, 15], next to high triglyceride levels, insulin resistance, and inflammation [16–19]. More recently, it has been postulated that environmental factors may explain part of the known differences in circulating leptin and adiponectin levels between ethnic groups [20, 21]. However, the effect of environmental factors on adipokines goes beyond the observed ethnic differences warranting the need to evaluate the impact of environmental factors among a homogenous population living in different environmental contexts. Further, migration and urbanization are known key drivers of changes in exposure to environmental factors.
Various studies, including our own, have shown an association between urbanization and migration and the serum leptin and adiponectin levels of individuals of a genetically homogenous population [22, 23] possibly mediated by the difference in the prevalence of obesity and the environmental factors in the different locations. However, limited data exist on the geographical differences in leptin resistance and in the association between cardiometabolic traits and these adipokines. In this present study, we aimed to examine the influence of geographical context on leptin resistance and the association between serum leptin and adiponectin with cardiometabolic traits among a relatively homogenous sub-Saharan African population.
Materials and Methods
Study Design, Study Participants, and Ethical Considerations
We performed a cross-sectional analysis on baseline data from the multi-centre Research on Obesity and Diabetes among African Migrants (RODAM). To summarize, the RODAM study is a population-based study that collected baseline data from 2012 to 2015 among Ghanaians (aged 18–96 years) living in rural and urban Ghana as well as in Amsterdam, the Netherlands; Berlin, Germany; and London, UK. The majority of the Ghanaian migrants belong to the Akan ethnic group, the largest ethnic group in Ghana. To ensure homogeneity, participants from rural and urban Ghana were recruited from Kumasi and its environs where majority of Akans reside. For further details, the rationale, conceptual framework, design, and methods of the RODAM study have been extensively described elsewhere [24]. Data collection and handling, processing, storage of samples in the study was standardized across sites. This cross-sectional analysis was carried out on a subset of the total RODAM study population and included a random selection of 2,640 participants from Amsterdam (1,189) and rural (906) and urban (545) Ghana with available data on serum leptin and adiponectin concentrations. A detailed overview of participants included in this cross-sectional analysis is illustrated in Figure 1. The approval for the study protocol was granted at all sites from the respective Ethical Committees: in Ghana (School of Medical Sciences/Komfo Anokye Teaching Hospital Committee on Human Research, Publication and Ethical Review Board) and in the Netherlands (Institutional Review Board of the Academic Medical Centre, University of Amsterdam), before data collection began in each country. Written informed consent was obtained from each participant before enrolment. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline [25].
Fig. 1.
Flow chart of participants included for analysis.
Measurement of Variables and Definitions
Information on demographics, medical history, and lifestyle factors was obtained through a structured questionnaire using appropriately validated instruments. Physical examination was done with validated devices according to standardized operational procedures across all study sites. Weight to the nearest 0.1 kg and height to the nearest 0.1 cm were measured in light clothing without shoes using SECA 877 weighing scales and SECA 217 portable stadiometers, respectively (manufactured by SECA, Hamburg, Germany). BMI (kg/m2) was calculated by dividing the weight in kilograms by the squared height in metres. Waist circumference (WC) (cm) and hip circumference (cm) were measured to the nearest millimetre using a standard measuring tape. For WC, the measurement was taken at the midpoint between the lower rib and the upper margin of the iliac crest, and hip circumference was measured around the major trochanter. Waist-to-hip ratio (WHR) was calculated by dividing WC by hip circumference. All anthropometric measurements were performed twice, and the average of the two measurements was used for the analysis.
Trained research assistants took fasting (overnight fast of 10–14 h) venous blood samples at all locations and samples were handled, processed, and stored according to standardized procedures. All samples were processed and divided into aliquots immediately after collection (within 1 h to a maximum of 3 h of the vena puncture) and then temporarily stored at the local research location at −20°C. The separated samples were then transported to the local research centres’ laboratories, where they were checked and registered and stored at − 80°C, before being shipped to their final destination for analysis. All samples were taken to Berlin, Germany, for analysis to avoid inter-laboratory bias. After the initial analysis, samples were then randomly selected from the Amsterdam and Ghanaian participants and sent to the Endocrinology Research Laboratory at the Johns Hopkins University School of Medicine, to further analyse for adiponectin and leptin concentrations. Adiponectin and leptin were measured in duplicate samples using Enzyme-Linked Immunosorbent Assay (ELISA) kits (Crystal Chem, Elk Grove, IL, USA): adiponectin, Cat No. 80571 (RRID:AB_2800326), sensitivity 0.3 ng/mL, intra-assay coefficient of variability (CV) 3.8%, inter-assay CV 5.7%; leptin, Cat No. 80968, sensitivity 0.42 ng/mL, intra-assay CV 4.5%, inter-assay CV 4.7%.
Measurement of fasting plasma glucose (FPG) concentration was obtained using an enzymatic method (hexokinase method by colourimetry). The Mercodia ELISA kit (Mercodia, Uppsala, Sweden) was used to measure plasma insulin levels. Based on the fasting plasma glucose and the plasma insulin, insulin resistance was assessed using the Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) method with early insulin resistance and overt insulin resistance defined as an HOMA-IR value of ≥1.9 and ≥2.9, respectively [26]. The glycated haemoglobin (HbA1C) was measured by high-performance liquid chromatography (TOSOH G8 HPLC analyser). The high-sensitivity C-reactive protein (hs-CRP) concentration, as a marker of inflammation, was measured in heparin plasma by a particle-enhanced immunoturbidimetric assay. Fasting total serum cholesterol, triglycerides, HDL-C, and LDL-C, as markers of lipid metabolism, were measured using enzymatic colourimetric assays (ABX Pentra 400 chemistry analyser, HORIBA ABX, Germany). Associations between measures of insulin resistance and dyslipidaemia with adiponectin and leptin were evaluated based on insights obtained from the HOORN study and a previous analysis in RODAM on differences between leptin and adiponectin that focused on geographical differences [10, 23].
Data Analysis
The data were analysed using RStudio version 2021.09.1 Build 372 for Mac Os and IBM SPSS version 26 for Windows. All analyses were stratified by sex because of the well-established differences in adiponectin and leptin levels between women and men [27, 28]. Data on continuous variables were presented as mean with standard deviation for normally distributed data, with p values for group differences determined using unpaired Student’s t test, and as median with interquartile range for non-normally distributed data with p values determined using Mann-Whitney U test. Normality was determined graphically with histograms and QQ plots. Data on categorical variables were presented as frequencies and percentages (proportions) with p values determined using a chi-square test. In addition, the distribution of serum leptin and adiponectin was normalized by natural log transformation to obtain a normal distribution. The bivariate analysis was done per sex, with the relationship between the log-transformed adipokines and each categorized cardiometabolic trait (variables of body mass [BMI, WC, WHR], insulin resistance [HOMA-IR, FPG, HbA1C], inflammation [hs-CRP], and lipid metabolism [HDL and LDL]) presented as correlation plots with Pearson’s correlation coefficients or Spearman’s rho to depict the strength of the correlation, where appropriate. Correlation coefficients were compared between geographic location using Fischer’s r-to-z transformation. Subsequently, after stratifying per location, the association between serum leptin and adiponectin levels and each categorized cardiometabolic trait was explored whilst adjusting for age. For this multivariable linear regression analysis, the analysis was done per sex, with each cardiometabolic trait category adjusted for age serving as a model. As the distribution of serum leptin and adiponectin was normalized by natural log transformation to obtain a normal distribution, back transformation of the beta values obtained was done to depict a unit change in the independent variable associated with a percentage change in the dependent variable. Results were presented as back-transformed beta values with confidence intervals, p values, and adjusted R2 (for multivariable analysis). Adjusted R2 for each model was expressed as the percentage of explained variance in the dependent variable. Interaction analyses were done to evaluate whether the relationship between the adipokines and variables of body composition, variables of insulin resistance, and variables of inflammation was modified by location/environmental factors. Missing values were handled by listwise deletion, with participants with complete data deemed representative of the total sample following subset analysis.
Results
Population Baseline Characteristics
Baseline characteristics stratified by sex and geographical location are shown in Table 1. The mean age was higher among Ghanaian women living in rural Ghana (48.5 years [95% CI: 47.2–49.8]) compared to their counterparts living in urban Ghana (45.6 years [95% CI: 44.5–46.7]) and Amsterdam (44.10 [95% CI: 43.34–44.85]), with no age differences observed amongst the men per location. In general, rural Ghanaians had a better cardiometabolic risk profile: a lower mean BMI, WC, HbA1C, serum total cholesterol, and serum LDL cholesterol compared to urban Ghanaians and Ghanaians living in Amsterdam. Women and men living in rural Ghana had a more favourable adipokine profile compared to the other geographic locations, with the highest median serum adiponectin (9.9 μg/mL and 6.9 μg/mL, respectively [p < 0.001, comparison between locations]) and the lowest median serum leptin concentrations (10.32 ng/mL and 0.28 ng/mL, respectively [p < 0.001]).
Table 1.
Baseline characteristics of the included population
| | Amsterdam Ghanaians | Urban Ghanaians | Rural Ghanaians | p value |
|---|---|---|---|---|
| Women | ||||
| Site, n (%) | 710 (43.0) | 385 (23.3) | 557 (33.7) | |
| Age, mean (95% CI), years | 44.10 (43.34–44.85) | 45.57 (44.45–46.69) | 48.51 (47.18–49.83) | <0.001 |
| Serum adipokines | ||||
| Serum leptin, median (IQR), ng/mL | 28.28 (16.16–41.64) | 25.82 (12.09–42.42) | 10.32 (4.10–23.44) | <0.001 |
| Serum adiponectin, median (IQR), μg/mL | 8.51 (5.23–12.95) | 6.21 (3.89–9.67) | 9.90 (6.26–15.62) | <0.001 |
| Variables related to body mass | ||||
| BMI, mean (95% CI), kg/m2 | 29.72 (29.33–30.10) | 28.01 (27.45–28.56) | 23.49 (23.05–23.93) | <0.001 |
| WC, mean (95% CI) | 94.09 (93.17–95.01) | 90.60 (89.41–91.78) | 83.54 (82.46–84.62) | <0.001 |
| WHR, mean (95% CI) | 0.89 (0.88–0.89) | 0.90 (0.89–0.91) | 0.90 (0.89–0.90) | <0.001 |
| Variables related to insulin resistance | ||||
| FPG, mmol/L, mean (95% CI), mmol/L | 5.33 (5.25–5.42) | 5.66 (5.44–5.89) | 5.21 (5.03–5.38) | <0.001 |
| HbA1C, mean (95% CI), mmol/mol | 39.16 (38.52–39.79) | 39.34 (37.80–40.88) | 32.00 (31.04–32.96) | <0.001 |
| Calculated insulin resistance (HOMA-IR), median (IQR) | 1.58 (1.07–2.30) | 1.73 (1.07–2.88) | 1.07 (0.70–1.72) | <0.001 |
| Variables related to inflammation | ||||
| CRP, median (IQR), mg/L | 0.90 (0.30–3.00) | 1.00 (0.30–3.30) | 0.90 (0.20–3.30) | 0.675 |
| Variables related to lipid metabolism | ||||
| Triglycerides, mean (95% CI), mmol/L | 0.79 (0.77–0.82) | 1.14 (1.08–1.20) | 1.14 (1.09–1.20) | <0.001 |
| HDL-C, mean (95% CI), mmol/L | 1.45 (1.43–1.48) | 1.28 (1.25–1.31) | 1.19 (1.16–1.23) | <0.001 |
| LDL-C, mean (95% CI), mmol/L | 3.22 (3.15–3.29) | 3.41 (3.32–3.51) | 3.00 (2.91–3.09) | <0.001 |
| Men | ||||
| Site, n (%) | 479 (48.5) | 160 (16.2) | 349 (35.3) | |
| Age, mean (95% CI), years | 47.49 (46.49–48.49) | 47.16 (45.28–49.04) | 49.37 (47.57–51.17) | 0.505 |
| Serum adipokines | ||||
| Serum leptin, median (IQR), ng/mL | 6.01 (2.58–11.45) | 6.73 (1.50–14.25) | 0.28 (0.11–1.19) | <0.001 |
| Serum adiponectin, median (IQR), μg/mL | 4.61 (2.68–8.17) | 5.17 (3.47–8.45) | 6.91 (4.03–11.26) | <0.001 |
| Variables related to body mass | ||||
| BMI, mean (95% CI), kg/m2 | 27.02 (26.67–27.37) | 24.18 (23.60–24.77) | 20.78 (20.42–21.13) | <0.001 |
| WC, mean (95% CI) | 92.88 (91.88–93.88) | 84.66 (83.09–86.23) | 76.83 (75.84–77.82) | <0.001 |
| WHR, mean (95% CI) | 0.94 (0.93–0.95) | 0.91 (0.90–0.92) | 0.87 (0.88–0.89) | <0.001 |
| Variables related to insulin resistance | ||||
| FPG, mean (95% CI), mmol/L | 5.75 (5.51–5.98) | 5.48 (5.24–5.73) | 5.18 (5.00–5.36) | <0.001 |
| HbA1C, mean (95% CI), mmol/mol | 40.22 (39.24–41.20) | 37.17 (34.86–39.49) | 30.93 (29.66–32.20) | <0.001 |
| Calculated insulin resistance (HOMA-IR), median (IQR) | 1.49 (0.95–2.31) | 1.27 (0.68–1.80) | 0.65 (0.40–0.98) | <0.001 |
| Variables related to inflammation | ||||
| CRP, median (IQR), mg/L | 0.50 (0.20–1.40) | 0.50 (0.10–1.80) | 0.70 (0.10–2.10) | 0.325 |
| Variables related to lipid metabolism | ||||
| Triglycerides, mean (95% CI), mmol/L | 1.03 (0.97–1.09) | 1.18 (1.09–1.27) | 1.06 (1.00–1.13) | 0.013 |
| HDL-C, mean (95% CI), mmol/L | 1.32 (1.29–1.35) | 1.18 (1.13–1.23) | 1.20 (1.15–1.25) | <0.001 |
| LDL-C, mean (95% CI), mmol/L | 3.29 (3.20–3.37) | 3.40 (3.24–3.56) | 2.53 (2.44–2.64) | <0.001 |
n = 1,652 [women] and n = 988 [men]. Values are presented as means (SD or 95% CI), medians (IQR), n (%), or percentages (%) per site. Hypertension defined as systolic BP ≥140 mm Hg and/or diastolic BP ≥90 mm Hg and/or current use of antihypertensive agents.
CI, confidence interval; IQR, interquartile range; BMI, body mass index; HOMA-IR, Homeostatic Model Assessment for Insulin Resistance.
Relationship between Adipokines and Cardiometabolic Traits per Geographical Location
Interaction analysis revealed a statistically significant influence of geographical location on the association between measures of body mass (BMI and WC) and serum leptin (p < 0.001) for both women and men, with the relationship between BMI and serum leptin in women affected by location, less so for men, while the same was true for relationship between WC and serum leptin in men (online suppl. Tables 1, 2; for all online suppl. material, see https://doi.org/10.1159/000550389). The regression lines for serum leptin with BMI and WC were steeper for both men and women in rural Ghana compared to those living in urban Ghana and Amsterdam (Fig. 2, 3a, b), whereas no statistically significant differences were observed in the association with BMI and WC and serum leptin according to latitude (online suppl. Tables 3, 4). There was no significant differences in the association with measures of insulin resistance, lipid metabolism and inflammation, and serum leptin per geographic location (data not shown). Additionally, there was a significant influence of geographical location in the association between variables related to body mass and serum adiponectin in men but not for women (online suppl. Tables 5, 6), while there was no significant influence of geographical location on the association between variables of insulin resistance, lipid metabolism and inflammation, and the adiponectin (data not shown).
Fig. 2.
a, b Correlation between serum leptin and body mass index (BMI) in women (a) and men (b) per geographical location (site). Correlation coefficients: Amsterdam Ghanaians (women: 0.447, 95% CI: 0.386–0.504; men: 0.557, 95% CI: 0.492–0.616), Urban Ghanaians (women: 0.424, 95% CI: 0.338–0.503; men: 0.443, 95% CI: 0.309–0.560), and Rural Ghanaians (women: 0.734, 95% CI: 0.693–0.770; men: 0.557, 95% CI: 0.479–0.625).
Fig. 3.
a, b Correlation between serum leptin and waist circumference (WC) in women (a) and men (b) per geographical location (site). Correlation coefficients: Amsterdam Ghanaians (women: 0.429, 95% CI: 0.367–0.487; men: 0.570, 95% CI: 0.507–0.628), Urban Ghanaians (women: 0.375, 95% CI: 0.286–0.458; men: 0.532, 95% CI: 0.411–0.635), and Rural Ghanaians (women: 0.687, 95% CI: 0.641–0.729; men: 0.641, 95% CI: 0.575–0.699).
Leptin
Details of the relationship between serum leptin and cardiometabolic traits per geographical location are shown in Table 2. Variables related to body mass explained the largest proportion of the variance in serum leptin across the different locations with the explained variance of variables related to body mass being disproportionately higher in rural Ghanaians. Consistently across all geographic locations, a unit change in BMI and WC was associated with a statistically significant (p value <0.05) percentage increase in serum leptin in women and men, respectively. To a lesser extent, variables related to insulin resistance also significantly explained the variance in serum leptin across all geographic locations (p value <0.05), except in urban Ghanaian men. Variables related to inflammation (CRP) did not have a statistically significant association with serum leptin except in participants living in Amsterdam. The same was true for HDL-cholesterol, which was associated with serum leptin in participants living in Amsterdam, but not or less so in urban and rural Ghanaian men and women. There was a clear association between LDL-cholesterol and leptin independent of location in both men and women.
Table 2.
Linear regression of serum leptin and cardiometabolic traits adjusted for age per geographical location and sex
| Cardiometabolic traits | Amsterdam Ghanaians | Urban Ghanaians | Rural Ghanaians | ||||||
|---|---|---|---|---|---|---|---|---|---|
| beta (95% CI) | p value | R 2 | beta (95% CI) | p value | R 2 | beta (95% CI) | p value | R 2 | |
| Women | |||||||||
| Variables related to body mass | | | 0.201 | | | 0.176 | | | 0.547 |
| BMI | 1.02 (1.01–1.03) | 0.006 | | 1.03 (1.01–1.05) | 0.001 | | 1.06 (1.04–1.08) | <0.001 | |
| WC | 1.01 (1.00–1.01) | 0.078 | | 1.00 (0.99–1.01) | 0.541 | | 1.02 (1.01–1.02) | <0.001 | |
| WHR | 1.00 (0.99–1.00) | 0.605 | | 1.00 (0.99–1.01) | 0.677 | | 0.99 (0.98–0.99) | 0.010 | |
| Variables related to insulin resistance | | | 0.069 | | | 0.050 | | | 0.207 |
| HOMA-IR | 1.09 (1.06–1.12) | <0.001 | | 1.08 (1.05–1.11) | <0.001 | | 1.20 (1.15–1.25) | <0.001 | |
| FPG | 0.92 (0.88–0.96) | <0.001 | | 0.97 (0.93–1.02) | 0.196 | | 0.87 (0.84–0.91) | <0.001 | |
| HbA1C | 1.01 (1.00–1.01) | 0.010 | | 1.00 (0.99–1.01) | 0.776 | | 1.02 (1.01–1.03) | <0.001 | |
| Variables related to inflammation | | | 0.058 | | | 0.004 | | | 0.005 |
| CRP | 1.02 (1.01–1.03) | <0.001 | | 1.00 (1.00–1.01) | 0.917 | | 1.00 (1.00–1.01) | 0.439 | |
| Variables related to lipid metabolism | | | 0.035 | | | 0.024 | | | 0.084 |
| HDL | 0.89 (0.82–0.96) | 0.004 | | 0.80 (0.68–0.94) | 0.007 | | 0.87 (0.76–0.99) | 0.043 | |
| LDL | 1.04 (1.01–1.07) | 0.005 | | 1.07 (1.02–1.12) | 0.011 | | 1.20 (1.14–1.26) | <0.001 | |
| Men | |||||||||
| Variables related to body mass | | | 0.338 | | | 0.285 | | | 0.415 |
| BMI | 1.03 (1.00–1.06) | 0.082 | | 1.00 (0.93–1.06) | 0.911 | | 1.04 (1.00–1.08) | 0.082 | |
| WC | 1.03 (1.01–1.04) | <0.001 | | 1.04 (1.01–1.07) | 0.020 | | 1.06 (1.03–1.07) | <0.001 | |
| WHR | 0.99 (0.98–0.99) | 0.043 | | 1.02 (1.00–1.05) | 0.103 | | 0.99 (0.98–1.00) | 0.227 | |
| Variables related to insulin resistance | | | 0.147 | | | 0.018 | | | 0.043 |
| HOMA-IR | 1.14 (1.10–1.18) | <0.001 | | 1.12 (1.01–1.24) | 0.043 | | 1.06 (1.01–1.13) | 0.042 | |
| FPG | 0.94 (0.91–0.97) | <0.001 | | 0.86 (0.74–0.99) | 0.036 | | 0.97 (0.89–1.06) | 0.460 | |
| HbA1C | 1.01 (1.00–1.01) | 0.007 | | 1.01 (1.00–1.03) | 0.117 | | 1.01 (1.00–1.02) | 0.167 | |
| Variables related to inflammation | | | 0.028 | | | 0.009 | | | 0.021 |
| CRP | 1.01 (1.00–1.01) | 0.026 | | 1.00 (0.98–1.01) | 0.681 | | 1.00 (0.99–1.01) | 0.786 | |
| Variables related to lipid metabolism | | | 0.066 | | | 0.032 | | | 0.103 |
| HDL | 0.74 (0.64–0.86) | <0.001 | | 0.82 (0.55–1.22) | 0.322 | | 0.86 (0.71–1.04) | 0.115 | |
| LDL | 1.09 (1.03–1.14) | 0.002 | | 1.20 (1.05–1.37) | 0.007 | | 1.29 (1.18–1.41) | <0.001 | |
Multivariable analysis was done per cardiometabolic trait category, adjusted for age.
Adiponectin
Details of the relationship between serum adiponectin and cardiometabolic traits per geographical location are shown in Table 3. Variables related to insulin resistance and to a lesser extent variables related to body mass explained the variance in serum adiponectin across all geographic locations. In both women and men, there was largely an inverse relation between BMI and serum adiponectin, and between WC and serum adiponectin, although not statistically significant. The association with HOMA-IR was comparable between men and women across the different locations except in men living in rural Ghana. The association between CRP and adiponectin was only apparent in Amsterdam and Urban Ghanaian women but not in rural Ghanaian women and Ghanaian men in all locations. Regardless of location, serum adiponectin was associated with HDL-cholesterol in both men and women, while the association with LDL-cholesterol was weak or non-existent.
Table 3.
Linear regression of serum adiponectin and cardiometabolic traits adjusted for age per location and sex
| Cardiometabolic traits | Amsterdam Ghanaians | Urban Ghanaians | Rural Ghanaians | ||||||
|---|---|---|---|---|---|---|---|---|---|
| beta (95% CI) | p value | R 2 | beta (95% CI) | p value | R 2 | beta (95% CI) | p value | R 2 | |
| Women | |||||||||
| Variables related to body mass | | | 0.064 | | | 0.117 | | | 0.097 |
| BMI | 1.00 (0.98–1.02) | 0.941 | | 0.99 (0.98–1.00) | 0.219 | | 0.98 (0.96–0.99) | 0.026 | |
| WC | 0.99 (0.98–1.00) | 0.142 | | 0.99 (0.99–1.00) | 0.133 | | 1.00 (0.99–1.01) | 0.781 | |
| WHR | 0.99 (0.98–1.00) | 0.061 | | 1.00 (0.23–0.62) | 0.537 | | 0.99 (0.98–0.99) | 0.004 | |
| Variables related to insulin resistance | | | 0.092 | | | 0.135 | | | 0.076 |
| HOMA-IR | 0.87 (0.84–0.90) | <0.001 | | 0.94 (0.92–0.96) | <0.001 | | 0.94 (0.91–0.97) | <0.001 | |
| FPG | 1.12 (1.06–1.18) | <0.001 | | 1.02 (0.99–1.05) | 0.104 | | 1.01 (0.98–1.04) | 0.547 | |
| HbA1C | 0.99 (0.98–0.99) | 0.002 | | 1.00 (0.99–1.00) | 0.118 | | 0.99 (0.98–0.99) | 0.009 | |
| Variables related to inflammation | | | 0.026 | | | 0.009 | | | 0.009 |
| CRP | 0.98 (0.97–0.99) | <0.001 | | 1.00 (1.00–1.01) | 0.021 | | 1.00 (1.00–1.01) | 0.481 | |
| Variables related to lipid metabolism | | | 0.062 | | | 0.051 | | | 0.064 |
| HDL | 1.43 (1.29–1.59) | <0.001 | | 1.27 (1.15–1.40) | <0.001 | | 1.31 (1.18–1.45) | <0.001 | |
| LDL | 0.98 (0.94–1.02) | 0.232 | | 0.98 (0.95–1.01) | 0.125 | | 0.92 (0.89–0.96) | <0.001 | |
| Men | |||||||||
| Variables related to body mass | | | 0.026 | | | 0.105 | | | 0.199 |
| BMI | 0.98 (0.95–1.02) | 0.329 | | 0.97 (0.94–0.99) | 0.018 | | 0.96 (0.94–0.99) | 0.001 | |
| WC | 1.01 (0.99–1.03) | 0.274 | | 1.00 (0.99–1.01) | 0.813 | | 1.00 (0.99–1.01) | 0.832 | |
| WHR | 0.98 (0.96–0.99) | 0.002 | | 1.00 (0.99–1.01) | 0.737 | | 1.00 (0.99–1.00) | 0.241 | |
| Variables related to insulin resistance | | | 0.013 | | | 0.141 | | | 0.070 |
| HOMA-IR | 0.96 (0.92–0.99) | 0.024 | | 0.93 (0.90–0.97) | <0.001 | | 0.98 (0.95–1.01) | 0.258 | |
| FPG | 1.01 (0.98–1.04) | 0.401 | | 1.04 (0.99–1.09) | 0.133 | | 1.03 (0.98–1.08) | 0.164 | |
| HbA1C | 0.99 (0.99–1.00) | 0.163 | | 0.99 (0.98–0.99) | 0.003 | | 0.99 (0.98–0.99) | 0.007 | |
| Variables related to inflammation | | | 0.002 | | | 0.009 | | | 0.066 |
| CRP | 1.00 (0.99–1.01) | 0.676 | | 1.00 (0.99–1.01) | 0.660 | | 1.00 (1.00–1.01) | 0.608 | |
| Variables related to lipid metabolism | | | 0.029 | | | 0.065 | | | 0.109 |
| HDL | 1.38 (1.17–1.63) | <0.001 | | 1.21 (1.05–1.40) | 0.008 | | 1.20 (1.10–1.31) | <0.001 | |
| LDL | 1.03 (0.97–1.09) | 0.319 | | 0.93 (0.89–0.97) | 0.002 | | 0.97 (0.93–1.01) | 0.129 | |
Multivariable analysis was done per cardiometabolic trait category, adjusted for age.
Combined Age-Adjusted and Sex-Stratified Relation between Adipokines and Cardiometabolic Traits
BMI and WC explained the largest variations in serum leptin in both women and men, with a higher explained variance in men as compared to women for both parameters (online suppl. Table 7). Insulin resistance (HOMA-IR), total and LDL-cholesterol were also associated with leptin concentration and only explained 8.5%, 5.4%, and 5.6% of the variance in serum leptin, respectively, in women; 12.3%, 13.8%, and 14.9% of the variance, respectively, in men.
BMI and WC also explained the largest proportion of the variance in the age-adjusted concentrations of serum adiponectin in men (online suppl. Table 7). However, in women, variables of insulin resistance (HOMA-IR, FPG, HbA1C) explained 11.7% of the variance in the age-adjusted concentrations of serum adiponectin whilst variables of body mass (BMI, WC, WHR) only explained 8.5%.
Discussion
Key Findings
In this homogenous Ghanaian population living in different geographical context in Ghana and the Netherlands, we show that the relationship between leptin and measures of body mass (BMI and WC) was significantly different between urban and rural living Ghanaians and between Ghanaians living in Amsterdam and rural Ghana with a steeper association between leptin and measures of body mass in rural Ghanaians and a weaker association in urban Ghanaians and Ghanaians living in Amsterdam, while there were no differences in the association between leptin and body mass according to latitude. Further, among all the cardiometabolic traits examined, variables related to body mass were strongly associated with serum leptin and to a lesser extent adiponectin concentrations. In addition, BMI was the most important determinant of a percentage change in serum leptin levels among Ghanaian women, while WC was the key determinant of a percentage change in serum leptin among Ghanaian men. Variables related to insulin resistance were strongly associated with serum adiponectin and to a lesser extent serum leptin with geographical differences in the strength of the association. HOMA-IR was statistically significant in predicting a percentage change in serum leptin and adiponectin in all locations except among rural Ghanaian men. Finally, variables related to inflammation and lipid metabolism explained the least amount of variance in serum adipokines, with LDL cholesterol statistically significant in predicting percentage change only in serum leptin in all locations.
Discussion of Key Findings
Our finding of sharp differences observed in the relationship between serum leptin and measures of body mass during transition from rural to urban or cross-continental regions, with no differences seen across latitudes, suggests that leptin resistance develops following exposure to obesogenic environments as a result of rural to urban migration, while ambient temperature does not seem to play an important role. Several studies have described impaired transport of leptin across the blood-brain barrier; dysfunction in leptin receptor signalling pathways; and blockades in downstream neuronal circuitries in the pathophysiology of developing leptin resistance in in vivo and in vitro models [16, 29]. However, the exact mechanism for the development of leptin resistance during the rural-to-urban transition is poorly understood. Several studies have shown a strong association between migration and metabolic changes, including the development of leptin resistance [30–32]. In a study carried out in Peru by Frenken et al. [31], the migrant population showed a stronger correlation between leptin concentration and adiposity, compared to satiety signalling, suggesting disrupted central leptin sensitivity. Similarly in our previous study, we showed that post-migration individuals tend to exhibit higher BMI, circulating leptin levels, and features consistent with leptin resistance when compared to non-migrants, with little differences between urban Ghanaian dwellers and Ghanaians living in Amsterdam [23]. Our findings therefore corroborate these previous studies, leaving room for further research into the exact mechanism by which migration causes leptin resistance.
The strong relation between variables related to body mass, specifically BMI and WC, and serum leptin but less so with adiponectin is in keeping with findings of previous studies. Many studies have demonstrated a strong association between serum leptin and variables related to body mass, irrespective of age, sex, geographical location, and ethnicity [33–36]. The finding that BMI was the most significant determinant of a percentage change in serum leptin in Ghanaian women, and WC the key determinant of a percentage change in serum leptin in Ghanaian men, is in line with a population study in the Netherlands and a study among West Africans in Benin. Here, WC was the strongest determinant of serum leptin in European-Dutch men and BMI in European-Dutch women [10]; similarly, Awede et al. [37] showed that WC was more strongly correlated with serum leptin in West-African men. The sex differences in the key determinants of percentage change serum leptin, in terms of variables related to body mass, is in keeping with known differences in sex distribution of body mass. First, several studies have demonstrated sex differences in serum leptin attributable to the differences in total body fat, with women having significantly higher total body fat and a significantly lower WHR driven primarily by less central fat deposition compared to men [27, 38]. Second, the production of leptin differs per distribution of fat with visceral fat producing less leptin compared to subcutaneous fat [39]. Women have more subcutaneous fat compared to men, with WC being more closely related to visceral fat; hence, the sex difference in fat distribution also contributes to the differences as seen in our study. Finally, in vitro studies suggest that the rate of production of leptin from adipose tissue differs per sex, with adipose tissues in women producing leptin at a faster rate compared to men, which is also responsible for the higher values in women [38].
We observed in our study variables related to insulin resistance (HOMA-IR) were strongly associated with serum adiponectin and to a lesser extent serum leptin with geographical differences in the strength of the association. A possible explanation for this is the fact that adiponectin has been shown to be a more robust marker and mediator of metabolic health in obesity, while leptin is more indicative of overall adiposity and energy regulation [35, 40]. A possible explanation for the effect of geographical location on the strength of the association between serum adiponectin and variables related to insulin resistance and an alternative explanation for the geographical differences in the association between leptin and measures of body mass may relate to differences in gut microbiota composition. Migration and urbanization may lead to significant lifestyle modifications such as dietary changes and reduced physical activity leading to obesity and its related diseases [41, 42]. Mediated by changes in diet and the geographical environment, migration and urbanization affect the gut microbiome [43, 44]. Both in vitro and in vivo studies have shown a strong association between the gut microbiome and serum adipokines [45, 46]. In a recent study, a causal association via Mendelian randomization was shown to exist between the gut microbiota and circulating adipokine concentrations [47]. Clostridium bacteria, found abundantly in the gut, enhance short-chain fatty acids production, stimulating mRNA expression of adiponectin by adipocytes in T2DM [47]. In addition, the presence of abundant Olsenella species and reduced Enterohabdus species in the gut causes increased expression of leptin by adipocytes [47]. Finally, migration and urbanization strongly influence environmental factors such as temperature an individual is exposed to. Studies have shown a white-to-brown adipocyte differentiation (browning) as an adaptation to prolonged exposure to cold temperatures mediated by serum leptin [48, 49], depicting leptin as a potent thermoregulator in the human body. However, conflicting data exist on the effect of environmental temperature on serum leptin levels in different geographical contexts. While some studies describe higher leptin levels in men living in cold temperate regions [50], other studies have shown an increase in serum leptin levels in hot temperatures and a decrease in cold temperatures [51, 52]. In our previous study, migration and urbanization have been shown to be independently associated with serum adiponectin and leptin [23].
Except in rural Ghanaian men, HOMA-IR was consistently statistically significantly associated with serum leptin and adiponectin in this study. Previous data on the relationship between adipokines and HOMA-IR in African populations vary, with one study demonstrating a robust correlation between HOMA-IR and adiponectin, less so with leptin, and another showing no significant correlation between adiponectin and HOMA-IR [37, 53]. Our findings add to the ongoing debate, leaving room for further research into the relationship between HOMA-IR and adipokines. While some of the observed associations between the adipocytokines and other cardiometabolic traits besides variables related to body mass and insulin resistance in this study were not statistically significant, they may be clinically relevant. For example, HDL-cholesterol was associated with leptin in Amsterdam women and men but less so in urban and rural Ghanaian men and women. Further, there was a clear association between LDL-cholesterol and leptin independent of location in both men and women. Another example was with regard to adiponectin: regardless of location, serum adiponectin was associated with HDL-cholesterol in both men and women, while the association with LDL-cholesterol was weak/non-existent. Conflicting data exist on the relationship between serum leptin and markers of lipid metabolism such as LDL-cholesterol [54, 55]. In addition, limited data exist on the effect of migration and urbanization on the association between serum leptin and LDL-cholesterol in a homogenous population. Concerning variables related to inflammation, CRP was statistically significant in predicting percentage in serum leptin in Ghanaians living in Amsterdam and serum adiponectin in Ghanaian women living in Amsterdam but not in urban and rural Ghanaian participants suggesting that the relation between CRP and leptin is mediated by obesity-induced inflammation.
Strengths and Limitations
A key strength of the RODAM study was the use of well-standardized approaches in data collection across the various study sites. Second, the use of an ethnically homogenous Ghanaian population of predominantly Akan ancestral heritage living in different geographical locations in West Africa and Europe gave the study a unique strength in that it provided an unparalleled opportunity to assess the impact of migration and urbanization-related factors on the association between serum adiponectin and leptin and some cardiometabolic traits. Using an ethnically homogenous population attenuated the effect of ethnic variations believed to be associated with variations in serum adiponectin and leptin levels.
A key limitation of this study is its cross-sectional nature, prohibiting the possibility for investigating temporality in observed associations. In addition, although power analysis for sample calculation for the RODAM study was done [24], power analysis for sample size calculation for this specific subset analysis was not done, and this can also be a limitation to the study; however, this limitation was mitigated by doing a post hoc power analysis for the subset; in addition, this limitation is likely to be minimal as this study involved a significantly large number of participants. A final limitation to this study was the fact that the analysis was done on a subset of the RODAM participants leading to the possibility of selection bias; however, this was mitigated by randomly selecting participants for this subset analysis. Large population-based studies focused on specific ethnic groups in specific geographic locations may provide insight for further evaluating the role of migration and urbanization and its ensuing environmental factors on determinants of serum adiponectin and leptin providing context-specific insight to clinicians for interventions and predictions.
Conclusion
Across locations, there was significant variability in the relationship between serum adipokines and the cardiometabolic traits examined. Variables related to body mass were strongly associated with serum leptin and to a lesser extent to serum adiponectin with the relationship significantly modified per geographic location. Rural to urban migration was associated with an increase in leptin resistance. Similarly, variables related to insulin resistance were strongly associated with adiponectin and to a lesser extent serum leptin with relationship significantly modified per geographic location. There was significant heterogeneity on the effect of location on the association between serum adipokines and variables related to lipid metabolism and inflammation that may be influenced by our environment. These findings suggest that environmental factors significantly affect these associations. Migration and urbanization strongly influence the environmental factors individuals are exposed to. Thus, in developing interventions or preventive strategies against cardiometabolic disease, environmental factors must be considered. Further research needs to be done to fully understand the mechanism by which environmental factors affect these adipokines to aid in developing therapeutic interventions or robust markers of disease progression.
Acknowledgments
The authors are very grateful to the advisory board members for their valuable support in shaping the methods, to the research assistants, interviewers, and other staff of the five research locations who have taken part in gathering the data, and, most of all, to the Ghanaian volunteers participating in this project. We gratefully thank Jan van Straalen from the Academic Medical Centre for his valuable support with standardization of the laboratory procedures and the AMC Biobank for support in biobank management and storage of collected samples.
Statement of Ethics
This study was performed in accordance with the Declaration of Helsinki. The approval for the study protocol was granted at all sites from the respective Ethical Committees: in Ghana (School of Medical Sciences/Komfo Anokye Teaching Hospital Committee on Human Research, Publication and Ethical Review Board, September 2012; CHRPE/AP/200/12) and in the Netherlands (Institutional Review Board of the Academic Medical Centre, University of Amsterdam, April 2012; W12_062 # 12.17.0086), before data collection began in each country. Written informed consent was obtained from each participant before enrolment.
Conflict of Interest Statement
The authors declared they have nothing to disclose.
Funding Sources
This work was supported by the European Commission under the Framework Programme, Grant No. 278901. RSA and FA-D were supported by Bloomberg Professorship and institutional funds. KACM is supported by the NIH Pathway to Independence Award (K99/R00, DK131018).
Author Contributions
Bert-Jan van den Born and Yaw Kusi-Mensah conceived the study. Fredrick Ankoye-Danso and Rexford S. Ahimah performed adiponectin and leptin ELISAs. Yaw A. Kusi-Mensah performed the statistical analysis and wrote the manuscript, supervised by Charles Agyemang, Bert-Jan van den Born, Sean Chetty, and Charles Hayfron-Benjamin. Erik Beune, Karlijn Meeks, and Eva van den Linden reviewed and made inputs to the final manuscript. All authors read and approved the final version of the manuscript.
Funding Statement
This work was supported by the European Commission under the Framework Programme, Grant No. 278901. RSA and FA-D were supported by Bloomberg Professorship and institutional funds. KACM is supported by the NIH Pathway to Independence Award (K99/R00, DK131018).
Data Availability Statement
Some or all datasets generated during and/or analysed during the current study are not publicly available due to privacy concerns and in keeping with research integrity practices but are available from the corresponding author on reasonable request. In addition, any researcher can request for the data by submitting a proposal as outlined at https://www.rod-am.eu/.
Supplementary Material.
References
- 1. World Health Organization . Obesity and overweight. Geneva: WHO; 2024. [Google Scholar]
- 2. Reed J, Bain S, Kanamarlapudi V. Recent advances in understanding the role of glucagon-like peptide 1. F1000Res. 2020;9:F1000 Faculty Rev-239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Würfel M, Blüher M, Stumvoll M, Ebert T, Kovacs P, Tönjes A, et al. Adipokines as clinically relevant therapeutic targets in obesity. Biomedicines. 2023;11(5):1427. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Hui X, Lam KS, Vanhoutte PM, Xu A. Adiponectin and cardiovascular health: an update. Br J Pharmacol. 2012;165(3):574–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Varma B, Ogunmoroti O, Ndumele CE, Zhao D, Szklo M, Sweeney T, et al. Higher leptin levels are associated with coronary artery calcium progression: the Multi-Ethnic Study of Atherosclerosis (MESA). Diabet Epidemiol Manag. 2022;6:100047. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Beltowski J. Leptin and atherosclerosis. Atherosclerosis. 2006;189(1):47–60. [DOI] [PubMed] [Google Scholar]
- 7. Bełtowski J. Leptin and the regulation of endothelial function in physiological and pathological conditions. Clin Exp Pharmacol Physiol. 2012;39(2):168–78. [DOI] [PubMed] [Google Scholar]
- 8. Meeks KAC, Bentley AR, Doumatey AP, Adeyemo AA, Rotimi CN. Mendelian randomization study reveals a causal relationship between adiponectin and LDL cholesterol in Africans. Sci Rep. 2022;12(1):18955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Begum M, Choubey M, Tirumalasetty MB, Arbee S, Mohib MM, Wahiduzzaman M, et al. Adiponectin: a promising target for the treatment of diabetes and its complications. Life. 2023;13(11):2213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Ruige JB, Dekker JM, Blum WF, Stehouwer CDA, Nijpels G, Mooy J, et al. Leptin and variables of body adiposity, energy balance, and insulin resistance in a population-based study: the hoorn study. Diabetes Care. 1999;22(7):1097–104. [DOI] [PubMed] [Google Scholar]
- 11. Flier JS, Ahima RS. Leptin physiology and pathophysiology: knowns and unknowns 30 years after its discovery. J Clin Investig. 2024;134(1):e174595. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Zhao S, Zhu Y, Schultz RD, Li N, He Z, Zhang Z, et al. Partial leptin reduction as an insulin sensitization and weight loss strategy. Cell Metab. 2019;30(4):706–19.e6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Knight ZA, Hannan KS, Greenberg ML, Friedman JM. Hyperleptinemia is required for the development of leptin resistance. PLoS One. 2010;5(6):e11376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Dutra SCP, Moura EG, Rodrigues AL, Lisboa PC, Bonomo I, Toste FP, et al. Cold exposure restores the decrease in leptin receptors (OB-Rb) caused by neonatal leptin treatment in 30-day-old rats. J Endocrinol. 2007;195(2):351–8. [DOI] [PubMed] [Google Scholar]
- 15. Harris RBS, Mitchell TD, Kelso EW, Flatt WP. Changes in environmental temperature influence leptin responsiveness in low- and high-fat-fed mice. Am J Physiol Regul Integr Comp Physiol. 2007;293(1):R106–15. [DOI] [PubMed] [Google Scholar]
- 16. Banks WA, Coon AB, Robinson SM, Moinuddin A, Shultz JM, Nakaoke R, et al. Triglycerides induce leptin resistance at the blood-brain barrier. Diabetes. 2004;53(5):1253–60. [DOI] [PubMed] [Google Scholar]
- 17. Tomiyama AJ. Stress and obesity. Annu Rev Psychol. 2019;70:703–18. [DOI] [PubMed] [Google Scholar]
- 18. Luc K, Schramm-Luc A, Guzik TJ, Mikolajczyk TP. Oxidative stress and inflammatory markers in prediabetes and diabetes. J Physiol Pharmacol. 2019;70(6):809–24. [DOI] [PubMed] [Google Scholar]
- 19. Bouillon-Minois JB, Trousselard M, Thivel D, Benson AC, Schmidt J, Moustafa F, et al. Leptin as a biomarker of stress: a systematic review and meta-analysis. Nutrients. 2021;13(10):3350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Mente A, Razak F, Blankenberg S, Vuksan V, Davis AD, Miller R, et al. Ethnic variation in adiponectin and leptin levels and their association with adiposity and insulin resistance. Diabetes Care. 2010;33(7):1629–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Retnakaran R, Hanley AJG, Raif N, Connelly PW, Sermer M, Zinman B. Hypoadiponectinaemia in South Asian women during pregnancy: evidence of ethnic variation in adiponectin concentration. Diabet Med. 2004;21(4):388–92. [DOI] [PubMed] [Google Scholar]
- 22. Kurniawan F, Manurung MD, Harbuwono DS, Yunir E, Tsonaka R, Pradnjaparamita T, et al. Urbanization and unfavorable changes in metabolic profiles: a prospective cohort study of Indonesian young adults. Nutrients. 2022;14(16):3326. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Kusi-Mensah YA, Hayfron-Benjamin C, Chetty S, van der Linden EL, Meeks KAC, Beune E, et al. Serum adiponectin and leptin among Ghanaian migrants in Amsterdam and their compatriots in rural and urban Ghana: the RODAM Study. Clin Med Insights Endocrinol Diabetes. 2023;16:11795514231218592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Agyemang C, Beune E, Meeks K, Owusu-Dabo E, Agyei-Baffour P, Aikins AG, et al. Rationale and cross-sectional study design of the research on obesity and type 2 diabetes among African migrants: the RODAM study. BMJ Open. 2014;4(3):e004877. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Altman ME. STROBE Statement- checklist observational studies. J Clin Epidemiol. 2008;61(4):344–9. [DOI] [PubMed] [Google Scholar]
- 26. Wallace T, Matthews D, Wallace TM, Matthews DR. The assessment of insulin resistance in man. Diabet Med. 2002;19:527–34. [DOI] [PubMed] [Google Scholar]
- 27. Ohman-Hanson RA, Cree-Green M, Kelsey MM, Bessesen DH, Sharp TA, Pyle L, et al. Ethnic and sex differences in adiponectin: from childhood to adulthood. J Clin Endocrinol Metab. 2016;101(12):4808–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Christen T, Trompet S, Noordam R, van Klinken JB, van Dijk KW, Lamb HJ, et al. Sex differences in body fat distribution are related to sex differences in serum leptin and adiponectin. Pept. 2018;107:25–31. [DOI] [PubMed] [Google Scholar]
- 29. Banks WA, Farrell CL. Impaired transport of leptin across the blood-brain barrier in obesity is acquired and reversible. Am J Physiol Endocrinol Metab. 2003;285(1):E10–5. [DOI] [PubMed] [Google Scholar]
- 30. Najera H, Nandy S, Carrillo-Larco RM, Miranda JJ. Within-country migration and obesity dynamics: analysis of 94,783 women from the Peruvian demographic and health surveys. BMC Public Health. 2019;19(1):263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Lindgärde F, Ercilla MB, Correa LR, Ahrén B. Body adiposity, insulin, and leptin in subgroups of Peruvian Amerindians. High Alt Med Biol. 2004;5(1):27–31. [DOI] [PubMed] [Google Scholar]
- 32. Misra A, Ganda OP. Migration and its impact on adiposity and type 2 diabetes. Nutrition. 2007;23(9):696–708. [DOI] [PubMed] [Google Scholar]
- 33. Rosenbaum M, Nicolson M, Hirsch J, Murphy E, Chu F, Leibel RL. Effects of weight change on plasma leptin concentrations and energy expenditure. J Clin Endocrinol Metab. 1997;82(11):3647–54. [DOI] [PubMed] [Google Scholar]
- 34. Rosenbaum M, Nicolson M, Hirsch J, Leibel RL. Effects of gender, body composition, concentrations and menopause on plasma concentrations of leptin. J Clin Endocrinol Metab. 1996;81(9):3424–7. [DOI] [PubMed] [Google Scholar]
- 35. Gómez JM, Maravall FJ, Gómez N, Navarro MA, Casamitjana R, Soler J. Interactions between serum leptin, the insulin-like growth factor-I system, and sex, age, anthropometric and body composition variables in a healthy population randomly selected. Clin Endocrinol. 2003;58(2):213–9. [DOI] [PubMed] [Google Scholar]
- 36. Shah A, Hernandez A, Mathur D, Budoff MJ, Kanaya AM. Adipokines and body fat composition in south Asians: results of the Metabolic Syndrome and Atherosclerosis in South Asians living in America (MASALA) study. Int J Obes. 2012;36(6):810–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Awede B, Adovoekpe D, Adehan G, MacFarlane NG, Azonbakin S, Dossou E, et al. 2018, Adiponectin, in contrast to leptin, is not associated with body mass index, waist circumference and HOMA-IR in subjects of a West-African population. Physiol Rep. 6(11):e13718. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Hellström L, Wahrenberg H, Hruska K, Reynisdottir S, Arner P. Mechanisms behind gender differences in circulating leptin levels. J Intern Med. 2000;247(4):457–62. [DOI] [PubMed] [Google Scholar]
- 39. Van Harmelen V, Reynisdottir S, Eriksson P, Thörne A, Hoffstedt J, Lönnqvist F, et al. Leptin secretion from subcutaneous and visceral adipose tissue in women. Diabetes. 1998;47(6):913–7. [DOI] [PubMed] [Google Scholar]
- 40. Baratta R, Amato S, Degano C, Farina MG, Patanè G, Vigneri R, et al. Adiponectin relationship with lipid metabolism is independent of body fat mass: evidence from both cross-sectional and intervention studies. J Clin Endocrinol Metab. 2004;89(6):2665–71. [DOI] [PubMed] [Google Scholar]
- 41. Riha J, Karabarinde A, Ssenyomo G, Allender S, Asiki G, Kamali A, et al. Urbanicity and lifestyle risk factors for cardiometabolic diseases in rural Uganda: a cross-sectional Study. PLoS Med. 2014;11(7):e1001683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. O’connor MY, Thoreson CK, Ricks M, Courville AB, Thomas F, Yao J, et al. Worse cardiometabolic health in African immigrant men than African American Men: reconsideration of the healthy immigrant effect. Metab Syndr Relat Disord. 2014;12(6):347–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Shad NS, Shaikh NI, Cunningham SA. Migration spurs changes in the human microbiome: a review. J Racial Ethn Health Disparities. 2024;11(6):3618–32. [DOI] [PubMed] [Google Scholar]
- 44. Vangay P, Johnson AJ, Ward TL, Al-Ghalith GA, Shields-Cutler RR, Hillmann BM, et al. US immigration westernizes the Human Gut microbiome. Cell. 2018;175(4):962–72.e10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Sanz Y, Santacruz A, Gauffin P. Gut microbiota in obesity and metabolic disorders. Proc Nutr Soc. 2010;69(3):434–41. [DOI] [PubMed] [Google Scholar]
- 46. Queipo-Ortuño MI, Seoane LM, Murri M, Pardo M, Gomez-Zumaquero JM, Cardona F, et al. Gut Microbiota composition in Male rat models under different nutritional status and physical activity and its Association with Serum Leptin and Ghrelin levels. PLoS One. 2013;8(5):e65465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Zhang T, Liu J, Liu X, Wang Q, Zhang H. The causal impact of gut microbiota on circulating adipokine concentrations: a two-sample Mendelian randomization study. Hormones. 2024;23(4):789–99. [DOI] [PubMed] [Google Scholar]
- 48. Petrovic N, Walden TB, Shabalina IG, Timmons JA, Cannon B, Nedergaard J. Chronic peroxisome proliferator-activated receptor gamma (PPARgamma) activation of epididymally derived white adipocyte cultures reveals a population of thermogenically competent, UCP1-containing adipocytes molecularly distinct from classic brown adipocytes. J Biol Chem. 2010;285(10):7153–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Kajimura S, Seale P, Kubota K, Lunsford E, Frangioni JV, Gygi SP, et al. Initiation of myoblast to brown fat switch by a PRDM16–C/EBP-β transcriptional complex. Nature. 2009;460(7259):1154–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50. Nikanorova AA, Barashkov NA, Nakhodkin SS, Pshennikova VG, Solovyev AV, Romanov GP, et al. The role of leptin levels in adaptation to cold climates. Int J Environ Res Public Health. 2020;17(6):1854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Mandic I, Ahmed M, Rhind S, Goodman L, L’Abbe M, Jacobs I. The effects of exercise and ambient temperature on dietary intake, appetite sensation, and appetite regulating hormone concentrations. Nutr Metab. 2019;16(1):29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. Zeyl A, Stocks JM, Taylor NAS, Jenkins AB. Interactions between temperature and human leptin physiology in vivo and in vitro. Eur J Appl Physiol. 2004;92(4–5):571–8. [DOI] [PubMed] [Google Scholar]
- 53. Ayina CNA, Noubiap JJN, Etoundi Ngoa LS, Boudou P, Gautier JF, Mengnjo MK, et al. Association of serum leptin and adiponectin with anthropomorphic indices of obesity, blood lipids and insulin resistance in a Sub-Saharan African population. Lipids Health Dis. 2016;15(1):96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Akinjinmi A, Amballi A, Abdulrahman A, Aderinto N, Adeniyi AA, Akindele A, et al. Relationship between Serum Leptin, lipid metabolism, HbA1c, and renal function in individuals with type 2 diabetes mellitus and obesity and in individuals with type 2 diabetes mellitus without obesity. Medicine. 2025;104(46):e45526. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. Oguz S, Aytan H, Tapisiz OL, Günyeli I, Erdem S, Tuncay G, et al. Is leptin associated with serum lipid profile or any other parameters of blood biochemistry in post-menopausal women with osteoporosis? J Int Med Res. 2004;32(4):359–66. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Some or all datasets generated during and/or analysed during the current study are not publicly available due to privacy concerns and in keeping with research integrity practices but are available from the corresponding author on reasonable request. In addition, any researcher can request for the data by submitting a proposal as outlined at https://www.rod-am.eu/.