The incidence of co-morbidities related to obesity and overweight: A systematic review and meta-analysis

Background Overweight and obese persons are at risk of a number of medical conditions which can lead to further morbidity and mortality. The primary objective of this study is to provide an estimate of the incidence of each co-morbidity related to obesity and overweight using a meta-analysis. Methods A literature search for the twenty co-morbidities identified in a preliminary search was conducted in Medline and Embase (Jan 2007). Studies meeting the inclusion criteria (prospective cohort studies of sufficient size reporting risk estimate based on the incidence of disease) were extracted. Study-specific unadjusted relative risks (RRs) on the log scale comparing overweight with normal and obese with normal were weighted by the inverse of their corresponding variances to obtain a pooled RR with 95% confidence intervals (CI). Results A total of 89 relevant studies were identified. The review found evidence for 18 co-morbidities which met the inclusion criteria. The meta-analysis determined statistically significant associations for overweight with the incidence of type II diabetes, all cancers except esophageal (female), pancreatic and prostate cancer, all cardiovascular diseases (except congestive heart failure), asthma, gallbladder disease, osteoarthritis and chronic back pain. We noted the strongest association between overweight defined by body mass index (BMI) and the incidence of type II diabetes in females (RR = 3.92 (95% CI: 3.10–4.97)). Statistically significant associations with obesity were found with the incidence of type II diabetes, all cancers except esophageal and prostate cancer, all cardiovascular diseases, asthma, gallbladder disease, osteoarthritis and chronic back pain. Obesity defined by BMI was also most strongly associated with the incidence of type II diabetes in females (12.41 (9.03–17.06)). Conclusion Both overweight and obesity are associated with the incidence of multiple co-morbidities including type II diabetes, cancer and cardiovascular diseases. Maintenance of a healthy weight could be important in the prevention of the large disease burden in the future. Further studies are needed to explore the biological mechanisms that link overweight and obesity with these co-morbidities.


Background
A substantial literature has emerged which has found that overweight and obesity are major causes of co-morbidities which can lead to further morbidity and mortality [1][2][3]. The related health care costs are substantial [4][5][6]. As the number of associated co-morbidities continues to increase, systematic reviews and meta-analysis are important tools to summarize the findings and produce more precise estimates of risk associated with overweight and obesity.
The primary objective of this study is to provide a comprehensive review of the incidence of co-morbidities related to obesity and overweight. We have identified a number of recent systematic reviews and meta-analyses on type II diabetes [7], cardiovascular diseases [8,9], cancer [10], breast cancer [11,12], esophageal or cardia adenocarcinoma [13], pancreatic cancer [14] and prostate cancer [15]. The rationale for re-conducting a review is threefold. Firstly, it has been reported that abdominal obesity, defined by waist circumference (WC) measurement in comparison to the more traditional obesity definition, based on Body Mass Index (BMI) measurement, is an even better predictor of many cardiovascular diseases and type II diabetes [16][17][18][19][20][21][22][23][24]. However, most recent reviews have only focused on obesity defined by BMI instead of WC. For example, a recent meta-analysis study has compared BMI and WC as risk factors for ischaemic heart disease and stroke but it only included studies with population from the Asia Pacific region [8]. Also, only meta-analysis studies on BMI and type II diabetes have been conducted [7].
Secondly, associating the incidence of co-morbidities with overweight and obesity can be done in many ways since there are many different definitions. For instance, many previous reviews have combined studies that have found the association with per unit change of BMI (kg/m 2 ) and WC (cm) measurements [7][8][9]11,14,15]. We are interested in measuring the incidence by categorization of overweight and obesity defined by BMI and WC measurements, which has not been the focus of a majority of previous reviews.
Thirdly, the previous meta-analysis studies primarily focused on individual co-morbidities and they were conducted by different authors and using different search strategies, inclusion criteria and analysis methods. Only one recent meta-analysis study conducted by Katzmarzyk and Janssen comprehensively estimated the incidence of eight different chronic diseases associated with obesity [5]. An objective of this review is to apply a consistent methodology across all relevant co-morbidities. This enables us to compare the number of studies and size of effect across all co-morbidities.
The articles obtained from the literature search were then evaluated according to criteria set out in Figure 1. Criteria for inclusion were: prospective cohort study of the general population of a Western country (countries in Europe or North America, Australia or New Zealand), relevant outcomes, a sample size of at least 200 subjects, and risk estimate based on the incidence of disease instead of the mortality rate of disease. For large cohorts with multiple articles meeting the defined criteria, the most recent article or the article with the most usable information was used. Studies were excluded if they did not provide enough data to allow calculation of unadjusted relative risks (RRs) with 95% confidence intervals (CI) for the overweight and obese groups compared to the normal group. Data extracted for study characteristics included study design, the RRs based on WC measurements for these co-morbidities were available, they were pooled separately from those based on BMI and used as the final RRs for these comorbidities. When both IRR and RR-P estimates were available, both estimates were presented. Final results of RRs were selected based on the number of pooled studies, the duration of study follow-up and the sample size of included studies

Results
A total of 89 relevant and unique studies were identified; several studies featured for more than one co-morbidities. Of the 20 co-morbidities, 18 were identified to have at least one study meeting the inclusion criteria. Some studies were applicable to more than one co-morbidity. No studies were found for dislipidaemia and sleep apnea. The total numbers of studies included for each co-morbidity varied from 1 to 14. Reasons for exclusion are given in Table 1.
The majority of the studies were conducted in US (55%) and in European countries (40%). Study characteristics such as age criteria, study follow-up, ascertainment of exposure and outcome variables were reported by the majority of the studies. However, only a small number of studies reported sample ethnicity and of those, the majority was US studies. Among those US studies, one study (endometrial cancer) was about the black women [31] while for the remaining US studies, the proportion of whites ranged from 81% to 95%. The mean duration of study follow-up was 12.5 (SD = 7.2) years. Over half of the studies (53%) were longer than 10 years while less than 10% of the studies were shorter than 5 years. BMI and WC measurements were clinically measured on 43% of the studies and were self-reported on 56% of the studies while one study did not provide such information. Regarding the ascertainment of cases, 43 (48%) studies identified cases from registry, database centre or clinical evaluation; 34 (38%) studies were based on subject selfreported information with some kind of confirmation method such as medical records review; 6 studies were based on medical records review and 5 studies (4 for asthma and 1 for type-2 diabetes) were relying on selfreported information alone. Note that cancer cases were identified from cancer registry/database on 66% of the studies. Table 2 summarized our final results. Figures 2 to 18 presented the detailed results including study-specific and pooled estimates. Results from the meta-analysis were summarized in the following sections for each co-morbidity. Articles meeting inclusion criteria and information extracted for metaanalysis: n = E Articles included in final metaanalysis calculations: n = G Articles not amenable to meta-analysis: n = F -Required information not provided: n = F (i) -Same cohort already used: n = F (ii)

Endometrial cancer
In total, 10 studies were identified to meet the inclusion criteria for endometrial cancer ( Figure  3) [19,31,35,36,42,[45][46][47][48][49]. The study on the US black women was not included in the final result as it showed systematic difference from other studies mainly on Cauca-sian population [31].

Kidney cancer
We identified 5 studies meeting the inclusion criteria relating overweight and obesity to kidney cancer ( Figure  7)

Pancreatic cancer
The search identified 6 studies giving information on the risk of pancreatic cancer attributable to overweight and obesity ( Figure 8)

Prostate cancer
The search identified 8 studies giving information on the risk of prostate cancer attributable to overweight and obesity ( Figure 9)

Coronary Artery Disease
Eleven studies were identified with evidence for coronary artery disease related to obesity (Figures 13 and 14) [20,21,79,91,[93][94][95][96][97][98][99]. The pooled RR-P estimates for coronary artery disease across categories of WC were more conservative RR estimates for men than the corresponding IRR estimates and thus were presented in the summary

Asthma
Four studies were identified with evidence for asthma related to obesity ( Figure 16)

Chronic back pain
Only 1 study was identified to meet the inclusion criteria for chronic back pain [107]. The study identified the association for the overweight and obesity with early retirement due to chronic back pain.

Osteoarthritis
We identified three studies meeting the inclusion criteria relating overweight and obesity to osteoarthritis ( Figure  17) [108][109][110]. The studies identified the risk of joint replacement attributable to being overweight and obese.

Meta-analysis of studies for coronary artery disease-males
Our sensitivity analyses showed that our results were in general robust with the following exceptions. For ovarian cancer, associations for both overweight and obesity were slightly weaker in US studies compared to European studies. Similar country differences were found in pancreatic cancers; in addition, weaker associations were observed in older population. Studies with shorter follow-up time and of older population showed slightly weaker association of obesity with prostate cancer. In coronary artery disease for females, studies with shorter follow-up showed weaker associations of both overweight and obese. In coronary artery disease for males, weaker associations were observed in US and Canadian studies. Studies of post-menopausal and senior women on congestive heart failure showed weaker associations for both overweight and obesity.

Discussion
We have comprehensively reviewed 20 co-morbidities for high quality cohort studies which determine risk factors associated with overweight or obesity. 18 co-morbidities were identified and meta-analysis was performed where at least 1 study was found. A summary of the results can be found in Table 2.
There are a number of alternative meta-analyses with which we can compare our results. . However, each study uses different definitions of overweight and obesity, includes varying quality of study designs, uses different methods for meta-analysis and ultimately only focuses on individual co-morbidities. Hence, the objective of our study is not only to provide up to date estimates of the risk of all possible co-morbidities attributable to overweight and obesity, but also to do it using consistent definitions and methodology.
In assessing whether obesity is related to a given co-morbidity, the occurrence timing of co-morbidities with respect to exposure of obesity is important in determining the causal pathway. Therefore, we included only the pro-Meta-analysis of studies for congestive heart failure Figure 15 Meta-analysis of studies for congestive heart failure. *Q-statistic(p-value); **post-menopause; square shape: study-and gender-specific risk estimates; diamond shape: pooled risk estimates. Meta-analysis of studies for osteoarthritis Figure 17 Meta-analysis of studies for osteoarthritis. *Q-statistic(p-value); F-up is follow-up in years; square shape: study-and gender-specific risk estimates; diamond shape: pooled risk estimates. All studies (2) All studies (2) All studies (1) All studies (2) Meta-analysis of studies for asthma Figure 16 Meta-analysis of studies for asthma. *Q-statistic(p-value); F-up is follow-up in years; square shape: study-and gender-specific risk estimates; diamond shape: pooled risk estimates.  US/Canada studies (3) F-up>=10 yrs (2) F-up<10 yrs (2) All studies (4) All studies (1) US/Canada studies (2) F-up>=10 yrs (2) F-up<10 yrs (1) All studies (3) All studies (1)  spective cohort studies and excluded the cross-sectional studies and case control studies to minimize the associated potential biases. In addition, WC measurements were considered to be the better risk predictor for type II diabetes, hypertension, coronary artery disease, congestive heart failure, stroke and gallbladder disease [16][17][18][19][20][21][22][23][24]. In our studies, the risk for type II diabetes, female hypertension, coronary artery disease, and male gallbladder disease were estimated based on WC measurements.
Some limitations are worthy of further consideration. Firstly, other variables not included in our analysis might potentially confound our results. Most important is the exclusion of the level of physical inactivity which is a known risk factor for some co-morbidities and related to overweight and obesity [127]. Physical inactivity is often poorly reported and requiring its inclusion would have reduced the number of included studies. Secondly, for certain co-morbidities, we only identified 1 or 2 prospective cohort studies that adopted the WC measurements as the risk predictor. Further studies are required to determine the association between WC and some co-morbidities before an estimate of the risk can be calculated through a meta-analysis. Thirdly, given the sizable literature and that we were searching for non RCT studies for which search filters are more complex, we determined to use Medline and Embase as the electronic databases, and complement the search with checking reference lists and thorough searching the internet. We did not search other databases such as CINHAL, HealthSTAR, AMED, and BIO-SIS. Therefore, bias might have occurred due to our search strategy. However, given the nature of the studies we are looking for, i.e., prospective cohort studies with high quality, we consider our search within Medline and Embase sufficient. Lastly, due to the small number of studies for most co-morbidities, assessment of potential publication bias was infeasible. However, we did not find evidence of publication bias in those meta-analyses where the number of studies was relatively large except for prostate cancer.

Conclusion
In conclusion, this study provides a comprehensive estimate of the incidence of 18 co-morbidities attributable to overweight and obesity using standardized and consistent definitions and methodologies. Our findings confirm that overweight and obesity carry a profound health burden and will have a significant impact on health expenditures.

Competing interests
This manuscript is part of larger project funded by sanofiaventis Canada Inc.
Meta-analysis of studies for gallbladder disease Figure 18 Meta-analysis of studies for gallbladder disease. *Q-statistic(p-value); F-up is follow-up in years; square shape: study-and gender-specific risk estimates; diamond shape: pooled risk estimates. All studies (3) All studies (1) All studies (2) US studies (2) All studies (3) All studies (1) All studies (2) All studies (1) All studies (1)