Obesity, unfavourable lifestyle and genetic risk of type 2 diabetes: a case-cohort study

Abstract

Aims/hypothesis

We aimed to investigate whether the impact of obesity and unfavourable lifestyle on type 2 diabetes risk is accentuated by genetic predisposition.

Methods

We examined the joint association of genetic predisposition, obesity and unfavourable lifestyle with incident type 2 diabetes using a case-cohort study nested within the Diet, Cancer and Health cohort in Denmark. The study sample included 4729 individuals who developed type 2 diabetes during a median 14.7 years of follow-up, and a randomly selected cohort sample of 5402 individuals. Genetic predisposition was quantified using a genetic risk score (GRS) comprising 193 known type 2 diabetes-associated loci (excluding known BMI loci) and stratified into low (quintile 1), intermediate and high (quintile 5) genetic risk groups. Lifestyle was assessed by a lifestyle score composed of smoking, alcohol consumption, physical activity and diet. We used Prentice-weighted Cox proportional-hazards models to test the associations of the GRS, obesity and lifestyle score with incident type 2 diabetes, as well as the interactions of the GRS with obesity and unfavourable lifestyle in relation to incident type 2 diabetes.

Results

Obesity (BMI ≥ 30 kg/m2) and unfavourable lifestyle were associated with higher risk for incident type 2 diabetes regardless of genetic predisposition (p > 0.05 for GRS–obesity and GRS–lifestyle interaction). The effect of obesity on type 2 diabetes risk (HR 5.81 [95% CI 5.16, 6.55]) was high, whereas the effects of high genetic risk (HR 2.00 [95% CI 1.76, 2.27]) and unfavourable lifestyle (HR 1.18 [95% CI 1.06, 1.30]) were relatively modest. Even among individuals with low GRS and favourable lifestyle, obesity was associated with a >8-fold risk of type 2 diabetes compared with normal-weight individuals in the same GRS and lifestyle stratum.

Conclusions/interpretation

Having normal body weight is crucial in the prevention of type 2 diabetes, regardless of genetic predisposition.

figurea

Introduction

Type 2 diabetes is a common disease with a rapidly increasing global prevalence that has been largely attributed to the ongoing pandemic of obesity and a sedentary lifestyle [1,2,3]. Public health strategies to prevent type 2 diabetes focus on weight management and promotion of healthy lifestyles [4, 5]. Lifestyle interventions designed for weight loss through intensive lifestyle counselling have been shown to delay the onset of type 2 diabetes among individuals with impaired glucose tolerance [6, 7]. However, the effects of lifestyle behaviours and weight loss on type 2 diabetes risk may vary between individuals depending on genetic variation [6, 7]. To understand the role of genetic variation in the prevention of type 2 diabetes, it is important to elucidate the interaction between genetic predisposition, obesity and lifestyle behaviours.

Recently, a case-cohort study that examined the joint association of genetic predisposition and unfavourable lifestyle with incident type 2 diabetes in the UK Biobank found that type 2 diabetes risk is increased by >10-fold in individuals with an unfavourable lifestyle, independent of genetic risk [8]. However, unfavourable lifestyle was defined using a multifactorial score that considered obesity as a risk factor equal to smoking behaviour, physical activity and diet. Hence, the study did not allow distinguishing between the effects of obesity and other alterable lifestyle factors in the development of type 2 diabetes. Furthermore, lifestyle recommendations designed for the prevention of cardiovascular disease were applied rather than type 2 diabetes-specific recommendations, and genetic risk was estimated using 38 known type 2 diabetes risk variants rather than ~200 risk variants identified in the most recent genome-wide association study (GWAS) [9].

In the present study, we address these shortcomings by examining the joint association of genetic risk, obesity and unfavourable lifestyle with incident type 2 diabetes within a case-cohort sample of 4729 cases and a randomly selected subcohort of 5402 individuals from Denmark.

Methods

Study population

The present case-cohort study is nested within the Danish Diet, Cancer and Health cohort established in 1993–1997 [10]. Altogether 160,725 native Danish citizens living in the urban areas of Copenhagen and Aarhus were invited to participate, and 57,053 eligible participants without a previous cancer diagnosis were finally recruited. The study complies with the Declaration of Helsinki and has been approved by the Health Research Ethics, Capital Region of Denmark and the Danish Data Protection Agency. All participants gave written informed consent at inclusion.

The participants were followed up from baseline until the date of the diagnosis of diabetes, death, emigration, date of change of personal identification number, or the end of follow-up on 31 December 2011, whichever came first. Incident cases of type 2 diabetes were identified through the National Diabetes Register, where participants were defined as diabetic if they met one of the following criteria [11,12,13,14]: registration with a diabetes diagnosis in the National Patient Register; registration of chiropody (as a diabetic patient); five blood glucose measurements in a one year period or two blood glucose measurements per year in five consecutive years; purchase of prescribed insulin recorded in the National Health Service Register; or purchase of oral glucose-lowering drugs in the Danish National Prescription Registry (electronic supplementary material [ESM] Table 1). Overall, the positive predictive value for the identification of diabetes cases using the criteria was 89%, and the sensitivity was 86% [14].

We genotyped a total of 4771 individuals with incident diabetes diagnosis by 31 December 2011 and a randomly drawn subcohort of 5655 individuals who also included 689 individuals who developed diabetes (due to the random selection). The genotyping was performed using the Illumina Human Core Exome BeadChip (Illumina, San Diego, CA, USA) and genotypes were called using the Illumina BeadStudio algorithm. The genotype data were imputed to the Haplotype Reference Consortium (HRC) reference panel 1.1 on the Michigan server using Minimac3.

We excluded closely related individuals, samples with extreme inbreeding coefficients, mislabelled sex or call rate <95%, duplicates, and individuals identified as ethnic outliers based on genome-wide principal component analysis. As indicated in ESM Fig. 1, we also excluded individuals with type 2 diabetes at baseline and participants that had missing information on lifestyle behaviour or covariates. After the exclusions, a total of 4729 incident type 2 diabetes cases and a randomly selected subcohort of 5402 individuals, of whom 575 developed incident type 2 diabetes, remained in the study population. The median follow-up time was 14.7 years (interquartile range 9.5–15.7 years).

Genetic risk score construction

The SNPs included in the genetic risk score (GRS) were selected based on the most recent DIAMANTE consortium meta-analysis of 32 European-descent GWAS including 74,124 individuals with type 2 diabetes and 824,006 control participants [9]. Of the 231 SNPs that reached the genome-wide significance threshold (p < 5× 10−8) in the BMI-unadjusted primary discovery analysis [9], we included 218 loci with minor allele frequency (MAF) >1% and good imputation quality (INFO > 0.7) in the Diet, Cancer and Health cohort. We excluded all genetic variants (n = 25) that had a genome-wide significant association with BMI or were in strong linkage disequilibrium (r2 > 0.8 in 1000 Genomes European panel) with a BMI locus in the largest and most recently published GWAS for BMI [15], except for the strongest known type 2 diabetes risk locus in TCF7L2 [16]. We constructed a weighted type 2 diabetes risk-increasing GRS consisting of the remaining 193 loci by summing the number of type 2 diabetes-increasing alleles weighted by the OR of the selected SNPs estimated in the discovery GWAS [9]. Information on the SNPs included 25 in the GRS, the risk alleles, risk-allele frequencies, imputation INFO scores, and respective ORs that were used as weights for the calculation of the GRS are presented in ESM Table 2. The GRS was stratified into low (lowest 20%), intermediate and high risk (top 20%) groups.

Assessment of body weight

BMI was calculated from weight and height measured at baseline by dividing body weight in kg by height in metres squared. We defined individuals as normal weight, overweight and obese if they had a BMI <25 kg/m2, ≥25 – <30 kg/m2 and ≥ 30 kg/m2, respectively.

Assessment of lifestyle

We adapted lifestyle intervention guidelines established by an initiative of the Danish Ministry of Health [4] and recommended for type 2 diabetes prevention by the Danish Diabetes Association, to create a multifactorial lifestyle score for type 2 diabetes risk (Table 1). Lifestyle was assessed by first defining dichotomous variables (0/1) for adherence to regular physical activity, healthy dietary pattern, smoking status and moderate alcohol consumption, as described in detail below, where 1 point was given for adherence to each favourable lifestyle behaviour. The points were subsequently summed to calculate a lifestyle score for the combined adherence to favourable lifestyle behaviours, where unfavourable lifestyle was defined as adherence to none or one favourable lifestyle behaviour, intermediate lifestyle as adherence to two favourable lifestyle behaviours, and favourable lifestyle as adherence to three or four favourable lifestyle behaviours.

Table 1 Baseline characteristics of the overall study population and type 2 diabetes cases comprising the 9556 participants included in the present analysis of the Danish Diet, Cancer and Health cohort

Physical activity

Physical activity was assessed by a previously validated questionnaire [17] that includes questions on walking, cycling, housework, sports, do-it-yourself activities and gardening. As described previously [18], each type of physical activity was assigned a metabolic equivalent of task (MET) estimate according to the compendium of physical activities [19], which allows estimation and classification of the energy costs of different activities based on their rate of energy expenditure. MET min/week were calculated from the median MET of physical activity performed in the summer and winter and multiplied by the number of minutes spent in the activity per week. Individuals were defined as having low or high physical activity levels based on the median 55.0 MET min/week in the study sample.

Dietary pattern

The participants completed a 192-item Food Frequency Questionnaire at baseline [20,21,22]. The intake of specific foods and nutrients was calculated by the FoodCalc software [23]. We adapted the Danish official food-based dietary guidelines that are based on the Nordic Nutrition Recommendations [24] and referred to by the Danish Diabetes Association [25], to calculate a healthy diet index comprising seven items (ESM Table 3). One of the original recommendations, concerning sodium intake, was excluded in the present analyses, due to lack of data on sodium intake in the Diet, Cancer and Health cohort. One point was given for reaching the recommended consumption of fruits and vegetables, whole grains, fish, low-fat (maximum fat content of 1.5%) dairy products, as well as for recommended low consumption of red meat (including unprocessed and processed meat and meat products), saturated fat, and sugar-sweetened beverages or juices (including fruit and vegetable juice). The points were summed to assess adherence to healthy diet (range 0 [no adherence] to 7 [highest adherence]), and a dichotomous healthy diet variable was formed by defining low adherence as 0 to 2 points and middle to high adherence as 3 to 7 points (Table 1). The cut-off for the healthy diet index was set at three out of seven items to ensure sufficient statistical power, as this value was closest to the median of the study population and thus balanced the sample sizes of the two strata.

Alcohol consumption, smoking and educational level

Alcohol consumption was calculated from the 192-item Food Frequency Questionnaire. Based on current Nordic Nutrition Recommendations [24], moderate alcohol consumption was defined as ≤6 units/week for women and ≤12 units/week for men, and high alcohol consumption as >6 units/week for women and >12 units/week for men, where 1 unit is equivalent to 12 g of pure alcohol. Information on smoking habit (current smoker, non-smoker) and educational level defined by its duration (≤7, 8–9, and ≥10 years) [26, 27] was obtained from the baseline questionnaire.

Statistical analysis

We used Prentice-weighted Cox proportional-hazards models using the ‘cch’ command integrated into the R package ‘survival’ to test the associations of GRS, obesity and lifestyle score with incident type 2 diabetes, as well as the interactions of the GRS with obesity and unfavourable lifestyle in relation to incident type 2 diabetes. Consistent with previous studies [8, 28], we determined whether participants had a low (quintile 1), intermediate (quintiles 2 to 4) or high (quintile 5) genetic risk for type 2 diabetes and a favourable (3 or 4 favourable lifestyle factors), intermediate (2 favourable lifestyle factors) or an unfavourable lifestyle (0 or 1 favourable lifestyle factors). All analyses were adjusted for age at baseline, sex and educational level. The analyses including the GRS were additionally adjusted for the first three genome-wide principal components to correct for population stratification. Analyses including the GRS and lifestyle score were performed with and without adjustment for BMI. All analyses were performed using RStudio software, version 3.3.1 (2016-06-21; Boston, MA, USA).

Power calculations and assessment of predictive utility of the models

We performed power calculations through 2000 simulation analyses using the simplified approximating setting of logistic regression, hence modelling the odds of incident diabetes (being a case vs a non-case) during the follow-up. The GRS*BMI setting was mimicked using estimated variables from a corresponding fitted model while adjusting for sex and age. We derived the least detectable interaction effect based on the current sample size (n = 9556), significance level 5% and 80% power. The power calculations were performed in Stata 15.1 (StataCorp, College Station, TX, USA; www.stata.com). We also assessed the predictive performance of BMI, the lifestyle score and the genetic risk score for incident type 2 diabetes by performing a receiver operating characteristic (ROC) analysis (ESM Table 4).

Results

Population characteristics

The baseline characteristics of the 4726 individuals with type 2 diabetes and the randomly selected subcohort of 5402 participants included in the presented analyses are provided in Table 1. The mean age of all participants was 56.1 years (range 50–65) and 49.6% were women. Overall, 40.0% of all participants had a favourable lifestyle, 34.6% had an intermediate lifestyle and 25.4% had an unfavourable lifestyle, and 21.8% were classified as obese, 43.0% as overweight and 35.2% as having normal weight.

Associations of the GRS, obesity and lifestyle with incident type 2 diabetes

As provided in ESM Table 5, participants with high or intermediate genetic risk had an HR of 2.00 (95% CI 1.76, 2.27) or an HR of 1.49 (95% CI 1.34, 1.66) for incident type 2 diabetes, respectively, compared with individuals with low genetic risk. Participants with an unfavourable or intermediate lifestyle had an HR of 1.18 (95% CI 1.06, 1.30) or an HR of 1.10 (95% CI 1.00, 1.20) for incident type 2 diabetes, respectively, compared with participants with a favourable lifestyle. Individuals who were overweight or obese had an HR of 2.37 (95% CI 2.15, 2.62) or an HR of 5.81 (95% CI 5.16, 6.55) for higher risk of incident type 2 diabetes, respectively, than individuals with normal body weight. In sensitivity analyses that excluded 62 individuals who were underweight (BMI < 18.5 kg/m2), the associations of being overweight and obese with type 2 diabetes risk remained virtually unchanged (HR 2.38 [95% CI 2.16, 2.62] and HR 5.82 [95% CI 5.17, 6.56], respectively).

The associations of high genetic risk and unfavourable lifestyle with type 2 diabetes risk remained similar after adjustment for BMI (HR 2.15 [95% CI 1.85, 2.50] and HR 1.29 [95% CI 1.15, 1.43], respectively), suggesting that the associations of GRS and unfavourable lifestyle with type 2 diabetes were not mediated by BMI.

Interaction of the GRS with obesity and lifestyle on incident type 2 diabetes

There was no significant interaction between the GRS and continuous BMI (p = 0.35) or the lifestyle score (p = 0.72) on incident type 2 diabetes (ESM Table 6); higher BMI and unfavourable lifestyle were similarly associated with increased risk of type 2 diabetes across all genetic risk groups (Figs 1 and 2). We also did not find significant interactions between the GRS and the four individual lifestyle components comprising the lifestyle score on the risk of type 2 diabetes (ESM Table 7). In a sensitivity analysis excluding 62 individuals who were underweight (BMI < 18.5 kg/m2), the interaction between the GRS and continuous BMI on incident type 2 diabetes (p = 0.49) remained virtually unchanged (ESM Fig. 2).

Fig. 1
figure1

Associations of GRS and lifestyle with incident type 2 diabetes. Analyses were adjusted for age, sex, educational level and the first three genome-wide principal components. n, number of incident type 2 diabetes cases

Fig. 2
figure2

Associations of GRS and body-weight status with incident type 2 diabetes. Analyses were adjusted for age, sex, educational level and the first three genome-wide principal components. n, number of incident type 2 diabetes cases

Through power calculation, we estimated that the least detectable GRS*BMI interaction effect that could be detected in the present sample with significance level 0.05 and 80% power is OR = 1.0025.

Combined association of the GRS, obesity and lifestyle with incident type 2 diabetes

Individuals who ranked high for all three risk factors, with obesity, high GRS and unfavourable lifestyle, had an HR of 14.54 (95% CI 8.09, 26.13) for incident type 2 diabetes, compared with normal-weight, individuals low GRS and favourable lifestyle. Notably, even among individuals with low GRS and favourable lifestyle, obesity was strongly associated with higher type 2 diabetes risk with an HR of 8.44 (95% CI 5.43, 13.14) compared with normal weight individuals in the same GRS and lifestyle stratum (Table 2). ROC analyses suggested that the GRS and the lifestyle score have very little predictive utility (AUC = 72.85 and 72.05, respectively) on top of BMI, age and sex (AUC = 71.81).

Table 2 Combined associations of GRS, lifestyle and body-weight status with incident type 2 diabetes

Discussion

In the present study, we found that obesity and unfavourable lifestyle are associated with increased risk of incident type 2 diabetes regardless of genetic risk. We also found that the associations of the GRS and lifestyle score with risk of incident type 2 diabetes are relatively modest compared with the association of obesity with diabetes risk, underscoring the importance of weight management in diabetes prevention.

Our results show that there was no significant interaction between behavioural lifestyle and genetic risk of type 2 diabetes. An unfavourable lifestyle was associated with similar increase in relative type 2 diabetes risk across each stratum of genetic risk. These findings are consistent with previous studies within the InterAct Consortium [29] and the UK Biobank [8]. Overall, the results indicate that a favourable lifestyle should be universally recommended in the prevention of type 2 diabetes, regardless of genetic predisposition, thus supporting current public health guidelines. The study within UK Biobank investigated the effect of a lifestyle score on genetically determined type 2 diabetes risk, finding that a poor lifestyle is associated with a >10-fold increased risk of incident type 2 diabetes, even in individuals with a low genetic risk [8]. However, the study incorporated being overweight/obese as one component of the multifactorial lifestyle score, which did not allow distinguishing between the effects specific to obesity and other lifestyle factors on type 2 diabetes risk. In the present study, we showed that the association with incident type 2 diabetes is dominated by obesity over the impact of unfavourable lifestyle, with a >8-fold risk of type 2 diabetes found for individuals who were obese, despite a favourable lifestyle.

We found no significant interaction between the GRS and BMI in relation to the risk of incident type 2 diabetes. In contrast, a previous analysis within the InterAct Consortium reported that the effect of a GRS for type 2 diabetes (based on 49 established type 2 diabetes loci) on diabetes risk is modified by BMI, such that the effect of the GRS is significantly greater among individuals who are leaner at baseline [29]. It is possible that the present analyses were underpowered to detect an interaction between the GRS and BMI. Alternatively, the inclusion of >150 additional recently identified type 2 diabetes risk variants in the GRS used in the present analyses may have incorporated a larger number of genetic loci that have an equal effect between lean and obese individuals, whereas the earlier discovered loci may have included a relatively larger proportion of variants with primary effects on insulin secretion and a larger effect in lean individuals, due to the exclusion of obese individuals in early GWAS for type 2 diabetes [30]. Indeed, a BMI-stratified analysis, published in 2012, showed a larger OR in normal-weight individuals compared with individuals who were obese for 29 of 36 diabetes loci known at the time [31]. The InterAct Consortium also did not exclude known obesity risk variants when constructing the GRS, which could lead to spurious interactions between the GRS and BMI due to gene–environment dependence [32].

Strengths of the present study are the large number of incident type 2 diabetes cases, retrieved objectively from the Danish Diabetes Registry, and the long follow-up period. Information on lifestyle was collected before type 2 diabetes diagnosis, which ensures that recall bias did not influence the present results. Furthermore, we comprised a lifestyle score specific for type 2 diabetes and took advantage of ~200 loci recently identified to be associated with type 2 diabetes [9]. As a limitation, the present analyses were performed in a population of European genetic ancestry and cannot immediately be generalised to other ancestry groups.

To conclude, we found that individuals with obesity and an unfavourable lifestyle are at greater risk of incident type 2 diabetes regardless of their genetic risk. The results suggest that type 2 diabetes prevention by weight management and healthy lifestyle is critical across all genetic risk groups. Furthermore, we found that the effect of obesity on type 2 diabetes risk is dominant over other risk factors, highlighting the importance of weight management in type 2 diabetes prevention.

Data availability

Due to restrictions related to Danish law and protecting patient privacy, the combined set of data as used in this study can only be made available through a trusted third party, Statistics Denmark. This state organisation holds the data used for this study. University-based Danish scientific organisations can be authorised to work with data within Statistics Denmark and such organisations can provide access to individual scientists inside and outside of Denmark. Requests for data may be sent to Statistics Denmark (www.dst.dk/en/kontakt) or the Danish Data Protection Agency (www.datatilsynet.dk/english).

Abbreviations

GRS:

Genetic risk score

GWAS:

Genome-wide association study

MET:

Metabolic equivalent of task

ROC:

Receiver operating characteristic

References

  1. 1.

    Vos T, Barber RM, Bell B et al (2015) Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet 386(9995):743–800. https://doi.org/10.1016/S0140-6736(15)60692-4

    Article  Google Scholar 

  2. 2.

    Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH (1999) The disease burden associated with overweight and obesity. JAMA 282(16):1523–1529. https://doi.org/10.1001/jama.282.16.1523

  3. 3.

    Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F, Bales VS, Marks JS (2003) Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA 289(1):76–79. https://doi.org/10.1001/jama.289.1.76

  4. 4.

    Christensen AI, Severin M, Holmberg T et al (2009) KRAM-undersøgelsen i tal og billeder, 1st edn. National Institute of Public Health, University of Southern Denmark, p 54 [article in Danish]

  5. 5.

    World Health Organization (2016) Global report on diabetes. Available from: https://apps.who.int/iris/handle/10665/204871. Accessed 23 April 2019

  6. 6.

    Tuomilehto J, Lindstrom J, Eriksson JG et al (2001) Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med 344(18):1343–1350. https://doi.org/10.1056/NEJM200105033441801

    CAS  Article  PubMed  Google Scholar 

  7. 7.

    Knowler WC, Barrett-Connor E, Fowler SE et al (2002) Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 346(6):393–403. https://doi.org/10.1056/NEJMoa012512

    CAS  Article  PubMed  Google Scholar 

  8. 8.

    Said MA, Verweij N, van der Harst P (2018) Associations of combined genetic and lifestyle risks with incident cardiovascular disease and diabetes in the UK Biobank Study. JAMA Cardiol 3(8):693–702. https://doi.org/10.1001/jamacardio.2018.1717

    Article  PubMed  PubMed Central  Google Scholar 

  9. 9.

    Mahajan A, Taliun D, Thurner M et al (2018) Fine-mapping type 2 diabetes loci to single-variant resolution using high-density imputation and islet-specific epigenome maps. Nat Genet 50(11):1505–1513. https://doi.org/10.1038/s41588-018-0241-6

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  10. 10.

    Tjonneland A, Olsen A, Boll K et al (2007) Study design, exposure variables, and socioeconomic determinants of participation in Diet, Cancer and Health: a population-based prospective cohort study of 57,053 men and women in Denmark. Scand J Public Health 35(4):432–441. https://doi.org/10.1080/14034940601047986

    Article  PubMed  Google Scholar 

  11. 11.

    Carstensen B, Kristensen JK, Marcussen MM, Borch-Johnsen K (2011) The National Diabetes Register. Scand J Public Health 39(7 Suppl):58–61. https://doi.org/10.1177/1403494811404278

    Article  PubMed  Google Scholar 

  12. 12.

    Kyro C, Tjonneland A, Overvad K, Olsen A, Landberg R (2018) Higher whole-grain intake is associated with lower risk of type 2 diabetes among middle-aged men and women: the Danish Diet, Cancer, and Health Cohort. J Nutr 148(9):1434–1444. https://doi.org/10.1093/jn/nxy112

    Article  PubMed  Google Scholar 

  13. 13.

    Sørensen M, Andersen ZJ, Nordsborg RB et al (2013) Long-term exposure to road traffic noise and incident diabetes: a cohort study. Environ Health Perspect 121(2):217–222. https://doi.org/10.1289/ehp.1205503

    Article  PubMed  Google Scholar 

  14. 14.

    Kristensen JK, Drivsholm TB, Carstensen B, Steding-Jensen M, Green A (2007) Validation of methods to identify known diabetes on the basis of health registers. Ugeskr Laeger 169(18):1687–1692 [article in Danish]

  15. 15.

    Yengo L, Sidorenko J, Kemper KE et al (2018) Meta-analysis of genome-wide association studies for height and body mass index in approximately 700000 individuals of European ancestry. Hum Mol Genet 27(20):3641–3649. https://doi.org/10.1093/hmg/ddy271

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  16. 16.

    Grant SF, Thorleifsson G, Reynisdottir I et al (2006) Variant of transcription factor 7-like 2 (TCF7L2) gene confers risk of type 2 diabetes. Nat Genet 38(3):320–323. https://doi.org/10.1038/ng1732

    CAS  Article  PubMed  Google Scholar 

  17. 17.

    Wareham NJ, Jakes RW, Rennie KL et al (2003) Validity and repeatability of a simple index derived from the short physical activity questionnaire used in the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Public Health Nutr 6(4):407–413. https://doi.org/10.1079/PHN2002439

    Article  PubMed  Google Scholar 

  18. 18.

    Aadahl M, Zacho M, Linneberg A, Thuesen BH, Jorgensen T (2013) Comparison of the Danish step test and the watt-max test for estimation of maximal oxygen uptake: the Health2008 study. Eur J Prev Cardiol 20(6):1088–1094. https://doi.org/10.1177/2047487312462825

    Article  PubMed  Google Scholar 

  19. 19.

    Ainsworth BE, Haskell WL, Leon AS et al (1993) Compendium of physical activities: classification of energy costs of human physical activities. Med Sci Sports Exerc 25(1):71–80. https://doi.org/10.1249/00005768-199301000-00011

    CAS  Article  PubMed  Google Scholar 

  20. 20.

    Lacoppidan SA, Kyrø C, Loft S et al (2015) Adherence to a healthy Nordic food index is associated with a lower risk of type-2 diabetes—the Danish Diet, Cancer and Health Cohort Study. Nutrients 7(10):8633–8644. https://doi.org/10.3390/nu7105418

    Article  PubMed  PubMed Central  Google Scholar 

  21. 21.

    Overvad K, Tjonneland A, Haraldsdottir J, Ewertz M, Jensen OM (1991) Development of a semiquantitative food frequency questionnaire to assess food, energy and nutrient intake in Denmark. Int J Epidemiol 20(4):900–905. https://doi.org/10.1093/ije/20.4.900

    CAS  Article  PubMed  Google Scholar 

  22. 22.

    Tjonneland A, Overvad K, Haraldsdottir J, Bang S, Ewertz M, Jensen OM (1991) Validation of a semiquantitative food frequency questionnaire developed in Denmark. Int J Epidemiol 20(4):906–912. https://doi.org/10.1093/ije/20.4.906

    CAS  Article  PubMed  Google Scholar 

  23. 23.

    Lauritzen, J. Foodcalc. Available from: www.Ibt.Ku.Dk/jesper/foodcalc/default.Htm. Accessed 23 April 2019

  24. 24.

    Nordic Council of Ministers (2014) Nordic nutrition recommendations 2012: integrating nutrition and physical activity. Available from: https://doi.org/10.6027/Nord2014-002. Accessed 23 April 2019

  25. 25.

    Fødevarestyrelsen De officielle kostraad (2013) Available from: https://altomkost.dk/publikationer/publikation/pub/hent-fil/publication/de-officielle-kostraad/. Accessed 23 April 2019 [article in Danish]

  26. 26.

    Olsen A, Egeberg R, Halkjær J, Christensen J, Overvad K, Tjønneland A (2011) Healthy aspects of the Nordic diet are related to lower total mortality. J Nutr 141(4):639–644. https://doi.org/10.3945/jn.110.131375

    CAS  Article  PubMed  Google Scholar 

  27. 27.

    Kirkegaard H, Johnsen NF, Christensen J, Frederiksen K, Overvad K, Tjonneland A (2010) Association of adherence to lifestyle recommendations and risk of colorectal cancer: a prospective Danish cohort study. BMJ 341:c5504. https://doi.org/10.1136/bmj.c5504

  28. 28.

    Khera AV, Emdin CA, Drake I et al (2016) Genetic risk, adherence to a healthy lifestyle, and coronary disease. N Engl J Med 375(24):2349–2358. https://doi.org/10.1056/NEJMoa1605086

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  29. 29.

    Langenberg C, Sharp SJ, Franks PW et al (2014) Gene-lifestyle interaction and type 2 diabetes: the EPIC interact case-cohort study. PLoS Med 11(5):e1001647. https://doi.org/10.1371/journal.pmed.1001647

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  30. 30.

    Sladek R, Rocheleau G, Rung J et al (2007) A genome-wide association study identifies novel risk loci for type 2 diabetes. Nature. 445(7130):881–885. https://doi.org/10.1038/nature05616

    CAS  Article  PubMed  Google Scholar 

  31. 31.

    Perry JR, Voight BF, Yengo L et al (2012) Stratifying type 2 diabetes cases by BMI identifies genetic risk variants in LAMA1 and enrichment for risk variants in lean compared to obese cases. PLoS Genet 8(5):e1002741. https://doi.org/10.1371/journal.pgen.1002741

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  32. 32.

    Dudbridge F, Fletcher O (2014) Gene-environment dependence creates spurious gene-environment interaction. Am J Hum Genet 95(3):301–307. https://doi.org/10.1016/j.ajhg.2014.07.014

    CAS  Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

Parts of this work were presented in abstract form at the European Human Genetics Conference in Gothenburg, Sweden, 15–18 June 2019, and at the European Association for the Study of Diabetes Conference in Barcelona, Spain, 16–20 September 2019.

Funding

The Diet, Cancer and Health cohort is supported by the Danish Cancer Society. DNA purification and genotyping were done using a grant to OP from the UNIK – Food, Fitness & Pharma research initiative. The Novo Nordisk Foundation Center for Basic Metabolic Research is an independent research centre at the University of Copenhagen, partially funded by an unrestricted donation from the Novo Nordisk Foundation (cbmr.ku.dk; grant number NNF18CC0034900). TOK was supported by the Danish Council for Independent Research (DFF – 1333-00124 and Sapere Aude programme grant DFF – 1331-00730B) and the Novo Nordisk Foundation (NNF17OC0026848). HJ, TMS and GDC were supported by the Danish Diabetes Academy, which is funded by the Novo Nordisk Foundation (grant number NNF17SA0031406).

Author information

Affiliations

Authors

Contributions

The research initiative and study concept were taken and designed by OP and TOK. TMS, HJ and TOK designed the present analysis. TMS, HJ and LÄ performed statistical analysis. TMS, HJ and TOK interpreted the data and drafted the manuscript. AT and KO collected data and contributed to discussions. GDC, NG, TIAS, OP and TH contributed to discussions. All authors interpreted the results, edited and revised the manuscript and read and approved the final version of the manuscript. TMS and HJ are the guarantors of the work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and accuracy of the data analysis.

Corresponding author

Correspondence to Hermina Jakupović.

Ethics declarations

The authors declare that there are no relationships or activities that might bias, or be perceived to bias, their work.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

ESM

(PDF 360 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Schnurr, T.M., Jakupović, H., Carrasquilla, G.D. et al. Obesity, unfavourable lifestyle and genetic risk of type 2 diabetes: a case-cohort study. Diabetologia 63, 1324–1332 (2020). https://doi.org/10.1007/s00125-020-05140-5

Download citation

Keywords

  • Body weight
  • Gene–environment interaction
  • Genetic risk score
  • Healthy lifestyle
  • Obesity
  • Type 2 diabetes