Journal Article Cohort Profile: The Ontario Health Study (OHS) Get access Victoria A Kirsh, Victoria A Kirsh Ontario Institute for Cancer Research, Toronto, ON, CanadaDalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada (formerly for N.K.) https://orcid.org/0000-0003-2422-2225 Search for other works by this author on: Oxford Academic PubMed Google Scholar Kimberly Skead, Kimberly Skead Ontario Institute for Cancer Research, Toronto, ON, CanadaDepartment of Molecular Genetics, University of Toronto, Toronto, ON, Canada Search for other works by this author on: Oxford Academic PubMed Google Scholar Kelly McDonald, Kelly McDonald Ontario Institute for Cancer Research, Toronto, ON, Canada Search for other works by this author on: Oxford Academic PubMed Google Scholar Nancy Kreiger, Nancy Kreiger Dalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada (formerly for N.K.)Prevention and Cancer Control, Ontario Health, Cancer Care Ontario, Toronto, ON, Canada Search for other works by this author on: Oxford Academic PubMed Google Scholar Julian Little, Julian Little Faculty of Medicine, School of Epidemiology and Public Health, University of Ottawa, Ottawa, ON, Canada Search for other works by this author on: Oxford Academic PubMed Google Scholar Karen Menard, Karen Menard Office of Institutional Research and Planning, University of Guelph, Guelph, ON, Canada Search for other works by this author on: Oxford Academic PubMed Google Scholar John McLaughlin, John McLaughlin Dalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada (formerly for N.K.) Search for other works by this author on: Oxford Academic PubMed Google Scholar Sutapa Mukherjee, Sutapa Mukherjee Adelaide Institute for Sleep Health, Flinders University, Adelaide, South Australia, Australia https://orcid.org/0000-0001-5021-1648 Search for other works by this author on: Oxford Academic PubMed Google Scholar Lyle J Palmer, Lyle J Palmer School of Public Health, University of Adelaide, Adelaide, South Australia, Australia Search for other works by this author on: Oxford Academic PubMed Google Scholar Vivek Goel, Vivek Goel Dalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada (formerly for N.K.)Office of the President, University of Waterloo, Waterloo, ON, Canada Search for other works by this author on: Oxford Academic PubMed Google Scholar ... Show more Mark P Purdue, Mark P Purdue Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Rockville, MD, USA Search for other works by this author on: Oxford Academic PubMed Google Scholar Philip Awadalla Philip Awadalla Ontario Institute for Cancer Research, Toronto, ON, CanadaDalla Lana School of Public Health, University of Toronto, Toronto, ON, Canada (formerly for N.K.)Department of Molecular Genetics, University of Toronto, Toronto, ON, Canada Corresponding author. Ontario Institute for Cancer Research, 661 University Ave, Suite 510, Toronto, ON, M5G 0A3, Canada. E-mail: philip.awadalla@oicr.on.ca https://orcid.org/0000-0001-9946-6393 Search for other works by this author on: Oxford Academic PubMed Google Scholar International Journal of Epidemiology, Volume 52, Issue 2, April 2023, Pages e137–e151, https://doi.org/10.1093/ije/dyac156 Published: 13 August 2022 Article history Received: 22 December 2021 Editorial decision: 01 July 2022 Accepted: 20 July 2022 Published: 13 August 2022

This paper examines the influence of an initial face-to-face meeting in an e-mail delivered physical activity intervention. RMANCOVA was conducted to examine differences between e-mail and face-to-face arms in mean weekly steps adjusted for baseline steps, sex, and body mass index (BMI). Overall, 68 participants were analyzed (face-to-face: (n=35); e-mail: (n=33). Participants in the face-to-face arm walked a greater number of steps each week; however, differences between arms pre- and post-intervention were not statistically significant (F (1, 60) = 0.31, p = .5796). These findings suggest that a physical activity intervention where initial instructions are delivered via e-mail is not statistically significantly different from a face-to-face approach.

Objectives The causes of kidney cancer are not well understood though occupational exposures are thought to play a role. Crystalline silica is a known human carcinogen, and despite previous links with kidney disease, there have been few studies investigating its association with kidney cancer. We addressed this research gap using a population-based case-control study of Canadian men. Methods Questionnaire data were obtained from individuals with histologically confirmed kidney cancer, and population-based controls recruited from eight Canadian provinces (1994-1997). An industrial hygienist characterised participants' lifetime occupational exposure, and their confidence in the assessment (possibly, probably or definitely exposed) to silica on three dimensions (intensity, frequency and duration), and cumulative exposure was estimated. Logistic regression was used to estimate ORs and 95% CIs, adjusting for potential confounders. Results Nearly half of the 689 kidney cancer cases (49%) and 2369 controls (44%) had ever been occupationally exposed to crystalline silica. In a fully adjusted model, workers ever-exposed to silica had a slightly increased risk of kidney cancer relative to those who were unexposed (OR 1.10, 95%CI 0.92 to 1.32). Odds were modestly (and generally not statistically significantly) increased for models with duration of exposure and cumulative exposure, though exposure-response relationships were not evident. Conclusions Our findings do not provide evidence that occupational exposure to crystalline silica increases risk of kidney cancer in men.

Although the short-term impact of incident fragility fractures on health-related quality of life (HRQL) of older people has been confirmed, we lack long-term evidence. We explored the impact of incident fragility fractures on HRQL, among people aged 50 years and older, using 10-year prospective data from the Canadian Multicentre Osteoporosis Study (CaMos). This study was based on data from 7753 (2187 men and 5566 women) participants of CaMos. The HRQL, measured through the Health Utility Index (HUI), was captured at baseline and year 10. The incident fragility fractures were recorded over 10 years of follow-up at spine, hip, rib, shoulder, pelvis, or forearm. Multivariable regression analysis was conducted to measure the mean difference, termed as deficit, in the HUI scores for participants with and without fractures. We examined the effects of single or multiple fragility fractures, time (fractures that occurred between year 1 to 5 and 6 to 10) and recovery to the prefracture level. Incident spine and hip fractures were associated with significant deficits (varied from -0.19 to -0.07) on the HUI scores. Hip and spine fractures were associated with negative impact on mobility, self-care, and ambulation. Fractures that occurred closer to the follow-up assessment were associated with significant impact on HRQL compared to fractures occurring a long time before it, except for hip fracture (deficits lasted 5 years or longer). Similarly, multiple hip (-0.14), spine (-0.16), and rib (-0.21) fractures significantly impacted the HRQL of women. Women with a hip fracture never recovered to their prefracture level score (OR = 0.41; 95% confidence interval [CI], 0.19 to 0.98). Our analysis suggests that single and multiple hip fractures as well as multiple spine and rib fractures strongly impact the HRQL of older people over a prolonged period of time. (c) 2019 American Society for Bone and Mineral Research.

This analysis uses data from the Community Noise and Health Study developed by Statistics Canada to investigate the association between residential proximity to wind turbines and health-related outcomes in a dataset that also provides objective measures of wind turbine noise. The findings indicate that residential proximity to wind turbines is correlated with annoyance and health-related quality of life measures. These associations differ in some respects from associations with noise measurements. Results can be used to support discussions between communities and wind-turbine developers regarding potential health effects of wind turbines.

Abstract Background Although several studies have suggested that obesity and other modifiable factors influence breast carcinogenesis, there is limited evidence on the combined effect of these exposures on risk of breast cancer. Thus, to gain further insight into whether modifiable risk factors may operate collectively to alter risk of breast cancer, we assessed the association of a healthy lifestyle index (HLI) score (a combination of diet, physical activity, smoking, alcohol consumption and anthropometry) with risk of invasive breast cancer among women in the Canadian Study of Diet, Lifestyle, and Health (CSDLH). Method We performed a case-cohort study using a cohort of 39,532 female participants in the CSDLH who were recruited between 1995 and 1998. The study included an age-stratified subcohort of 3,027 women and 1,018 incident invasive breast cancer cases. Lifestyle and dietary information was collected at baseline using self-administered questionnaires. We estimated hazard ratios (HR) and 95 % confidence intervals (CI) for the association of the HLI score and the individual HLI components with risk of breast cancer using Cox proportional hazards regression modified for the case cohort design. Results: Every unit increase in the HLI score was associated with a 3% reduced risk of breast cancer (HR: 0.97; 95% CI: 0.95-1.00). When considering quintiles of the HLI score, those in the highest quintile had a 26% lower risk for breast cancer compared to those with HLI score in the lowest quintile (HRq5 vs q1: 0.74; 95% CI: 0.59-0.93). In analyses stratified by menopausal status, there was a similar inverse association between the HLI score and risk of breast cancer among postmenopausal women (HR: 0.97; 95% CI: 0.94-1.00) and a tendency towards an inverse association among premenopausal women (HR per unit increase in the HLI score: 0.98; 95% CI: 0.95-1.01). Among the individual components of the HLI score, there was some suggestion of a reduction in risk of breast cancer in association with a relatively high physical activity level in the overall study population (HRq5 vs. q1: 0.81: 95% CI: 0.63-1.04, p trend: 0.08), but consuming a healthy diet, smoking, having high alcohol consumption and being obese were not associated with altered risk of breast cancer. Conclusion: Our findings indicate that adherence to a healthy lifestyle may be associated with reduced risk of breast cancer. These findings suggest the need to further investigate how modifiable risk factors may act jointly to contribute to the development of breast cancer. Such knowledge may have important implications for the development of interventions designed to promote a healthy lifestyle to aid in the primary prevention of breast cancer. Citation Format: Arthur R, Kirsh V, Kreiger N, Rohan T. The joint effect of modifiable risk factors on risk of invasive breast cancer among Canadian women [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P5-12-01.

Several modifiable risk factors have been associated with risk of female cancers, but there is limited data regarding their combined effect on risk among Canadian women. Therefore, we assessed the joint association of modifiable risk factors, using a healthy lifestyle index (HLI) score, with risk of specific reproductive cancers.

BACKGROUND Dietary recommendations emphasize increased consumption of fruit, vegetables, and whole grain cereals for prevention of chronic disease.OBJECTIVES This study assessed the effect of dietary advice and/or food provision on body weight and cardiovascular disease risk factors.METHODS Healthy overweight men (n = 209) and women (n = 710), mean age 44.7 years, body mass index [BMI] 32.4 kg/m(2), were randomized between November 2005 and August 2009 to receive Health Canada's food guide (control, n = 486) or 1 of 3 interventions: dietary advice consistent with both Dietary Approaches to Stop Hypertension (DASH) and dietary portfolio principles (n = 145); weekly food provision reflecting this advice (n = 148); or food delivery plus advice (n = 140). Interventions lasted 6 months with 12-month follow-up. Semiquantitative food frequency questionnaires and fasting blood, anthropometric and blood pressure measurements were obtained at baseline, 6 months, and 18 months.RESULTS Participant retention at 6 and 18 months was 91% and 81%, respectively, after food provision compared to 67% and 57% when no food was provided (p < 0.0001). Test and control treatments showed small reductions in body weight (-0.8 to -1.2 kg), waist circumference (-1.1 to -1.9 cm), and mean arterial pressure (0.0 to -1.1 mm Hg) at 6 months and Framingham coronary heart disease risk score at 18 months (-0.19 to -0.42%), which were significant overall. Outcomes did not differ among test and control groups.CONCLUSIONS Provision of foods increased retention but only modestly increased intake of recommended foods. Current dietary recommendations showed small overall benefits in coronary heart disease risk factors. Additional dietary strategies to maximize these benefits are required. (Fruits, Vegetables, and Whole Grains: A Community-based Intervention; NCT00516620) (C) 2017 The Authors. Published by Elsevier on behalf of the American College of Cardiology Foundation.

There is evidence to suggest the burden of dementia may be greater in Indigenous populations than it is in non-Indigenous populations. The objective of this study is to characterize the epidemiology of dementia diagnoses and care in off-reserve, community-dwelling, Ontario Indigenous populations using data from the Canadian Community Health Survey (CCHS) and provincial health insurance claims datasets housed at the Institute of Clinical and Evaluative Sciences (ICES). We have established a Community Advisory Board (CAB) comprised of five members from a variety of research backgrounds, Indigenous communities and lived experiences to provide leadership, support and direction to all stages of our research. Indigenous and non-Indigenous populations will be identified using the CCHS. We will use a definition of dementia (including Alzheimer's disease) developed and previously validated by ICES. All analyses will be conducted using SAS 9.3. Differences in prevalence estimates and patterns of care and service usage for Indigenous and non-Indigenous participants will be compared using Chi-square tests for frequencies and t-tests for means. Risk factors for dementia will be identified using PROC GENMOD to build a multivariate logistic model for both Indigenous and non-Indigenous participants. Participant characteristics will be reported as frequencies and proportions for Indigenous and non-Indigenous respondents by sex. Age-specific dementia prevalence estimates will also be reported for each group. Odds ratios will be reported from the final multivariate model for Indigenous and non-Indigenous participants. Differences in frequencies of drug prescirption types (e.g. Donepezil, Galantamine, Memantine, Rivastigmine, Tacrine), hospital admissions, specialist care, home care, and physician visits will be will be identified using Chi-square tests between Indigenous and non-Indigenous participants. This study will be the first of its kind in Ontario. By characterizing the epidemiology of dementia cases and care in community-dwelling, off-reserve, Indigenous populations we hope to identify risk factors, identify patterns of care for dementia health services, and increase awareness of dementia among Indigenous populations in Ontario.

Introduction Même si certaines études suggèrent une association entre le travail par quarts et l’obésité, peu sont fondées sur la population ou tiennent compte de divers horaires de travail par quarts. L’obésité étant associée à plusieurs problèmes de santé chroniques, il est important de comprendre quelles formes de travail par quarts ont une incidence sur elle et d’effectuer des travaux permettant d’évaluer de façon plus détaillée l’exposition au travail par quarts. Méthodologie Au moyen d’une régression logistique polytomique multivariée, nous avons étudié l’association entre le travail par quarts (de soir ou nuit, par quarts ou autre type de rotation) et le surpoids et l’obésité, en fonction d’une mesure transversale de l’indice de masse corporelle chez 1 561 hommes. Ces hommes avaient déjà servi de témoins dans une étude cas-témoins sur le cancer de la prostate menée dans le nord-est de l’Ontario de 1995 à 1999. Nous avons obtenu l’information sur leurs antécédents de travail (notamment sur le travail par quarts), leur taille et leur poids à partir de données autodéclarées recueillies par questionnaire. Résultats Nous avons observé une association entre le fait d’avoir déjà travaillé par quarts (par opposition au fait de n’avoir jamais travaillé par quarts) et le surpoids (RC [rapport de cotes] = 1,34; IC [intervalle de confiance] à 95 % : 1,05 à 1,73) ainsi que l’obésité (RC = 1,57; IC à 95 % : 1,12 à 2,21). Nous avons également observé des associations statistiquement non significatives avec le fait d’avoir déjà travaillé (par opposition au fait de n’avoir jamais travaillé) de façon permanente selon un quart de soir ou de nuit. Nous avons par ailleurs observé une tendance à la hausse statistiquement significative en ce qui concerne le risque de surpoids et d’obésité en fonction de la durée du travail par quarts. Conclusion Tant l’association positive observée entre le travail par quarts et l’obésité que l’association positive suggérée dans notre étude en ce qui concerne le travail permanent selon un quart de soir ou nuit concordent avec ce qui a été observé antérieurement. D’autres études en population tenant compte de nos résultats seront à mener pour examiner d’autres caractéristiques du travail par quarts, afin de mieux déterminer si certains types de travail par quarts ont une plus grande incidence sur l’obésité que d’autres.

INTRODUCTION:While some studies have suggested associations between shift work and obesity, few have been population-based or considered multiple shift schedules. Since obesity is linked with several chronic health conditions, understanding which types of shift work influence obesity is important and additional work with more detailed exposure assessment of shift work is warranted.METHODS:Using multivariate polytomous logistic regression, we investigated the associations between shift work (evening/night, rotating and other shift schedules) and overweight and obesity as measured by body mass index cross-sectionally among 1561 men. These men had previously participated as population controls in a prostate cancer case-control study conducted in northeastern Ontario from 1995 to 1999. We obtained information on work history (including shift work), height and weight from the existing self-reported questionnaire data.RESULTS:We observed an association for ever (vs. never) having been employed in rotating shift work for both the overweight (OR [odds ratio] = 1.34; 95% CI [confidence interval]: 1.05-1.73) and obese (OR = 1.57; 95% CI: 1.12-2.21) groups. We also observed nonsignificant associations for ever (vs. never) having been employed in permanent evening/night shifts. In addition, we found a significant trend of increased risk for both overweight and obesity with increasing duration of rotating shift work.CONCLUSION:Both the positive association between rotating shift work and obesity and the suggested positive association for permanent evening/night shift work in this study are consistent with previous findings. Future population-based research that is able to build on our results while examining additional shift work characteristics will further clarify whether some shift patterns have a greater impact on obesity than others.

OBJECTIVE:Prostate cancer continues to be the most commonly diagnosed cancer in men, and there is limited knowledge on its preventable risk factors. A number of occupational exposures in natural resource-based industries are suspected to be related to prostate cancer risk. This study investigates associations between employment in these industries and prostate cancer.METHODS:Data were from a population-based, case-control study previously conducted in Northeastern Ontario. Incident cases (N=760) aged 45-85 years and diagnosed with prostate cancer between 1995 and 1998 were identified from the Ontario Cancer Registry. Controls (N=1632) were recruited using telephone listings, and were frequency matched to cases by age. Lifetime occupational history was collected for all participants. Logistic regression was used to estimate ORs and their associated 95% CIs.RESULTS:Elevated risks were observed for employment in forestry and logging industries (OR=1.87, 95% CI 1.32 to 2.73) and occupations (OR=1.71, 95% CI 1.24 to 2.35), and these risks increased with duration of employment for ≥10 years. Elevated risks were also found for employment in wood products industries (OR=1.45, 95% CI 1.07 to 1.97), and paper and allied products industries (OR=1.43, 95% CI 1.03 to 2.00), and when duration of employment was ≥10 years. There were also elevated risks in agriculture and mining-related work; however, these findings were not consistent across industry and occupation categories.CONCLUSIONS:Prostate cancer risk may be associated with work in several natural resource industries, primarily in the forest industries. To further evaluate observed associations, studies should focus on natural resource-based exposures in larger populations with improved exposure assessment.

OBJECTIVES:Preventable risk factors for prostate cancer are poorly understood; sun exposure is a possible protective factor. The goal of this study was to investigate prostate cancer risk in outdoor workers, a population with high sun exposure.METHODS:Prostate cancer cases and controls from a large study (conducted between 1994 and 1997) were used for this analysis. A job exposure matrix (JEM) was used to assign solar ultraviolet radiation (UVR) at work as moderate (2 to <6 hours outside/day) or high (≥6 hours). Average daily satellite UV-B measures were linked to the latitude/longitude of the residences of each participant. Several other exposure metrics were also examined, including ever/never exposed and standard erythemal dose by years (SED×years). Logistic regression was used to evaluate the association between solar UVR exposure and the odds of prostate cancer.RESULTS:A total of 1638 cases and 1697 controls were included. Men of Indian and Asian descent had reduced odds of prostate cancer (ORs 0.17 (0.08 to 0.35) and 0.25 (0.15 to 0.41), respectively) compared with Caucasian men, as did single men (OR 0.76 (0.58 to 0.98)) compared with married men. Overall, no statistically significant associations were observed between sun exposure and prostate cancer with 1 exception. In the satellite-enhanced JEM that considered exposure in high category jobs only, prostate cancer odds in the highest quartile of cumulative exposure was decreased compared with unexposed men (OR 0.68 (0.51 to 0.92)).CONCLUSIONS:This study found limited evidence for an association with prostate cancer, with the exception of 1 statistically significant finding of a decreased risk among workers with the longest term and highest sun exposure.

Objectives Prostate cancer is the second most common cancer in Canadian men, but preventable risk factors are poorly understood. Vitamin D has been hypothesised to be a protective agent, so the goal of this study was to investigate the potentially reduced risk of prostate cancer in outdoor workers, a population at risk of high exposure to the sun. Methods Prostate cancer cases and controls in a large Canadian population-based case-control study (carried out between 1994 and 1997) were used for this analysis. A job exposure matrix (SUNJEM) was used to assign exposure in 2 categories: moderate (2 ≤ 6 hours outside/day) or high (≥6). Satellite UV-B measures were linked to the latitude/longitude of each residence during the working life. Several exposure metrics were examined including a cumulative measure of standard erythemal dose – years (SED-years). Logistic regression was used to evaluate the association between the various measures of occupational solar UVR exposure and prostate cancer. Results A total of 3,335 men were included (1638 cases, 1697 controls). 737 cases (45%) and 713 controls (42%) were ever exposed to solar UVR at work. In the satellite-enhanced SUNJEM that only considered exposure in high category jobs, the risk of prostate cancer was decreased compared to unexposed men, and the highest exposure group had an OR for prostate cancer of 0.68 (95% CI: 0.51–0.92). Conclusion This study found that long-term outdoor workers may have a decreased risk of prostate cancer.

Ecosystems comprise all the living and non-living things in a particular area (e.g., rain forest, desert), which interact and maintain equilibrium. Loss of equilibrium (e.g., clear-cutting trees in a rain forest) can mean the decline of the ecosystem, unless it is able to adapt to the new circumstances. The term "knowledge ecosystem" describes an approach to managing knowledge in a particular field; the components of this system include the people, the technological skills and resources, and information or data. Epidemiology can be thought of as a knowledge ecosystem and, like ecological systems, its existence can be threatened, from both internal and external forces that may alter its equilibrium. This paper describes some threats to the epidemiology knowledge ecosystem, how these threats came about, and what responses we can make that may serve to mitigate those threats.

• Confirms much of what is known about barriers to breast screening in Canada, and adds unique information. • Hints that some of the known barriers to screening have underlying causes that may vary by region. • Women need to be better informed about screening benefits, accessibility, lack of cost, the process, and to know they may request a female technician. • Screening staff need to be informed of their importance in future screening attendance, as their behaviour affects a woman's likelihood of returning for screening.

We hypothesized that diet would have direct effects on glucose metabolism with direct and indirect effects on bone metabolism in a cohort of Canadian adults. We assessed dietary patterns (Prudent [fruit, vegetables, whole grains, fish, and legumes] and Western [soft drinks, potato chips, French fries, meats, and desserts]) from a semiquantitative food frequency questionnaire. We used fasting blood samples to measure glucose, insulin, homeostatic model assessment insulin resistance (HOMA-IR), 25-hydroxyvitamin D (25OHD), parathyroid hormone, bone-specific alkaline phosphatase (a bone formation marker), and serum C-terminal telopeptide (CTX; a bone resorption marker). We used multivariate regression models adjusted for confounders and including/excluding body mass index. In a secondary analysis, we examined relationships through structural equations models. The Prudent diet was associated with favorable effects on glucose metabolism (lower insulin and HOMA-IR) and bone metabolism (lower CTX in women; higher 25OHD and lower parathyroid hormone in men). The Western diet was associated with deleterious effects on glucose metabolism (higher glucose, insulin, and HOMA-IR) and bone metabolism (higher bone-specific alkaline phosphatase and lower 25OHD in women; higher CTX in men). Body mass index adjustment moved point estimates toward the null, indicating partial mediation. The structural equation model confirmed the hypothesized linkage with strong effects of Prudent and Western diet on metabolic risk, and both direct and indirect effects of a Prudent diet on bone turnover. In summary, a Prudent diet was associated with lower metabolic risk with both primary and mediated effects on bone turnover, suggesting that it is a potential target for reducing fracture risk.

Background/ObjectivesIt has been suggested that the association between shift work and chronic disease is mediated by an increase in obesity. However, investigations of the relationship between shift work and obesity reveal mixed findings. Using a recently developed exposure assessment tool, this study examined the association between shift work and obesity among Canadian women from two studies: a cohort of university alumni, and a population-based study.MethodsSelf-administered questionnaire data were used from healthy, currently employed females in a population-based study, the Ontario Women’s Diet and Health case-control study (n = 1611 controls), and from a subset of a of university alumni from the Canadian Study of Diet, Lifestyle, and Health (n = 1097) cohort study. Overweight was defined as BMI=25 to <30, and obesity as BMI=30. Reported occupation was converted to occupational codes and linked to a probability of shift work value derived from Survey of Labour and Income Dynamics data. Regular evenings, nights, or rotating work comprised shift work. Polytomous logistic regression estimated the association between probability of shift work, categorized as near nil, low, medium, and high probability of shift work, on overweight and obesity, controlling for detected confounders.ResultsIn the population-based sample, high probability of shift work was associated with obesity (reference = near nil probability of shift work, OR: 1.88, 95% CI: 1.01–3.51, p = 0.047). In the alumni cohort, no significant association was detected between shift work and overweight or obesity.ConclusionsAs these analyses found a positive association between high probability of shift work exposure and obesity in a population-based sample, but not in an alumni cohort, it is suggested that the relationship between shift work and obesity is complex, and may be particularly susceptible to occupational and education-related factors within a given population.