Questionnaire-Based Polyexposure Assessment Outperforms Polygenic Scores for Classification of Type 2 Diabetes in a Multiancestry Cohort

Our plain-English summary of Questionnaire-Based Polyexposure Assessment Outperforms Polygenic Scores for Classification of Type 2 Diabetes in a Multiancestry Cohort, by F.S. Akhtari, D. Lloyd, A. Burkholder, et al (Diabetes Care, 2023).

The study

• Authors: F.S. Akhtari, D. Lloyd, A. Burkholder, X. Tong, J.S. House, E.Y. Lee, J. Buse, S.H. Schurman, D.C. Fargo, C.P. Schmitt, J. Hall, A.A. Motsinger-Reif
• Journal: Diabetes Care
• Published: 2023;46(5):929-937
• Cited by: 10 articles, as counted by NIH iCite
• Read it: PubMed · Publisher · Free full text

Abstract

Objective: Environmental exposures may have greater predictive power for type 2 diabetes than polygenic scores (PGS). Studies examining environmental risk factors, however, have included only individuals with European ancestry, limiting the applicability of results. We conducted an exposome-wide association study in the multiancestry Personalized Environment and Genes Study to assess the effects of environmental factors on type 2 diabetes.

Research Design And Methods: Using logistic regression for single-exposure analysis, we identified exposures associated with type 2 diabetes, adjusting for age, BMI, household income, and self-reported sex and race. To compare cumulative genetic and environmental effects, we computed an overall clinical score (OCS) as a weighted sum of BMI and prediabetes, hypertension, and high cholesterol status and a polyexposure score (PXS) as a weighted sum of 13 environmental variables. Using UK Biobank data, we developed a multiancestry PGS and calculated it for participants.

Results: We found 76 significant associations with type 2 diabetes, including novel associations of asbestos and coal dust exposure. OCS, PXS, and PGS were significantly associated with type 2 diabetes. PXS had moderate power to determine associations, with larger effect size and greater power and reclassification improvement than PGS. For all scores, the results differed by race.

Conclusions: Our findings in a multiancestry cohort elucidate how type 2 diabetes odds can be attributed to clinical, genetic, and environmental factors and emphasize the need for exposome data in disease-risk association studies. Race-based differences in predictive scores highlight the need for genetic and exposome-wide studies in diverse populations.

© 2023 by the American Diabetes Association.

In plain English

This observational study used 9,414 participants in the Personalized Environment and Genes Study (PEGS), a diverse North Carolina cohort. Participants answered questionnaires about health, lifestyle, exposures at home and work, and any diabetes diagnosis. The authors tested 662 exposures one at a time against type 2 diabetes. Adjusting for age, BMI, income, sex and self-reported race, they found 76 significant associations. Self-reported asbestos exposure was associated with higher odds of type 2 diabetes (odds ratio 1.44, 95% CI 1.10 to 1.87). So was coal dust exposure (odds ratio 1.66, 95% CI 1.08 to 2.49); the authors called both associations novel. The article does not name an asbestos fiber type. The authors also compared a combined score from 13 exposures with a genetic score, both standardized to mean 0 and standard deviation 1. In the test data set, the exposure score had an odds ratio of 1.921 (95% CI 1.604 to 2.304), against 1.360 (95% CI 1.115 to 1.662) for the genetic score. They noted that self-reported exposures may carry recall bias and that the study shows associations, not causation.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Observational, questionnaire-based exposome-wide association study (testing many exposures one at a time with logistic regression), followed by multi-exposure modeling and comparison of clinical, exposure-based and genetic risk scores
• Who was studied: 9,414 PEGS participants in North Carolina, surveyed beginning in 2013. Diabetes status came from the survey question on whether participants had been diagnosed with diabetes. 4,737 had whole-genome sequencing. The score analyses used a derivation set of 3,611 (462 cases), a training set of 1,774 (139 cases) and a test set of 1,847 (144 cases).
• Exposure: 662 self-reported questionnaire exposure variables, covering lifestyle, occupational, residential, chemical, diet, medication, sleep and socioeconomic factors. These included asbestos exposure (no fiber type given) and coal dust exposure.
• Main results:
  • Across the three surveys, 76 exposures were significantly associated with type 2 diabetes (28 in the Health and Exposure Survey, 0 in the Internal Exposome Survey, 48 in the External Exposome Survey), after adjusting for age, BMI, income, sex and self-reported race.
  • In the whole cohort, self-reported asbestos exposure was associated with increased odds of type 2 diabetes (OR 1.44, 95% CI 1.10 to 1.87, adjusted P = 0.02). Coal dust exposure was also associated with increased odds (OR 1.66, 95% CI 1.08 to 2.49, P = 0.04).
  • Asbestos exposure and coal dust exposure were among the variables kept in the most parsimonious multi-exposure model for the Health and Exposure Survey data.
  • The clinical, exposure and genetic scores were each standardized to a mean of 0 and a standard deviation of 1 across all participants.
  • In the test data set, the polyexposure score (PXS) had an OR of 1.921 (95% CI 1.604 to 2.304) and an AUC of 0.775 (95% CI 0.737 to 0.813). The polygenic score (PGS) had an OR of 1.360 (95% CI 1.115 to 1.662) and an AUC of 0.731 (95% CI 0.692 to 0.770). These odds ratios apply to the standardized scores, not to any single exposure.
  • The overall clinical score (OCS, built from BMI and prediabetes, hypertension and high cholesterol status) had the highest discriminative power, followed by PXS and then PGS.
  • Classification accuracy of the clinical score model improved by 44.6% when PXS was added and by 17.4% when PGS was added (overall continuous net reclassification index).
  • In the test data set, the mean of each score was significantly higher for Black participants than for White participants.
• The authors' conclusion: The authors concluded that their exposome-wide analysis identified emerging risk factors for diabetes such as asbestos exposure. They described the asbestos and coal dust associations as novel. They also concluded that the questionnaire-based polyexposure score had a larger effect size and greater reclassification improvement than the polygenic score, and that the results differed by race.
• Limitations the authors state:
  • Exposure data were self-reported, so recall bias is possible.
  • Some exposure variables were left out because of insufficient sample sizes.
  • The study identifies associations, not causation, and reverse causality is possible for some exposures.
  • Because the data are observational, the direction of causality, potential confounders and comorbidities cannot be determined.
  • Small sample sizes in some stratified analyses may have reduced statistical power.
  • Participants with genotyping data may differ from those with exposure data only, which could introduce bias. The authors state that replication in independent data sets is needed.

Topics

Medical terms in this study, explained in our Lexicon: Environmental Exposure.

Related studies

• Metformin and survival of people with type 2 diabetes and pleural mesothelioma: A population-based retrospective cohort study — H. Wu, J. Walker, et al. (2016)
• Questionnaire-based exposome-wide association studies for common diseases in the Personalized Environment and Genes Study — D. Lloyd, J.S. House, et al. (2024)
• Volumetric PET Response Assessment Outperforms Conventional Criteria in Patients Receiving High-Dose Pembrolizumab for Malignant Mesothelioma — J. Ferdinandus, F. Barbato, et al. (2021)
• Retrospective assessment of asbestos exposure--I. Case-control analysis in a study of lung cancer: efficiency of job-specific questionnaires and job exposure matrices — W. Ahrens, K.H. Jöckel, et al. (1993)
• Comparison of exposure assessment methods in a lung cancer case-control study: performance of a lifelong task-based questionnaire for asbestos and PAHs — E. Bourgkard, P. Wild, et al. (2013)
• A randomised assessment of the use of a quality of life questionnaire with or without intervention in patients attending a thoracic cancer clinic — K. Nimako, B. Ayite, et al. (2017)

Browse return to medical research.