The Influence of Sun Exposure on T2DM-Related Outcomes
The Influence of Sun Exposure on T2DM-Related Outcomes
Where appropriate, PRISMA guidelines for the reporting of systematic reviews and meta-analyses were followed. Searches for pre-existing reviews on sun exposure and T2DM end-points, using Thomsom Reuters' Web of Science (formerly known as ISI Web of Knowledge), Scopus (Science Direct), PubMed and The Cochrane Collaboration, yielded no results. We undertook a search for full-text articles with sun exposure and either T2DM or glucose metabolism or both outcomes. The electronic search was conducted within Web of Science and Scopus; search terms are shown in Table 1. Restrictions were placed on title (to omit experimental rodent studies), language (to omit publications in languages other than English), article type (to omit reviews) and publication date (1966 through to July 2012). Briefly, the sun exposure terms captured exposure to radiation from the sun or artificial UVR and were measured as one or any combination of the following: time outdoors, time between certain latitudes or in different climates, composite indices comprising sun exposure duration and body surface area exposed to sun, solar irradiance at the skin measured by polysulphone ultraviolet wristwatch methods, ambient UVR, latitude, season, time of year or sun-seeking behaviour such as sunbathing or sun bed use. T2DM was defined by self-report, clinical diagnosis, hospital records, registry linkage, medication use or standard diagnostic glucose metabolism test results. Glucose metabolism was defined as any one or more of the following: fasting glucose, fasting insulin, HbA1c, test results from oral glucose tolerance and glucose clamp techniques or composite indices manipulating these glucose metabolism measurements.
Three researchers (CS-L, SLB, KMS) searched titles, abstracts and full articles of the resulting output to determine eligibility of the articles against predetermined criteria before checking reference lists of eligible publications for additional potentially eligible articles. Peer-reviewed full-text articles which were published in English and comprised cohort, case–control or cross-sectional study designs, and were population, community, clinic or institution based (with short-term institutionalization of <1 week) were eligible for inclusion. Articles on gestational diabetes mellitus, type 1 diabetes mellitus and diabetes insipidus were excluded, maintaining a more homogenous group of study populations and outcomes. Studies were also excluded if they (i) investigated T2DM in people younger than 18 years of age; (ii) consisted of a source population who were completely diabetic or a population in which all individuals experienced limited or excessive sun exposure or were matched on sun exposure (including studies conducted solely within the tropics); (iii) were intervention studies (for example vitamin D supplementation trials or ecological studies reporting diabetes outcomes in different geographical regions without analysis of latitudinal differences); (iv) provided only unadjusted data for an association between sun exposure and T2DM-related outcomes. Although age, obesity and physical activity are key confounders in the association between sun exposure and T2DM-related outcomes, adjustment for these factors did not determine article eligibility. Instead, the degree to which the eligible articles controlled for these factors influenced their risk-of-bias assessment and therefore their inclusion into the best evidence synthesis. We did not include abstracts or search the grey literature.
Methodology for Article Identification
Where appropriate, PRISMA guidelines for the reporting of systematic reviews and meta-analyses were followed. Searches for pre-existing reviews on sun exposure and T2DM end-points, using Thomsom Reuters' Web of Science (formerly known as ISI Web of Knowledge), Scopus (Science Direct), PubMed and The Cochrane Collaboration, yielded no results. We undertook a search for full-text articles with sun exposure and either T2DM or glucose metabolism or both outcomes. The electronic search was conducted within Web of Science and Scopus; search terms are shown in Table 1. Restrictions were placed on title (to omit experimental rodent studies), language (to omit publications in languages other than English), article type (to omit reviews) and publication date (1966 through to July 2012). Briefly, the sun exposure terms captured exposure to radiation from the sun or artificial UVR and were measured as one or any combination of the following: time outdoors, time between certain latitudes or in different climates, composite indices comprising sun exposure duration and body surface area exposed to sun, solar irradiance at the skin measured by polysulphone ultraviolet wristwatch methods, ambient UVR, latitude, season, time of year or sun-seeking behaviour such as sunbathing or sun bed use. T2DM was defined by self-report, clinical diagnosis, hospital records, registry linkage, medication use or standard diagnostic glucose metabolism test results. Glucose metabolism was defined as any one or more of the following: fasting glucose, fasting insulin, HbA1c, test results from oral glucose tolerance and glucose clamp techniques or composite indices manipulating these glucose metabolism measurements.
Three researchers (CS-L, SLB, KMS) searched titles, abstracts and full articles of the resulting output to determine eligibility of the articles against predetermined criteria before checking reference lists of eligible publications for additional potentially eligible articles. Peer-reviewed full-text articles which were published in English and comprised cohort, case–control or cross-sectional study designs, and were population, community, clinic or institution based (with short-term institutionalization of <1 week) were eligible for inclusion. Articles on gestational diabetes mellitus, type 1 diabetes mellitus and diabetes insipidus were excluded, maintaining a more homogenous group of study populations and outcomes. Studies were also excluded if they (i) investigated T2DM in people younger than 18 years of age; (ii) consisted of a source population who were completely diabetic or a population in which all individuals experienced limited or excessive sun exposure or were matched on sun exposure (including studies conducted solely within the tropics); (iii) were intervention studies (for example vitamin D supplementation trials or ecological studies reporting diabetes outcomes in different geographical regions without analysis of latitudinal differences); (iv) provided only unadjusted data for an association between sun exposure and T2DM-related outcomes. Although age, obesity and physical activity are key confounders in the association between sun exposure and T2DM-related outcomes, adjustment for these factors did not determine article eligibility. Instead, the degree to which the eligible articles controlled for these factors influenced their risk-of-bias assessment and therefore their inclusion into the best evidence synthesis. We did not include abstracts or search the grey literature.