Abstract
Type 2 diabetes (T2D) results from the interplay of genetic susceptibility and an unhealthy lifestyle, but their combined effects are not well studied. We examined whether unhealthy modifiable behaviors were associated with similar increases in the risk of incident T2D in individuals with different levels of genetic risk. Among 332,251 UK Biobank participants without diabetes, we constructed a multiancestry genetic risk score (GRS) based on 783 T2D-associated variants, categorized into tertiles. Lifestyle was classified as healthy, intermediate, or unhealthy based on baseline self-reported smoking status, BMI, physical activity level, and diet quality. Cox proportional hazards regression models were used to generate adjusted hazard ratios (HRs) for T2D and associated 95% CIs. During follow-up (median 13.6 years), 13,128 (4.0%) participants developed T2D. GRS (P < 0.001) and lifestyle classification (P < 0.001) were independently associated with increased risk of T2D. Compared with a healthy lifestyle, an unhealthy lifestyle was associated with increased risk in all genetic risk strata, with adjusted HRs ranging from 7.11 to 16.33. High genetic risk and an unhealthy lifestyle were the most significant contributors to T2D development. Individuals at all levels of genetic risk can substantially mitigate their T2D risk through lifestyle modifications.
Article Highlights
Both genetic susceptibility and an unhealthy lifestyle are known to be associated with elevated type 2 diabetes (T2D) risk. However, their combined effects on T2D risk are not well studied.
In this large prospective cohort study of more than 332,000 individuals, unhealthy lifestyle factors were associated with risk of incident T2D within and across different levels of genetic risk.
These findings suggest individuals at all levels of genetic risk can greatly mitigate their risk of T2D by adhering to a healthy lifestyle.
Graphical Abstract
Introduction
Type 2 diabetes (T2D) is caused by a complex interplay between genetic predisposition and lifestyle factors. Genome-wide association studies have identified more than 1,200 independent genetic variants associated with T2D, explaining approximately 20% to 40% of overall heritability (1). Lifestyle factors, such as BMI, physical activity level, diet quality, and smoking status, also play an important role in modulating disease risk (2–5).
Several studies have attempted to examine the potential joint effects between genetic risk and overall behavioral and/or lifestyle factors on the risk of T2D (6–13), demonstrating the strongest disease risk among those with the highest genetic risk and unhealthiest lifestyle. However, prior analyses were predominantly performed among individuals of East Asian (8,9,11,12) or European ancestry (6,7,10,13,14), limiting generalizability to multiancestry and other non-European populations, and differences by sex were not often considered. Additionally, a majority of studies included fewer than 100 genetic variants in calculating genetic risk scores (GRS) (8–14), used effect size weights from a mismatched ancestry population (9,12), or used variants and effect size weights that were not independent of the study population (6,11), all of which limit accuracy in measuring the genetic risk of T2D.
Therefore, the goal of this study was to determine whether unhealthy modifiable health behaviors were associated with similar increases in the risk of incident T2D among individuals with different levels of genetic risk using the most up-to-date list of independent T2D-associated genetic variants across all individuals in the UK Biobank (UKB).
Research Design and Methods
Data Source
The UKB study design and population have been described elsewhere (15–17). Briefly, the UKB is a population-based prospective cohort of more than 500,000 participants from across the U.K. designed to examine environmental, lifestyle, and genetic determinants of adult-onset diseases (16). At enrollment, participants aged 40 to 69 years provided extensive information on their demographics, health, and lifestyle through questionnaires, interviews, and physical assessments. Blood samples were collected for genotyping; genotypes were imputed into a merged UK10K/1000 Genomes phase 3 panel (17). Follow-up of participants is ongoing through linked health records. The UKB study was approved by the North West Multi-centre Research Ethics Committee, and all UKB participants provided written informed consent.
All UKB participants with complete data for relevant variables were eligible for inclusion in this study. Participants were excluded if they had a mismatch between their genetic and self-reported sex (n = 372), were missing genotypes for 7.5% or more of the included variants (n = 498), or had a BMI less than 18.5 kg/m2 (n = 2,505) or any kind of diabetes (n = 15,250) at enrollment. To limit our analyses to those who developed incident T2D, we further excluded individuals who developed incident type 1 (ICD-10 E10), malnutrition-related (ICD-10 E12), or other specified (ICD-10 E13) diabetes (n = 699) during follow-up. Our final sample size was 332,251 participants (Fig. 1).
Figure 1.
Flowchart of study population inclusion and exclusion. UKB, UK Biobank.
Exposure and Outcome Ascertainment and Measurement
To estimate genetic predisposition to T2D, GRS were created following an additive model using 783 genome-wide significant variants identified from the most recent multiancestry genome-wide association study meta-analysis (Supplementary Table 1), after excluding results from the UKB to avoid potential effect overestimation (18). GRS were calculated using a weighted method whereby the number of T2D risk–increasing alleles a person has at each variant is multiplied by the effect size estimate from the multiancestry fixed-effects meta-analysis; the products for each variant are then summed for each individual. GRS were analyzed as continuous and divided into tertiles (low, moderate, and high genetic risk groups). We also generated ancestry-specific GRS using the global ancestry group effect size estimates that most closely matched an individual’s self-reported ethnic background; individuals who self-reported their ethnic background as mixed, other, or missing were omitted from the ancestry-specific analyses.
Lifestyle factors (smoking status, BMI, physical activity level, and diet quality) were categorized (healthy, intermediate, or unhealthy) (Supplementary Tables 2 and 3) following the American Heart Association 2020 Strategic Impact Goal guidelines (19,20). Based on these categories, we assigned participants to an overall lifestyle category: healthy (three or more healthy lifestyle factors), unhealthy (three or more unhealthy lifestyle factors), or intermediate (all other combinations).
Incident cases of T2D were identified using first occurrence data, which indicate the first occurrence of any disease from primary care, hospital inpatient, death register, and self-reported data (16). For this study, we used the first record of ICD-10 diagnosis code E11 (i.e., T2D) and the corresponding date to define incident cases.
Descriptive statistics for participants were generated using baseline data and compared between censored observations and incident T2D cases using t tests for continuous variables and χ2 tests for categorical variables. Multivariable Cox regression models were used to calculate hazard ratios (HRs) and associated 95% CIs testing both the independent (both genetic risk and lifestyle as predictor variables) and joint associations of genetic risk and lifestyle with incident T2D; individuals with low GRS and a healthy lifestyle comprised the reference group. Follow-up time was calculated from enrollment until diagnosis of T2D, death, loss to follow-up, or censoring date, whichever came first. The proportional hazards assumptions were tested based on visualization of the survival probabilities over time and the scaled Schoenfeld residuals, and the assumptions were met (Supplementary Figs. 1–10). Adjusted models included the following covariates: age at baseline, biological sex, years of education (21), Townsend Deprivation Index (22), income, and the first 16 genetic principal components (23). All statistical tests, conducted using R 4.3.0, were two sided, and P values <0.05 were considered statistically significant.
Data and Resource Availability
All data analyzed are available through the UKB.
Results
Baseline characteristics of study participants are listed in Table 1. During a median follow-up of 13.56 (interquartile range 12.74, 14.25) years, 13,128 (4%) participants developed incident T2D, with a median time to onset of 7.98 years and the highest incidence rates occurring among those in the high GRS tertile and unhealthy lifestyle classifications (Table 2 and Supplementary Fig. 11 and Supplementary Tables 4 and 5).
Table 1.
Baseline characteristics of participants from UKB
| Variable | Overall* (N = 332,251) | Censored observations* (n = 319,123) | Incident cases* (n = 13,128) |
|---|---|---|---|
| Age at baseline, mean (SD), years | 55.19 (8.06) | 55.09 (8.06) | 57.56 (7.63) |
| Female sex | 177,869 (54) | 172,507 (54) | 5,362 (41) |
| Follow-up time, median (IQR), years | 13.56 (12.74, 14.25) | 13.61 (12.88, 14.27) | 7.98 (4.92, 10.82) |
| Multiancestry GRS, mean (SD) | 21.54 (0.56) | 21.53 (0.56) | 21.76 (0.55) |
| Multiancestry GRS tertile | |||
| Low | 110,753 (33) | 108,130 (34) | 2,623 (20) |
| Moderate | 110,755 (33) | 106,670 (33) | 4,085 (31) |
| High | 110,743 (33) | 104,323 (33) | 6,420 (49) |
| BMI, kg/m2 | |||
| Healthy (18.5–24.9) | 118,246 (36) | 117,038 (37) | 1,208 (9.2) |
| Intermediate (25.0–29.9) | 143,047 (43) | 138,249 (43) | 4,798 (37) |
| Unhealthy (≥30) | 70,958 (21) | 63,836 (20) | 7,122 (54) |
| Smoking status | |||
| Healthy (nonsmoker) | 188,546 (57) | 182,341 (57) | 6,205 (47) |
| Intermediate (past) | 111,190 (34) | 106,143 (33) | 5,047 (39) |
| Unhealthy (current) | 31,959 (9.6) | 30,104 (9.4) | 1,855 (14) |
| Physical activity level† | |||
| Healthy (regular) | 93,866 (29) | 90,532 (29) | 3,334 (27) |
| Intermediate (some) | 192,820 (60) | 185,844 (60) | 6,976 (57) |
| Unhealthy (none) | 33,588 (10) | 31,599 (10) | 1,989 (16) |
| Diet quality‡ | |||
| Healthy (adequate) | 159,553 (48) | 153,928 (48) | 5,625 (43) |
| Unhealthy (inadequate) | 172,686 (52) | 165,183 (52) | 7,503 (57) |
| Overall lifestyle | |||
| Healthy (≥3 healthy factors) | 70,854 (21) | 69,669 (22) | 1,185 (9.0) |
| Intermediate (all other combinations) | 248,927 (75) | 238,434 (75) | 10,493 (80) |
| Unhealthy (≥3 unhealthy factors) | 12,470 (3.8) | 11,020 (3.5) | 1,450 (11) |
| Years of education, mean (SD) | 15.54 (4.44) | 15.56 (4.44) | 14.97 (4.38) |
| Income, pound sterling | |||
| <18,000 | 54,537 (16) | 51,007 (16) | 3,530 (27) |
| 18,000–30,999 | 82,294 (25) | 78,494 (25) | 3,800 (29) |
| 31,000–51999 | 95,351 (29) | 91,947 (29) | 3,404 (26) |
| 52,000–100,000 | 78,705 (24) | 76,717 (24) | 1,988 (15) |
| >100,000 | 21,364 (6.4) | 20,958 (6.6) | 406 (3.1) |
| Self-reported ancestry | |||
| African or African American | 4,385 (1.3) | 3,963 (1.2) | 422 (3.2) |
| East Asian | 970 (0.3) | 914 (0.3) | 56 (0.4) |
| European | 316,747 (95) | 304,982 (96) | 11,765 (90) |
| Missing | 840 (0.3) | 796 (0.2) | 44 (0.3) |
| Mixed | 1,984 (0.6) | 1,888 (0.6) | 96 (0.7) |
| Other | 2,494 (0.8) | 2,299 (0.7) | 195 (1.5) |
| South Asian | 4,831 (1.5) | 4,281 (1.3) | 550 (4.2) |
Data are n (%) unless otherwise indicated. IQR, interquartile range.
*All P values <0.001.
†Healthy, ≥150 min/week moderate or ≥75 min/week vigorous or ≥150 min/week mixed; intermediate, 1–149 min/week moderate or 1–74 min/week vigorous or 1–149 min/week mixed; unhealthy, not performing any moderate or vigorous.
‡Healthy, adequate intake of at least half of certain dietary components; unhealthy, intake of fewer than half.
Table 2.
Incidence rates of T2D by category of combined genetic and lifestyle risk
| Multiancestry GRS tertile | Lifestyle | N (%) | Person-years | Incidence rate† | |
|---|---|---|---|---|---|
| Total | Cases* | ||||
| Low | Healthy | 24,408 (7.4) | 245 (1.0) | 325,304.3 | 0.75 |
| Intermediate | 82,402 (24.8) | 2,082 (2.5) | 1,083,869.8 | 1.92 | |
| Unhealthy | 3,943 (1.2) | 296 (7.5) | 50,042.0 | 5.92 | |
| Moderate | Healthy | 23,373 (7.0) | 347 (1.5) | 311,018.5 | 1.12 |
| Intermediate | 83,224 (25.1) | 3,250 (3.9) | 1,088,448.2 | 2.99 | |
| Unhealthy | 4,158 (1.3) | 488 (11.7) | 51,563.5 | 9.46 | |
| High | Healthy | 23,073 (6.9) | 593 (2.6) | 305,394.7 | 1.94 |
| Intermediate | 83,301 (25.1) | 5,161 (6.2) | 1,078,391.9 | 4.79 | |
| Unhealthy | 4,369 (1.3) | 666 (15.2) | 53,161.2 | 12.53 | |
*Cumulative incidence.
†Per 1,000 person-years.
GRS tertiles (moderate, HR 1.59 [95% CI 1.52–1.67]; high, HR 2.58 [95% CI 2.47–2.70]) and overall lifestyle categories (intermediate, HR 2.38 [95% CI 2.24–2.52]; unhealthy, HR 6.83 [95% CI 6.32–7.38]) were independently associated with T2D risk (Supplementary Tables 6 and 7). For the standardized GRS, a 1-SD increase was associated with a 53% increased risk of T2D (95% CI 1.50–1.55) (Supplementary Table 8). Results were similar for individual lifestyle factors and in models stratified by sex (Supplementary Tables 9 and 10); BMI had the strongest independent association with incident T2D (intermediate, HR 2.81 [95% CI 2.63–3.00]; unhealthy, HR 8.84 [95% CI 8.29–9.42]).
Across all GRS tertiles, individuals with an intermediate or unhealthy lifestyle were at substantially increased risk of T2D compared with those with a healthy lifestyle (Fig. 2 and Supplementary Tables 11–13 and Supplementary Figs. 12 and 13). Within a single genetic risk tertile (e.g., low GRS), individuals in the intermediate and unhealthy lifestyle category were at increased risk of T2D compared with those with a healthy lifestyle. Results were similar in sex-stratified analyses. We did not detect a significant interaction between GRS tertile and lifestyle classification; however, we did detect a significant interaction when considering the standardized GRS and lifestyle classifications (Supplementary Tables 14 and 15). Inclusion of both the GRS and lifestyle classifications improved T2D prediction beyond each exposure individually (Supplementary Table 16). Results showed similar trends across self-reported ancestry groups but were generally underpowered for analyses of individuals of non-European ancestry (Supplementary Table 17).
Figure 2.
Association of combined genetic (multiancestry) and lifestyle risk of T2D.
Similar trends in the associations between combined GRS and lifestyle risk with incident T2D were found in analyses using ancestry-specific GRS (Supplementary Tables 18–25 and Supplementary Figs. 14–18).
To evaluate the proportion of cases of incident T2D that would have been prevented if individuals with an intermediate or unhealthy lifestyle (also considered nonhealthy) instead had a healthy lifestyle, we calculated the population attributable fraction (Supplementary Table 26). Regardless of GRS, more than 55% of incident T2D cases in the UKB would have been prevented if all individuals in the nonhealthy lifestyle categories had been in the healthy lifestyle category (year 1, 95% CI 0.53–0.58). The population attributable fraction proportions were consistent across all time points during the 15-year follow-up.
Discussion
In this large population-based prospective cohort study, high GRS and an unhealthy lifestyle were independently associated with increased risk of T2D. Across and within different GRS tertiles, adherence to an intermediate or unhealthy lifestyle was associated with substantially increased risk of T2D compared with a healthy lifestyle. Overall, our analyses support the notion that although genetics play a large role in the risk of developing T2D, lifestyle factors play a substantially larger role, particularly BMI. Furthermore, we demonstrated that individuals with any level of genetic risk could greatly reduce their disease risk through modifiable healthy lifestyle behaviors.
To our knowledge, this study is the first to test the effect of combined lifestyle factors in different genetic risk levels for T2D based on nearly 800 genetic variants and the first to consider both multiancestry and ancestry-matched GRS. Consistent with findings from prior studies (6–14), we found high GRS and unhealthy lifestyle factors were independently and jointly associated with increased risk of developing T2D. However, there is wide variability in effect sizes across the prior studies, most likely because of fundamental differences in study design and methodology. Most similar to our study, the study by Said et al. (14) found strong effects of an unhealthy lifestyle across different GRS tertiles in the UKB, with adjusted HRs ranging from 10.82 to 15.46 in sex-combined analyses. Although both studies used a similar approach to categorize lifestyle factors, Said et al. (14) included only individuals of European ancestry, had a less restrictive definition of incident T2D that likely resulted in outcome misclassification, and calculated the GRS based on only 38 variants. Our study improved on these prior limitations. Additional strengths of our study include the prospective cohort design, large sample size, comprehensive measure of genetic risk, and American Heart Association–defined lifestyle classification, which allows for a more direct clinical interpretation of our results.
There are also several limitations to note. Measurements for all lifestyle factors were obtained at enrollment, of which three were based on self-reported data. Because they are all potentially time-varying covariates, misclassification of exposure is possible. However, because of the prospective design of the UKB, any misclassification would be nondifferential and would result in an underestimation of the true association. Second, incident T2D cases were identified using first occurrence data in the UKB, which includes self-reported outcomes. Furthermore, the suspected rate of undiagnosed T2D in the U.K. is estimated to be approximately 2% (24). Therefore, misclassification of some T2D cases as noncases is possible; however, we would expect this to bias the results toward the null. The genetic variants used in the GRS calculation may also have pleiotropic effects on lifestyle factors, including BMI; this may have resulted in an underestimation of the true effects (Supplementary Table 27). Although our study included individuals with diverse ethnic backgrounds, the generalizability of the findings remains somewhat limited because of the predominance of European participants in the UKB. Generalizability may also be limited because UKB participants are generally regarded as healthier than the general population.
Based on our results, it is clear that individuals who have either moderate or higher genetic risk of T2D with an intermediate or unhealthy lifestyle are at substantially increased risk of T2D. These findings indicate the strong potential benefits of adherence to multiple healthy lifestyle factors to mitigate disease risk, regardless of genetic risk. In fact, our analysis suggests that 55% of the incident T2D risk in the population could be theoretically eliminated if individuals with a nonhealthy lifestyle instead had a healthy lifestyle, highlighting the potential impact of individuals shifting from the nonhealthy to healthy lifestyle category. Although challenges remain in communicating individual genetic risk information to patients that is understandable and interpretable by the general population, knowledge of the strong impact a healthy lifestyle can have in mitigating genetic or familial risk of T2D may motivate patients to change behaviors.
This article contains supplementary material online at https://doi.org/10.2337/figshare.31170646.
Article Information
Acknowledgments. This research was conducted using the UKB under application 81009.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. C.Z., R.B., E.B.-J., A.P., S.R.S., N.V., C.H.L., and C.N.S. were involved in the conception, design, and conduct of the study. C.Z. and C.N.S. wrote the first draft of the manuscript. All authors were involved in the analysis and interpretation of results and edited, reviewed, and approved the final version of the manuscript. C.N.S. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Prior Presentation. This study was presented at the 84th Scientific Sessions of the American Diabetes Association, Orlando, FL, 21–24 June 2024, and the Annual Meeting of the American Society for Human Genetics, Denver, CO, 5–9 November 2024.
Funding Statement
This work was supported by National Institutes of Health grants R01DK118011 (C.N.S. and C.Z.), R01DK136671 (C.N.S.), R34MH118396 and K01DK123193 (E.B.-J.), and K01DK123193 (N.V.); American Diabetes Association grants 11-22-JDFPM-06 (C.N.S. and C.Z.) and 11-23-PDF-35 (A.H.-C.); and Commonwealth of Massachusetts grant INTF4107H78241733154 (E.B.-J.).
Supporting information
References
- 1. Prasad RB, Groop L.. Genetics of type 2 diabetes—pitfalls and possibilities. Genes (Basel) 2015;6:87–123 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Ye C, Niu J, Zhao Z, et al. Genetic susceptibility, family history of diabetes and healthy lifestyle factors in relation to diabetes: a gene-environment interaction analysis in Chinese adults. J Diabetes Investig 2021;12:2089–2098 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Langenberg C, Sharp SJ, Franks PW, et al. Gene-lifestyle interaction and type 2 diabetes: the EPIC interact case-cohort study. PLoS Med 2014;11:e1001647. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Klimentidis YC, Chen Z, Arora A, Hsu C-H.. Association of physical activity with lower type 2 diabetes incidence is weaker among individuals at high genetic risk. Diabetologia 2014;57:2530–2534 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Lim JE, Kang J-O, Ha T-W, et al. Gene-environment interaction in type 2 diabetes in Korean cohorts: interaction of a type 2 diabetes polygenic risk score with triglyceride and cholesterol on fasting glucose levels. Genet Epidemiol 2022;46:285–302 [DOI] [PubMed] [Google Scholar]
- 6. Zhao Y, Li Y, Zhuang Z, et al. Associations of polysocial risk score, lifestyle and genetic factors with incident type 2 diabetes: a prospective cohort study. Diabetologia 2022;65:2056–2065 [DOI] [PubMed] [Google Scholar]
- 7. Schnurr TM, Jakupović H, Carrasquilla GD, et al. Obesity, unfavourable lifestyle and genetic risk of type 2 diabetes: a case-cohort study. Diabetologia 2020;63:1324–1332 [DOI] [PubMed] [Google Scholar]
- 8. Yu H, Wang T, Zhang R, et al. Alcohol consumption and its interaction with genetic variants are strongly associated with the risk of type 2 diabetes: a prospective cohort study. Nutr Metab (Lond) 2019;16:64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Villegas R, Delahanty R, Gao Y-T, et al. Joint effect of genetic and lifestyle risk factors on type 2 diabetes risk among Chinese men and women. PloS One 2012;7:e49464. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Szczerbiński Ł, Gościk J, Bauer W, et al. Efficacy of family history, genetic risk score, and physical activity in assessing the prevalence of type 2 diabetes. Pol Arch Intern Med 2019;129:442–450 [DOI] [PubMed] [Google Scholar]
- 11. Li H, Khor C-C, Fan J, et al. Genetic risk, adherence to a healthy lifestyle, and type 2 diabetes risk among 550,000 Chinese adults: results from 2 independent Asian cohorts. Am J Clin Nutr 2020;111:698–707 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Han X, Wei Y, Hu H, et al. Genetic risk, a healthy lifestyle, and type 2 diabetes: the Dongfeng-Tongji cohort study. J Clin Endocrinol Metab 2020;105:dgz325. [DOI] [PubMed] [Google Scholar]
- 13. Ericson U, Hindy G, Drake I, et al. Dietary and genetic risk scores and incidence of type 2 diabetes. Genes Nutr 2018;13:13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Said MA, Verweij N, van der Harst P.. Associations of combined genetic and lifestyle risks with incident cardiovascular disease and diabetes in the UK Biobank study. JAMA Cardiol 2018;3:693–702 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Allen NE, Lacey B, Lawlor DA, et al. Prospective study design and data analysis in UK Biobank. Sci Transl Med 2024;16:eadf4428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Sudlow C, Gallacher J, Allen N, et al. UK Biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 2015;12:e1001779. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Bycroft C, Freeman C, Petkova D, et al. The UK Biobank resource with deep phenotyping and genomic data. Nature 2018;562:203–209 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Suzuki K, Hatzikotoulas K, Southam L, et al.; VA Million Veteran Program . Genetic drivers of heterogeneity in type 2 diabetes pathophysiology. Nature 2024;627:347–357 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Mozaffarian D. Dietary and policy priorities for cardiovascular disease, diabetes, and obesity: a comprehensive review. Circulation 2016;133:187–225 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Lloyd-Jones DM, Hong Y, Labarthe D, et al.; American Heart Association Strategic Planning Task Force and Statistics Committee . Defining and setting national goals for cardiovascular health promotion and disease reduction: the American Heart Association’s strategic impact goal through 2020 and beyond. Circulation 2010;121:586–613 [DOI] [PubMed] [Google Scholar]
- 21. UNESCO Institute for Statistics. International Standard Classification of Education: ISCED 2011. Accessed 15 July 2024. Available from https://www.uis.unesco.org/en/methods-and-tools/isced
- 22. Townsend P. Deprivation. J Soc Pol 1987;16:125–146 [Google Scholar]
- 23. Privé F, Luu K, Blum MGB, McGrath JJ, Vilhjálmsson BJ.. Efficient toolkit implementing best practices for principal component analysis of population genetic data. Bioinformatics 2020;36:4449–4457 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Whicher CA, O’Neill S, Holt RIG.. Diabetes in the UK: 2019. Diabet Med J Med 2020;37:242–247 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.