Increasing trends in adult obesity in Greenland: a serial cross-sectional study from 1999 through 2018 Authors: Helene Nielsen, Christian Landbo, Ninna Senftleber, Christina Viskum Lytken Larsen, Peter Bjerregaard, Thorkild I A Sørensen, Marit Eika Jørgensen, Dirk Lund Christensen, Stine Byberg Published: 2026 ABSTRACT ABSTRACTThe prevalence of obesity has increased among Inuit populations across the Arctic, although some studies indicate stabilisation. The Greenlandic Inuit may exhibit distinct obesity patterns due to unique genetic factors and rapid societal transitions. We investigated temporal trends in obesity prevalence among adults in Greenland from 1999 to 2018 by age, sex and place of residence. We analysed data from serial cross-sectional Greenland Population Health Surveys conducted in 1999–2001 (N = 1,004), 2005–2010 (N = 3,031), 2014 (N = 511), and 2018 (N = 859). Obesity was defined as a body mass index ≥ 30 kg/m². The prevalence of obesity increased among all groups. Among women, the prevalence rose from 20.0% (95% CI 16.6, 23.4) in 1999 to 29.6% (95% CI 25.0, 34.2) in 2018; among men, it rose from 14.4% (95% CI 11.2, 17.6) to 22.2% (95% CI 18.5, 26.0). Women maintained a higher prevalence than men across all surveys. Age-specific patterns showed increases up to age 40, then plateaued in both sexes. Geographic differences among women diminished by 2018, whereas men in settlements consistently showed the lowest prevalence. The persistent rise across all demographic groups, combined with geographic convergence among women but not men, highlights the need for targeted, culturally appropriate public health strategies.KEYWORDS: Obesity, Inuit Health, serial cross-sectional study, epidemiology, Greenland Introduction Obesity is recognised by the World Health Organization (WHO) as a global public health emergency [1] and has emerged as one of the most significant public health challenges in Greenland [2]. Obesity is caused by a multifactorial causal relationship between genetic, environmental and behavioural factors [3]. According to the WHO, obesity is characterised by excessive fat accumulation that can impair health and is practically defined as a body mass index (BMI) ≥ 30 kg/m2 [4]. BMI is the most commonly applied and reproducible marker of adiposity in epidemiological research. However, other adipose markers, such as waist circumference (WC) and waist-to-hip ratio (WHR), are also widely used. Greenland is home to approximately 56,000 inhabitants, most of whom live along the coastline in communities ranging from the capital Nuuk (population 19,903) [5] to small settlements with fewer than 450 inhabitants [6]. There are no connecting roads between communities, and marked disparities exist between towns and settlements in terms of socio-economic conditions, infrastructure, including healthcare access, and economic opportunities. The Greenlandic population is predominantly of Inuit origin, with approximately 90% classified as Inuit, defined as individuals born in Greenland with at least one parent also born in Greenland [7]. Immigrants account for approximately 10% of the population, primarily of Danish origin [8]. Greenland was a Danish colony from the early 18th century, and its indigenous Inuit population has historically been geographically isolated. In 1953, Greenland became an integral part of Denmark, and the modernisation policies that followed led to the rapid urbanisation of the Greenlandic society [2]. This urbanisation gradually shifted the traditional Inuit diet, which is rich in fish and marine mammals, to higher consumption of saturated fats and simple carbohydrates, and replaced a physically active lifestyle with increasingly sedentary behaviour [9,10]. This transition, along with the high impact of rare genetic variants unique to this population, such as the ADCY3 variant linked to increased obesity risk [11], underscores the need for targeted research in this population. Since the 1990s, the Greenland population health surveys have systematically monitored obesity indicators [12–14]. Data from the most recent population health survey documented a substantial increase in the crude adult obesity prevalence, measured by BMI, from 1999 to 2018, rising from 23.0% to 32% among women and from 17% to 24% among men [12–14]. Despite documented increases in obesity prevalence, little is known about how these trends vary by demographic subgroups. Therefore, we aimed to investigate the temporal trends in obesity prevalence among adults in Greenland and whether the prevalence differed across age, sex and place of residence. Additionally, in sensitivity analyses, we investigated whether the use of other anthropometric measurements to define obesity influenced the temporal trends. Materials and methods Study design and population We used data from the Greenland population health surveys conducted in 1999–2001 (Inuit Health in Greenland, B99) (N = 1,728, 62% participation), 2005–2010 (Inuit Health in Transition, IHIT) (N = 3,253, 67% participation), 2014 (The Greenland Population Health Survey 2014, B2014) (N = 2,188, 63% participation), and 2018 (The Greenland Population Health Survey 2018, B2018) (N = 2,539,52% participation) [12–15] and included data on objectively measured height, weight, and waist and hip circumference. The surveys used stratified random sampling, selecting participants to reflect various community sizes and geographical locations in Greenland, and were considered representative of the general adult population. Identifying as Inuit was not an inclusion criterion. In B99, participants were included if they had at least one Greenlandic grandparent, or, if this information was missing, at least one Greenlandic parent. In the IHIT, B2014 and B2018 surveys, participants consisted of persons born in Greenland or Denmark and living in Greenland. Based on self-identification and primary language, the majority of participants across all surveys were categorised as Inuit (~ 96%) [12–16]. The B99 survey included Greenlanders living in West Greenland and Denmark. The Greenlandic part of the survey used in this present study was conducted in seven locations, both larger towns and smaller settlements, ranging from the capital of Nuuk to small settlements. All persons included in the survey were adults (≥18 years) [12]. The IHIT survey was conducted in 9 towns and 13 settlements in Greenland, covering towns from the north to the south, and the west and east of Greenland [15]. The B2014 was conducted as a follow-up of participants in previous Greenland population health surveys. Adults from 11 towns and 8 settlements from across Greenland are represented in the survey [14]. The most recent survey, B2018, included participants aged 15 years and above living in 12 towns and 8 settlements across Greenland [13]. All surveys collected clinical and questionnaire data. For the present study, only participants living in Greenland were included. Participants were included at the date of their first participation in one of the population health surveys. Persons with missing information on BMI, with implausible BMI values (BMI ≤ 16.5 kg/m2 or BMI > 60 kg/m2), and children (age at inclusion < 18 years) were excluded from the study population. A flow diagram of the study population is presented in Figure 1. The study was approved by the ethical review committee in Greenland and the health authorities (project numbers: B99, 505-42; Inuit Health in Transition, 505-95; B2014, 2014-10; B2018, 2017-05). All participants provided both written and oral informed consent. Flow chart of the study population. a BMI ≥ 16.5 kg/m2 or BMI < 60 kg/m2. Abbreviations: BMI; body mass index, B99; Inuit Health in Greenland, IHIT, Inuit Health in Transition, B2014; The Greenland Population Health Survey 2014, B2018; The Greenland Population Health Survey 2018. Definition of obesity All data used to define obesity are based on objective measurements by researchers conducting the Greenland population health surveys. Height and weight were measured with participants wearing light clothing and no shoes. WC was measured midway between the iliac crest and the rib cage while standing, and hip circumference was measured at its widest point. BMI was calculated as weight divided by height squared (kg/m2), and the WHR was calculated as waist circumference (cm) divided by hip circumference (cm). Obesity was defined according to the WHO cut-off for BMI (BMI ≥ 30 kg/m2), WC (women ≥ 80 cm, men ≥ 94 cm), and WHR (women ≥ 0.85 cm, men ≥ 0.90 cm) [17]. Covariates Information on sex and date of birth was retrieved from the civil registration system. Age was calculated based on date of birth and date of survey participation. Place of residence was recorded by interviewers at the survey location and categorised into three groups: Nuuk (the capital), towns, or settlements. The majority of Greenland's population lives on the west coast, distributed across the capital Nuuk, smaller towns and remote settlements. These settings differ considerably in size, access to healthcare, and imported food supply, and the transition from a traditional to a Western diet and lifestyle has progressed at different rates across the three settings. We therefore used place of residence as a proxy for living conditions, which is consistent with previous studies that applied this measure to assess the degree of urbanisation in health research [18]. Statistical analyses Table 1 presents the baseline characteristics of the participants. Normally distributed continuous variables are reported as mean (SD), skewed continuous variables as median (IQR), and categorical variables as numbers (percentages). Obesity defined by BMI was the primary outcome, whereas obesity estimates based on WC and WHR were used as outcomes in sensitivity analyses. The crude prevalence of obesity with a 95% confidence interval (95% CI) was calculated for each survey according to sex and place of residence and is presented in Table 2. To assess potential differences in the age-specific prevalence of obesity across the surveys, we fitted a multiple logistic regression model. The model included sex, age, survey, and an interaction term between age and sex, survey and sex, and survey and age to account for potential variations in effects. To evaluate deviations from linearity in the association between age and obesity, quadratic and cubic terms for age were tested in the model. The association between age and obesity was not linear. Thus, age was included in the model as a natural spline. A natural spline is a piecewise polynomial function defined between a specified set of knots, where the function is restricted to be linear at the first and last knots. Knots are the specific points where the curve changes direction. Multiple models with varying knot configurations were fitted to determine the optimal number of knots. The Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were applied to evaluate the model fit. Preference was given to models exhibiting lower AIC or BIC values. Based on these criteria, age was included in the model as a natural spline with three knots. The placement of the knots was determined by evenly distributing the age between the knots. To examine whether the prevalence of obesity varied between individuals living in settlements, towns, or in the capital Nuuk, we created an additional multiple logistic regression model. The model included age, sex, residency, survey, and an interaction term between residency and survey, residency and sex, and residency and age to account for potential variations in the residency-specific effects. Similarly, the linearity of age was investigated. We found no linear association between age and obesity. Thus, age was included in the model as a natural spline with three knots, based on an even distribution of age between each knot. We provided the predicted prevalence estimates for 31-, 44- and 55-year-old women and men, corresponding to the first quartile, median, and third quartile ages, respectively. To investigate whether the choice of obesity measure influenced the estimation of temporal trends, we conducted sensitivity analysis using WC and WHR as alternative measures of obesity. All the statistical analyses were performed using R software, version 4.5.0 [19]. Table 1.Distribution of the study population on key variables stratified by survey. B99IHITB2014B2018OverallSurvey year1999–20012005–2010201420181999–2018N. of participants(N = 1004)(N = 3031)(N = 511)(N = 859)(N = 5405) BMI (kg/m 2 ) Median [IQR]25.2 [22.6, 28.6]25.6 [22.5, 29.6]25.7 [22.5, 29.8]26.1 [22.8, 30.0]25.6 [22.6, 29.5] Age (years) Median [IQR]42 [32, 55]44 [34, 54]44 [23, 53]36 [24, 53]43 [31, 54] Sex, n (%) Women545 (54.3%)1687 (55.7%)298 (58.3%)382 (44.5%)2912 (53.9%)Men459 (45.7%)1344 (44.3%)213 (41.7%)477 (55.5%)2493 (46.1%) Residency, n (%) Nuuk432 (43.0%)427 (14.1%)135 (26.4%)111 (12.9%)1105 (20.4%)Settlement255 (25.4%)732 (24.2%)40 (7.8%)217 (25.3%)1244 (23.0%)Town317 (31.6%)1872 (61.8%)336 (65.8%)531 (61.8%)3056 (56.5%) Weight (kg) Median [IQR]67.0 [59.3, 76.0]67.6 [58.2, 78.7]69.0 [60.0, 81.0]72.2 [61.5, 84.1]68.2 [59.0, 79.3] Height (cm) Mean (SD)162 (9.39)162 (9.24)164 (9.45)166 (9.69)163 (9.45) Waist circumference (cm) Median [IQR]86.0 [79.0, 95.8]90.5 [81.5, 101]89.0 [80.0, 101]91.0 [81.5, 101]89.5 [81.0, 100]Missing5 (0.5%)52 (1.7%)2 (0.4%)7 (0.8%)66 (1.2%) Hip circumference (cm) Median [IQR]97.0 [92.0, 103.0]98.5 [92.5, 105.0]101 [95.0, 108.0]99.0 [92.0, 105.0]98.5 [92.8, 105.0]Missing5 (0.5%)53 (1.7%)0 (0%)8 (0.9%)66 (1.2%)Abbreviations: BMI; Body mass index, B99; Inuit Health in Greenland, IHIT; Inuit Health in Transition, B2014; The Greenland Population Health Survey 2014, B2018; The Greenland Population Health Survey 2018. Distribution of the study population on key variables stratified by survey. Abbreviations: BMI; Body mass index, B99; Inuit Health in Greenland, IHIT; Inuit Health in Transition, B2014; The Greenland Population Health Survey 2014, B2018; The Greenland Population Health Survey 2018. Table 2.Crude prevalence (%) of adult obesity (BMI ≥ 30) in each survey. B99IHITB2014B2018Survey year1999–20012005–201020142018 Sex Women20.0 (16.6, 23.4)26.2 (24.1, 28.3)27.2 (22.1, 32.2)29.6 (25.0, 34.2)Men14.4 (11.2, 17.6)19.3 (17.2, 21.4)21.6 (16.1, 27.1)22.2 (18.5, 26) Residency Nuuk17.4 (13.8, 20.9)23.2 (19.2, 27.2)26.7 (19.2, 34.1)24.3 (16.3, 32.3)Settlement14.5 (10.2, 18.8)21.9 (18.9, 24.9)10.0 (0.7, 19.3)28.1 (22.1, 34.1)Town19.9 (15.5, 24.3)23.6 (21.7, 25.5)25.9 (21.2, 30.6)24.7 (21, 28.3)Abbreviations: BMI; Body Mass Index, B99; Inuit Health in Greenland, IHIT; Inuit Health in Transition, B2014; The Greenland Population Health Survey 2014, B2018; The Greenland Population Health Survey 2018. Crude prevalence (%) of adult obesity (BMI ≥ 30) in each survey. Abbreviations: BMI; Body Mass Index, B99; Inuit Health in Greenland, IHIT; Inuit Health in Transition, B2014; The Greenland Population Health Survey 2014, B2018; The Greenland Population Health Survey 2018. Results In total, 5,405 individuals were included in the study population. The median BMI increased over the survey years, peaking at 26.1 kg/m² in the B2018 surveys. There was a slight decrease in the mean age at inclusion, from 42 years in B99 to 36 years in B2018. The sex distribution remained relatively equal throughout the survey years, except for B2014, where 58.3% of participants were women. Additionally, a lower percentage of individuals living in settlements were included in the B2014 survey (see Table 1). Table 2 shows the crude prevalence and corresponding 95% CI of adult obesity in each survey by sex and residency. The prevalence of obesity increased consistently among both women (from 20.0% in B99 to 29.6% in B2018) and men (from 14.4% in B99 to 22.2% in B2018). Trends across places of residence were less consistent: Nuuk showed an increase through B2014 followed by a decline, settlements showed marked variation, and towns remained relatively stable across survey periods. Figure 2 shows the estimated age-specific prevalence and 95% CI (indicated by the shaded area) of obesity for each survey, stratified by sex, based on the multiple logistic regression model. Through the surveys, the lowest prevalence was found among the youngest ages, with an almost linear increase from ages 20 to 40 years, whereafter the increase plateaued for both women and men. Women consistently had a higher prevalence of obesity than men throughout the survey years. Among women, the age-specific obesity prevalence demonstrated an increasing temporal trend, with the highest prevalence of obesity occurring in the most recent survey period among all ages. However, there was an overlap in the confidence intervals, especially among the two latest surveys (indicated by the shaded area). A similar pattern of increase in the age-specific prevalence was found among men, with men in the B99 survey having the lowest prevalence across ages and those in B2018 having the highest age-specific prevalence. Age-specific prevalence of obesity in Greenland according to a survey stratified by sex. The lines show the estimated prevalence of obesity (%) for each survey with a shaded area corresponding to a 95% CI. Abbreviations: BMI; Body mass index, B99; Inuit Health in Greenland, IHIT,;Inuit Health in Transition, B2014; The Greenland Population Health Survey 2014, B2018; The Greenland Population Health Survey 2018. Figure 3 shows the predicted obesity prevalence with 95% CI by residency and survey, stratified by sex. Among women, obesity prevalence was comparable across all places of residency. Among men, those residing in settlements maintained a consistently lower prevalence, while men in Nuuk and towns exhibited similar levels, particularly in earlier survey years. Both sexes demonstrated their highest age-specific prevalence in the most recent survey. In additional analyses presented in the appendix, we estimated the temporal trends in obesity using WC and WHR. Despite differences in prevalence estimates, the overall trend of increasing obesity prevalence over the survey years remained consistent across all anthropometric measures of obesity. Age-specific prevalence (%) of obesity in Greenland according to residency by stratified by sex. The dots represent the estimated prevalence based on a logistic regression model including survey, age, residency and sex as covariates. The error bar represents the 95% CI, and the dotted line connects the prevalence estimates of the four surveys stratified by the place of residency. The prevalence is estimated for a 31-, 43- and 54-year-old man and woman living in the capital Nuuk, a town, or a settlement. Abbreviations: BMI; Body mass index, B99; Inuit Health in Greenland, IHIT; Inuit Health in Transition, B2014; The Greenland Population Health Survey 2014, B2018; The Greenland Population Health Survey 2018. Discussion This serial cross-sectional study examined the temporal trends in adult obesity using four population health surveys conducted in Greenland from 1999 to 2018. The crude prevalence of obesity increased from 20.0% in B99 to 29.6% in B2018 among women and from 14.4% to 22.2% among men. Using various models, we estimated the age-, sex- and residency-specific prevalence of obesity. We found a consistent increase in the prevalence of obesity across all ages, sexes, and types of residencies throughout the survey years. Women consistently had a higher obesity prevalence than men, and it increased with age in all surveys from an early age. Among men, those residing in settlements displayed a consistently lower obesity prevalence compared with their counterparts in Nuuk or towns across all age groups. Andersen et al. [20] analysed BMI trends among Inuit and non-Inuit participants aged 50–69 years in East Greenland (1963, 1998 and 2008) and Nuuk (1998 and 2008). While both studies documented increasing obesity prevalence, Andersen et al. reported a levelling off in Nuuk from 1998 to 2008 [20]. In contrast, our analysis demonstrates continued increases across all geographic regions. This difference likely reflects differences in the study populations, locations, and extended study period. Our findings extend those of Bjerregaard et al. [21], who analysed BMI trends using data from the 1993, B99, and IHIT Greenland Population Health Surveys. The study found that the prevalence of obesity, defined by BMI and WC, increased for women from 1993 to 2010. For men, only WC increased, whereas the prevalence of obesity defined by BMI remained unchanged [21]. Our study, however, finds a consistent increase in the prevalence of obesity for both men and women. This discrepancy may be due to the inclusion of data from different periods, particularly our inclusion of data from the most recent 2018 population health survey. Notably, our study did not include data from the 1993 population health survey, as it lacked objective measures of obesity. Several studies and surveys have investigated obesity trends among Inuit populations in the circumpolar region. In Nunavut, an Inuit self-governed territory in northern Canada, the prevalence of obesity has increased more rapidly than the Canadian national average, with an obesity prevalence of 33.7% in the period 2015–2018 [22]. In Nunavik, northern Quebec, obesity prevalence among Inuit rose from 19% in 1992 to 28% in 2004 [23]. Consistent with our findings, studies across these populations have documented higher obesity prevalence among Inuit women compared to men [23], underscoring a disproportionate disease burden among women. A recent study of the worldwide trends in underweight and obesity revealed that obesity accounts for the majority of the combined increased burden globally [24]. Further, the WHO estimates that 16% of adults aged 18 years and older worldwide had obesity in 2022 and that the global prevalence of obesity more than doubled between 1990 and 2022 [4]. The current study showed that Greenland is experiencing a higher prevalence of obesity than the global average, while the prevalence is more similar to that found in Canada. The observed increase in obesity prevalence likely has implications for the broader cardiometabolic disease burden in Greenland. Obesity is a well-established risk factor for type 2 diabetes, hypertension, cardiovascular disease, and various cancers, and recent studies suggest that these conditions are already prominent and increasing among the Greenlandic population. Type 2 diabetes was almost non-existent half a century ago, yet the 2018 population health survey found a prevalence of 16% among adults aged 35 years and above [13]. A 2022 cross-sectional study of antihypertensive medication use in Greenland found an increase from 16.7% in 2020 to 17.5% in 2021, and a hypertension prevalence, defined based on medical diagnosis codes, was 7.9% in 2021 [25]. The population health surveys also demonstrated an increase in hypertension, defined by measured blood pressure or self-reported antihypertensive treatment, from 49% to 76% in men and from 34% to 61% in women between 2005–2010 and 2018 [13]. The incidence of cardiovascular disease (CVD) prevalence is higher among Greenlanders with type 2 diabetes than among those without, and rates of ischaemic heart disease now mirror those reported in other Western populations [26,27]. However, a study of Greenlandic Inuit with a median follow-up of 10.5 years found no significant increase in CVD incidence with increasing BMI in models adjusted for demographic, lifestyle and cardiometabolic risk factors [28]. The incidence of obesity-related cancers, including colorectal and breast cancer, has also increased markedly among Greenlandic Inuit over recent decades [29]. If obesity prevalence continues to rise, these conditions are likely to increase further, although part of the observed increase may also reflect improved detection, better treatment, and longer survival with chronic conditions. The increasing use of pharmacological weight-lowering treatments, such as GLP-1 receptor agonists, may influence future obesity trends. In Greenland, all healthcare services and prescribed medication are free for all residents. The use of GLP-1 receptor agonists for the treatment of obesity has increased in recent years. However, the current clinical guidelines restrict eligibility to patients with a BMI of 40 kg/m² or above who have an obesity-related comorbidity and have completed a lifestyle intervention without achieving a BMI below 40 kg/m² [30]. The effect of these treatments on population-level obesity prevalence in Greenland, therefore, remains uncertain. Strengths and limitations This study has several strengths. First, the study is based on data from stratified random sampling from the civil registration systems, ensuring that the BMI of individuals is representative of the background population. This sampling method ensures that prevalence estimates reflect population-level distributions rather than healthcare utilisation patterns. Second, the 19-year observation period (1999–2018) captured long-term secular trends across diverse geographic regions, including remote settlements. Third, standardised anthropometric measurements conducted by trained personnel provided objective data, eliminating systematic underreporting associated with self-reported height and weight. Although a previous study demonstrated that the differences between self-reported and measured values of BMI were small and similar across sexes [21], self-reported data typically result in underreporting of weight and overreporting of height [31]. There are also some weaknesses in the present study. First, the cross-sectional design with varying geographical coverage across surveys introduces potential selection effects. Thus, observed temporal trends may partially reflect geographical population differences rather than true secular changes. To mitigate this limitation, we employed multiple logistic regression models adjusting for age, sex and residency to estimate temporal trends whilst controlling for demographic composition differences between surveys. Second, non-response bias may affect the generalisability of prevalence estimates. Nevertheless, the Greenland Population Health Surveys achieved participation rates between 52% and 65% across surveys [12–15], suggesting adequate population representativeness. Third, the main measure of obesity in this study was BMI, which is a composite measure that does not directly reflect the body fat mass [32]. As recently highlighted by the Lancet Diabetes and Endocrinology Commission, BMI provides limited information about the presence of obesity-related organ dysfunction at the individual level [33]. Nevertheless, using BMI to estimate obesity prevalence has several advantages for epidemiological research. The aim of this study was to estimate temporal trends at the population level, not to classify individual clinical risk, and the specific measure of obesity is less critical when estimating trends over time. The widespread use of BMI allows for comparison with other studies, and its straightforward assessment makes it suitable for use in remote areas, such as small settlements. To account for the possible overestimation of obesity, we also assessed the prevalence using WC and WHR. BMI are used to assess general obesity, while WHR and WC specifically measure abdominal fat. Although these measures capture different aspects of body fat distribution, they are indirect measures of obesity as compared to MRI or DXA image-based estimates [34]. The estimated prevalence of adult obesity using the cut-offs for WC and WHR deviated from the prevalence estimates based on BMI (Appendix Table 1). However, the trend of increasing prevalence persisted. Thus, the trend of increasing prevalence of obesity is consistent and independent of the measure of obesity applied. Conclusion This study showed a consistent increase in adult obesity across age, sex and place of residence from 1999 to 2018 in Greenland, indicating that obesity affects all parts of the Greenlandic adult population. The particularly high prevalence among women in all subgroups indicates a significant sex disparity. These findings underscore the need for research on the drivers of the changes in prevalence and thereby inform public health policies that address the rising prevalence of obesity in Greenland, taking into account the population and context. Prevention and treatment of obesity can reduce the risk of developing cardiometabolic diseases, including diabetes and cardiovascular disease, further highlighting the importance of addressing the trends identified in this study. Supplementary Material Supplementary Material Appendix.pdf Acknowledgements The authors thank the participants who contributed to this study and all staff who worked on the Greenland population health surveys. C.V.L., P.B. and M.E.J. designed and conducted the population health surveys that provided the data for this study. H.N., C.L., M.E.J., D.L.C., N.S. and S.B. conceived the study. H.N., M.E.J., D.L.C. and S.B. developed the analysis plan, and H.N. analysed the data and wrote the first draft. All authors provided critical revisions and approved the final manuscript. Claude Sonnet 4.5 (Anthropic) was used for grammar and spelling checks to improve the clarity of the manuscript. Funding Statement The Greenland population health surveys were funded by the Ministry of Health, Government of Greenland, The Novo Nordisk Foundation (NNF 17SH0027192, NNF 17OC0028136 and NNF15OC0016544), The Independent Research Fund Denmark, and Karen Elise Jensen's Foundation. The Novo Nordisk Foundation Centre for Basic Metabolic Research is supported by The Novo Nordisk Foundation (Grant number NNF23SA0084103). The funder(s) had no role in the study design, data collection and analysis, decision to publish, or manuscript preparation. Disclosure statement H.N., C.L., N.K.S., M.E.J. and S.B. were employed at Steno Diabetes Center Greenland (SDCG) and/or Steno Diabetes Center Copenhagen (SDCC) while the study was conducted. Both are public hospitals and research institutions partly funded by a grant from the Novo Nordisk Foundation (NNF20SA0064190). P.B., M.E.J. and S.B. own shares in Novo Nordisk A/S. C.V.L.L., T.I.A.S. and D.L.C. declare no conflicts of interest. Data availability statement Data are temporarily deposited at the Danish National Archives until data deposition facilities become available in Greenland. The data can be made available anonymously for analysis upon reasonable request. Ethics approval statement The study was performed in accordance with the principles of the Helsinki Declaration, and all participants gave written informed consent when participating in the surveys. Further, the study was approved by the Ethics Committee in Greenland and the health authorities (project numbers: B99, 505-42; Inuit Health in Transition, 505-95; B2014, 2014-10; B2018, 2017-05). Supplementary material Supplemental data for this article can be accessed at https://doi.org/10.1080/22423982.2026.2686030. References Associated Data Supplementary Materials Supplementary Material Appendix.pdf Data Availability Statement Data are temporarily deposited at the Danish National Archives until data deposition facilities become available in Greenland. The data can be made available anonymously for analysis upon reasonable request.