The magnitude of association between overweight and obesity and the risk of diabetes: A meta-analysis of prospective cohort studies

https://doi.org/10.1016/j.diabres.2010.04.012Get rights and content

Abstract

The objectives of this meta-analysis were to examine the magnitude of the relative risk (RR) of developing type 2 diabetes for overweight and obese populations, compared to those with normal weight, and to determine causes of the variation in RR between various cohort studies.

The magnitude of the RR was analyzed by combining 18 prospective cohort studies that matched defined criteria. The variance in RR between studies was explored. The overall RR of diabetes for obese persons compared to those with normal weight was 7.19, 95% CI: 5.74, 9.00 and for overweight was 2.99, 95% CI: 2.42, 3.72. The variation in RR among studies was explored and it was found that the effect of heterogeneity was highly related with sample size, method of assessment of body mass index (BMI) and method of ascertainment of type 2 diabetes. By combining only cohort studies with more than 400 cases of incident diabetes (>median), adjusted by at least three main confounding variables (age, family history of type 2 diabetes, physical activity), measured BMI, and diabetes determined by clinical diagnosis, the RR was 7.28, 95% CI: 6.47, 8.28 for obesity and 2.92, 95% CI: 2.57, 3.32 for overweight.

Introduction

Obesity has been associated with an increased risk of several chronic diseases, including type 2 diabetes. While most studies have shown some association between obesity and the risk of diabetes, the magnitude of the association differs between studies, ranging from relative risks of 2, through 5, 10, and 15–27 (Table 2). This suggests a meta-analysis study combining the results from different cohort studies is required in order to provide an accurate estimation of the magnitude of risk of developing type 2 diabetes for obese persons. The accurate estimation of diabetes risk for obese/overweight persons is not only useful for clinical practice, but also for public health policy.

A previous meta-analysis by Hartemink et al. [1] presented the overall relative risk of diabetes per unit of BMI and reported that diabetes risk increased by 1.18 (95% CI: 1.16, 1.20) per unit of BMI and assumes a linear relation between BMI and the relative risk of diabetes. However, several studies suggest the association may not be linear and instead that there is a threshold effect at a certain cut point of BMI. For example, Modan et al. [2] reported that the association with degree of overweight was non-linear, with threshold effects at BMI ≥31 for type 2 diabetes. The Gothenburg study [3] also found that when the incidence was plotted against the percentiles of the distribution of BMI, the relationship was non-linear. In addition, in the previous meta-analysis, the pooled relative risk of diabetes was derived from combining all types of epidemiological studies, including case control and cross sectional studies. A second meta-analysis was published by Ni Mhurchu et al. [4] but this analysis only pooled studies in the Asia-Pacific region.

The main objective of this study is to estimate the magnitude of the relationship between obesity or overweight, classified according to categories of body mass index (BMI), and the risk of developing type 2 diabetes by combining the available relative risks derived from prospective cohort studies. The second objective is to determine causes of the variation in relative risks across studies.

Section snippets

Data sources

The studies included in this analysis were obtained by performing a comprehensive electronic literature search using Embase, Medline, Pubmed and Web of Science from 1966 to 18th of June 2008 using the Medical Subject Heading Terms (MeSH) related terms of bodyweight, waist–hip ratio (WHR), body mass index (BMI), body fat distribution, overweight, obesity, weight change, weight gain, body weight and diabetes related keywords (i.e. diabetes; diabetic; NIDDM; diabetes mellitus). 27,140 articles

Characteristics of the Studies

Table 1 provides a summary of the characteristics of the studies included in the meta-analysis and Table 2 provides a detailed description of each study included in the meta-analysis. Among the studies included, 8 studies focused the analysis on men, 10 on women and 5 studied both men and women. The age range was 18–80 years. The studies included in the analysis were from different regions, 8 from the USA, and 5 each from the Asia-Pacific and Europe. Most studies recruited participants from the

Discussion

Within this meta-analysis, obesity was associated with a 7 times higher risk of diabetes compared to those with normal weight. Overweight was associated with a 3 times higher risk of diabetes compared to those with normal weight. The different magnitude of the relative risk across published studies was influenced by several study characteristics. It was found that relative risks varied according to the study population characteristics of gender and study region and according to the quality

Conclusion

After combining 12 relatively homogeneous, high quality studies, the relative risk of developing type 2 diabetes for obese persons, compared to those with normal weight was 7-fold. That for overweight was almost 3-fold. There is a clear risk associated with both overweight and obesity that demonstrates the importance of preventing population wide increases in body weight. In order to derive a reliable and consistent estimate of the health risks associated with increasing weight we recommend the

Conflict of interest

There is no conflict of interest.

Acknowledgments

A. Abdullah was supported by an AusAID scholarship and A. Peeters by a VicHealth Fellowship and an NH&MRC grant. We would also like to acknowledge the anonymous reviewers for their useful comments.

References (27)

  • S.A. Kaye et al.

    Increased incidence of diabetes mellitus in relation to abdominal adiposity in older women

    J. Clin. Epidemiol.

    (1991)
  • N. Hartemink et al.

    Combining risk estimates from observational studies with different exposure cutpoints: a meta-analysis on body mass index and diabetes type 2

    Am. J. Epidemiol.

    (2006)
  • M. Modan et al.

    Effect of past and concurrent body mass index on prevalence of glucose intolerance and type 2 (non-insulin-dependent) diabetes and on insulin response. The Israel study of glucose intolerance, obesity and hypertension

    Diabetologia

    (1986)
  • L.O. Ohlson et al.

    The influence of body fat distribution on the incidence of diabetes mellitus. 13. 5 years of follow-up of the participants in the study of men born in 1913

    Diabetes

    (1985)
  • C. Ni Mhurchu et al.

    Body mass index and risk of diabetes mellitus in the Asia-Pacific region

    Asia Pac. J. Clin. Nutr.

    (2006)
  • G.A. Colditz et al.

    Weight gain as a risk factor for clinical diabetes mellitus in women

    Ann. Int. Med.

    (1995)
  • V.J. Carey et al.

    Body fat distribution and risk of non-insulin-dependent diabetes mellitus in women. The Nurses’ Health Study

    Am. J. Epidemiol.

    (1997)
  • A.E. Field et al.

    Impact of overweight on the risk of developing common chronic diseases during a, 10-year period

    Arch. Intern. Med.

    (2001)
  • F.B. Hu et al.

    Diet, lifestyle, and the risk of type, 2 diabetes mellitus in women

    N. Engl. J. Med.

    (2001)
  • P.W.F. Wilson et al.

    Overweight and obesity as determinants of cardiovascular risk: the Framingham experience

    Arch. Intern. Med.

    (2002)
  • J.B. Meigs et al.

    Body mass index, metabolic syndrome, and risk of type, 2 diabetes or cardiovascular disease

    J. Clin. Endocrinol. Metab.

    (2006)
  • D. Badger, RevMan Exercise (For RevMan, 4.2). The Cochrane Collaboration Open Learning Material for Reviewer, 2003...
  • A.D. Rosenthal et al.

    Body fat distribution and risk of diabetes among Chinese women

    Int. J. Obes. (Lond.)

    (2004)
  • Cited by (541)

    View all citing articles on Scopus
    View full text