IJC International Journal of Cancer
Circulating adiponectin, leptin and adiponectin–leptin ratio and endometrial cancer risk: Evidence from a meta-analysis of epidemiologic studies Ting-Ting Gong1, Qi-Jun Wu2, Yong-Lai Wang1 and Xiao-Xin Ma1 1 2
Department of Obstetrics and Gynecology, Shengjing Hospital of China Medical University, Shenyang, Liaoning, People’s Republic of China Division of Clinical Epidemiology, Shengjing Hospital of China Medical University, Shenyang, Liaoning, People’s Republic of China
Endometrial cancer is the second most common gynecologic malignancy worldwide, with approximately 0.3 million new cases in 2012.1 There is no doubt that obesity is a wellrecognized risk factor for this disease. Recently, the biological mechanisms linking obesity and endometrial cancer have largely centered on the aromatization of androgens in adipose tissue, leading to increased circulating estradiol levels2 which have been the primary stimulants of endometrial proliferation.3 However, this mechanism cannot fully explain the incidence of endometrial cancer in obese women. Experimental studies have provided evidence that adipokines, which are
Key words: adiponectin, endometrial neoplasm, leptin, meta-analysis, obesity Additional Supporting Information may be found in the online version of this article. Grant sponsor: The Younger research fund of Shengjing Hospital; Grant number: 2014sj09; Grant sponsor: The National Natural Science Foundation of China; Grant numbers: 81472438, 81272874 DOI: 10.1002/ijc.29561 History: Received 1 Feb 2015; Accepted 2 Apr 2015; Online 9 Apr 2015 Correspondence to: Qi-Jun Wu, Division of Clinical Epidemiology, Shengjing Hospital of China Medical University, Address: No. 36, San Hao Street, Shenyang, Liaoning 110004, People’s Republic of China, Tel.: 86-24-96615-13648, E-mail: [email protected]
C 2015 UICC Int. J. Cancer: 137, 1967–1978 (2015) V
associated with hyperinsulinemia and degree of insulin resistance independent of adiposity, are also involved in the development of endometrial cancer. Adipose tissue was previously considered to be an energystorage site. During recent decades, this tissue has been found to be an endocrine organ producing and secreting several bioactive peptides, including adipokines such as adiponectin and leptin.4 Biological studies have shown that adiponectin, the most abundant adipokine, is not only inversely associated with obesity as well as hyperinsulinemia5,6 but appears to have anti-inflammatory, antiangiogenic and antidiabetic properties7,8; In contrast, the biological functions of leptin are involved in the cell proliferation, angiogenesis and metastasis in several cell lines.9 Furthermore, as increased adipocytes produce more adiponectin and less leptin, the circulating adiponectin–leptin (A/L) ratio or leptin–adiponectin (L/A) ratio has been suggested to be a useful, surrogate marker of insulin resistance which characterizes both adiponectin and leptin.10 To date, several studies have suggested that adiponectin, leptin and A/L ratio were associated with obesity-associated malignancies (e.g., breast, colorectal and pancreatic cancer).11–15 Evidence from epidemiologic studies investigating the relationship between circulating adiponectin and leptin concentrations and endometrial cancer risk has not been summarized.16–29 Additionally, limited evidence from observational studies evaluating the A/L or L/A ratio and risk of endometrial cancer has been available to date.16,19,21,22,24
Epidemiology
We performed this meta-analysis of epidemiologic studies to investigate the associations between circulating adiponectin, leptin and adiponectin–leptin (A/L) ratio and endometrial cancer risk. Relevant manuscripts were identified by searching PubMed and ISI Web of Science databases as well as by manual searching the references cited in retrieved manuscripts. Random-effects models were used to estimate summary odds ratio (SOR) and 95% confidence intervals (CIs) for aforementioned associations. Fourteen manuscripts with 13 studies (five nested case–control and eight case–control studies) cumulatively involving a total of 1,963 endometrial cancer cases and 3,503 noncases were included in the analyses. Overall, comparing persons with circulating concentrations of adiponectin, leptin and A/L ratio in the top tertile with persons with concentrations of these biomarkers in the bottom tertile yielded SORs of 0.47 (95% CI: 0.34–0.65; I2 5 63.7%; n 5 13), 2.19 (95% CI: 1.44–3.31; I2 5 64.2%; n 5 7),and 0.45 (95% CI: 0.24–0.86; I2 5 90.1%; n 5 5), respectively. Notably, there was an 18% reduction in risk for per each 5 lg/mL increment in circulating adiponectin concentrations (SOR 5 0.82; 95% CI: 0.74–0.90; I2 5 49%; n 5 8). Stratifying by study characteristics and whether these studies considered or adjusted for potential confounders, the findings were robust in the analyses of circulating adiponectin and leptin. No evidence of publication bias was detected. In conclusion, the findings from this metaanalysis suggest that increased circulating adiponectin and A/L ratio or decreased leptin concentrations were associated with reduced risk of endometrial cancer. Further prospective designed studies are warranted to confirm our findings.
1968
Adipokine-related biomarker and endometrial cancer
What’s new? How does obesity confer an increase in endometrial cancer risk? Some studies have suggested that adipokines could be involved, in particular, adiponectin, leptin, and the ratio between the two. Adiponectin is generally associated with anti-diabetic and anti-inflammatory actions, while leptin contributes to cell proliferation and metastasis. In this study, the authors collected the existing literature on adiponectin, leptin, and endometrial cancer risk, and analyzed the combined data. They conclude that increased circulating adiponectin or decreased leptin, or a higher A/L ratio, corresponds to a reduced risk of endometrial cancer.
Therefore, we carried out a meta-analysis of observational studies to systematically assess the overall aforementioned exposures with endometrial cancer risk.
Methods Literature search
This meta-analysis was planned, conducted and reported in adherence to the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) guidelines.30 We comprehensively searched the published manuscripts without restriction from the following databases: PubMed and ISI Web of Science. The following key words combined with their corresponding MeSH terms were used: (adipokines, adiponectin and leptin) and (endometrial and endometrium) and (cancer, carcinoma and neoplasm). Additionally, the references cited in retrieved articles were scrutinized by manual search.
Epidemiology
Study selection and exclusion criteria
The published manuscripts were included by the following selection criteria: (i) the study used an epidemiologic study design (e.g., case–control, case–cohort, nested case–control and cohort study) (n 5 52); (ii) the study provided information on circulating adiponectin, leptin concentrations or A/L ratio as the exposure of interest (n 5 44); (iii) the study reported endometrial cancer as the outcome of interest (n 5 137) and (iv) the study reported the usable risk estimates (e.g., odds ratio, risk ratio or relative risk with 95% confidence intervals [CIs] or necessary data for calculation) between circulating adiponectin, leptin level or A/L ratio and endometrial cancer risk (n 5 16). In addition, if more than one study was conducted in the same population, the most recent report or the report with the most applicable estimates was selected for our analysis (n 5 2). The published manuscripts were excluded by the following exclusion criteria: (i) nonepidemiologic studies (e.g., case– control, case–cohort, nested case–control and cohort study) (n 5 292), reviews without original data (n 5 7), ecological studies (n 5 10), editorials (n 5 4) and case reports (n 5 3); (ii) the study reported the risk estimates that could not be summarized (such as reported the risk estimates without 95% CIs) (n 5 8) and (iii) the study reported exclusively on endometrial cancer mortality (n 5 0). All abovementioned study selection and exclusion procedures were carried out by two independent investigators (T.T.G. and Q-.J.W.).
Data abstraction
Two independent investigators (T-.T.G. and Q-.J.W.) carried out the data abstraction by a predefined database. The disagreements of these two databases were resolved by joint reevaluation. The following key study characteristics were abstracted from each included study: the first author’s name, publication year, study region, study design, numbers of cases and controls, assay method, categories of exposure with corresponding risk estimates, the factors matched between cases and controls and potential confounders adjusted in the analysis. Statistical analysis
As the absolute risk of endometrial cancer is low and all included studies were nested case–control or case–control studies, and therefore we interpreted all risk estimates as odds ratio (OR) for simplicity. We converted the reported risk estimates into a standard scale of effect to compare persons with the levels of exposure in the top tertile with persons with those in the bottom tertile. This scaling method was first used by Danesh et al.,31 in essence giving an estimate per 2.18 standard deviation units of exposure, where 2.18 is the difference in the means of the highest and lowest tertile of the standard normal distribution. When the log OR for the comparison of the top and bottom tertiles of adiponectin or leptin was not directly available from the published report,14,20,23,26, it was estimated as (i) a scaling factor of 2.18 divided by 1.59 times the log OR for comparison of the top and bottom halves, (ii) a scaling factor of 2.18 divided by 2.54 times the log OR for comparison of the top and bottom quartiles or (iii) a scaling factor of 2.18 times the log OR for a 1 standard deviation difference in the baseline levels of adiponectin or leptin. More details about this method could be obtained elsewhere. Additionally, for the studies that did not use the lowest category of adiponectin as the reference,14,31,32 as well as reported the L/A ratio rather than the A/L ratio as the exposure,21,24 we used the effective-count method proposed by Hamling et al.33 to recalculate the ORs and 95% CIs. To examine the associations between the exposure of interest (i.e., adiponectin, leptin and A/L ratio) and endometrial cancer risk, the summary odds ratio (SOR) with 95% CIs were estimated by summarizing the risk estimates of each study using the random effect models, which considered both within- and between-study variation.34 We first C 2015 UICC Int. J. Cancer: 137, 1967–1978 (2015) V
1969
Gong et al.
C 2015 UICC Int. J. Cancer: 137, 1967–1978 (2015) V
Heterogeneity among studies was assessed with I2 statistics, and the results were defined as heterogeneous for an I2 value of >50%.40 Small-study bias, such as publication bias, was tested by visual inspection of funnel plot and testing by using Egger’s test41 and Begg’s test.42 Two-sided p-values of
Circulating adiponectin, leptin and adiponectin–leptin ratio and endometrial cancer risk: Evidence from a meta-analysis of epidemiologic studies Ting-Ting Gong1, Qi-Jun Wu2, Yong-Lai Wang1 and Xiao-Xin Ma1 1 2
Department of Obstetrics and Gynecology, Shengjing Hospital of China Medical University, Shenyang, Liaoning, People’s Republic of China Division of Clinical Epidemiology, Shengjing Hospital of China Medical University, Shenyang, Liaoning, People’s Republic of China
Endometrial cancer is the second most common gynecologic malignancy worldwide, with approximately 0.3 million new cases in 2012.1 There is no doubt that obesity is a wellrecognized risk factor for this disease. Recently, the biological mechanisms linking obesity and endometrial cancer have largely centered on the aromatization of androgens in adipose tissue, leading to increased circulating estradiol levels2 which have been the primary stimulants of endometrial proliferation.3 However, this mechanism cannot fully explain the incidence of endometrial cancer in obese women. Experimental studies have provided evidence that adipokines, which are
Key words: adiponectin, endometrial neoplasm, leptin, meta-analysis, obesity Additional Supporting Information may be found in the online version of this article. Grant sponsor: The Younger research fund of Shengjing Hospital; Grant number: 2014sj09; Grant sponsor: The National Natural Science Foundation of China; Grant numbers: 81472438, 81272874 DOI: 10.1002/ijc.29561 History: Received 1 Feb 2015; Accepted 2 Apr 2015; Online 9 Apr 2015 Correspondence to: Qi-Jun Wu, Division of Clinical Epidemiology, Shengjing Hospital of China Medical University, Address: No. 36, San Hao Street, Shenyang, Liaoning 110004, People’s Republic of China, Tel.: 86-24-96615-13648, E-mail: [email protected]
C 2015 UICC Int. J. Cancer: 137, 1967–1978 (2015) V
associated with hyperinsulinemia and degree of insulin resistance independent of adiposity, are also involved in the development of endometrial cancer. Adipose tissue was previously considered to be an energystorage site. During recent decades, this tissue has been found to be an endocrine organ producing and secreting several bioactive peptides, including adipokines such as adiponectin and leptin.4 Biological studies have shown that adiponectin, the most abundant adipokine, is not only inversely associated with obesity as well as hyperinsulinemia5,6 but appears to have anti-inflammatory, antiangiogenic and antidiabetic properties7,8; In contrast, the biological functions of leptin are involved in the cell proliferation, angiogenesis and metastasis in several cell lines.9 Furthermore, as increased adipocytes produce more adiponectin and less leptin, the circulating adiponectin–leptin (A/L) ratio or leptin–adiponectin (L/A) ratio has been suggested to be a useful, surrogate marker of insulin resistance which characterizes both adiponectin and leptin.10 To date, several studies have suggested that adiponectin, leptin and A/L ratio were associated with obesity-associated malignancies (e.g., breast, colorectal and pancreatic cancer).11–15 Evidence from epidemiologic studies investigating the relationship between circulating adiponectin and leptin concentrations and endometrial cancer risk has not been summarized.16–29 Additionally, limited evidence from observational studies evaluating the A/L or L/A ratio and risk of endometrial cancer has been available to date.16,19,21,22,24
Epidemiology
We performed this meta-analysis of epidemiologic studies to investigate the associations between circulating adiponectin, leptin and adiponectin–leptin (A/L) ratio and endometrial cancer risk. Relevant manuscripts were identified by searching PubMed and ISI Web of Science databases as well as by manual searching the references cited in retrieved manuscripts. Random-effects models were used to estimate summary odds ratio (SOR) and 95% confidence intervals (CIs) for aforementioned associations. Fourteen manuscripts with 13 studies (five nested case–control and eight case–control studies) cumulatively involving a total of 1,963 endometrial cancer cases and 3,503 noncases were included in the analyses. Overall, comparing persons with circulating concentrations of adiponectin, leptin and A/L ratio in the top tertile with persons with concentrations of these biomarkers in the bottom tertile yielded SORs of 0.47 (95% CI: 0.34–0.65; I2 5 63.7%; n 5 13), 2.19 (95% CI: 1.44–3.31; I2 5 64.2%; n 5 7),and 0.45 (95% CI: 0.24–0.86; I2 5 90.1%; n 5 5), respectively. Notably, there was an 18% reduction in risk for per each 5 lg/mL increment in circulating adiponectin concentrations (SOR 5 0.82; 95% CI: 0.74–0.90; I2 5 49%; n 5 8). Stratifying by study characteristics and whether these studies considered or adjusted for potential confounders, the findings were robust in the analyses of circulating adiponectin and leptin. No evidence of publication bias was detected. In conclusion, the findings from this metaanalysis suggest that increased circulating adiponectin and A/L ratio or decreased leptin concentrations were associated with reduced risk of endometrial cancer. Further prospective designed studies are warranted to confirm our findings.
1968
Adipokine-related biomarker and endometrial cancer
What’s new? How does obesity confer an increase in endometrial cancer risk? Some studies have suggested that adipokines could be involved, in particular, adiponectin, leptin, and the ratio between the two. Adiponectin is generally associated with anti-diabetic and anti-inflammatory actions, while leptin contributes to cell proliferation and metastasis. In this study, the authors collected the existing literature on adiponectin, leptin, and endometrial cancer risk, and analyzed the combined data. They conclude that increased circulating adiponectin or decreased leptin, or a higher A/L ratio, corresponds to a reduced risk of endometrial cancer.
Therefore, we carried out a meta-analysis of observational studies to systematically assess the overall aforementioned exposures with endometrial cancer risk.
Methods Literature search
This meta-analysis was planned, conducted and reported in adherence to the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) guidelines.30 We comprehensively searched the published manuscripts without restriction from the following databases: PubMed and ISI Web of Science. The following key words combined with their corresponding MeSH terms were used: (adipokines, adiponectin and leptin) and (endometrial and endometrium) and (cancer, carcinoma and neoplasm). Additionally, the references cited in retrieved articles were scrutinized by manual search.
Epidemiology
Study selection and exclusion criteria
The published manuscripts were included by the following selection criteria: (i) the study used an epidemiologic study design (e.g., case–control, case–cohort, nested case–control and cohort study) (n 5 52); (ii) the study provided information on circulating adiponectin, leptin concentrations or A/L ratio as the exposure of interest (n 5 44); (iii) the study reported endometrial cancer as the outcome of interest (n 5 137) and (iv) the study reported the usable risk estimates (e.g., odds ratio, risk ratio or relative risk with 95% confidence intervals [CIs] or necessary data for calculation) between circulating adiponectin, leptin level or A/L ratio and endometrial cancer risk (n 5 16). In addition, if more than one study was conducted in the same population, the most recent report or the report with the most applicable estimates was selected for our analysis (n 5 2). The published manuscripts were excluded by the following exclusion criteria: (i) nonepidemiologic studies (e.g., case– control, case–cohort, nested case–control and cohort study) (n 5 292), reviews without original data (n 5 7), ecological studies (n 5 10), editorials (n 5 4) and case reports (n 5 3); (ii) the study reported the risk estimates that could not be summarized (such as reported the risk estimates without 95% CIs) (n 5 8) and (iii) the study reported exclusively on endometrial cancer mortality (n 5 0). All abovementioned study selection and exclusion procedures were carried out by two independent investigators (T.T.G. and Q-.J.W.).
Data abstraction
Two independent investigators (T-.T.G. and Q-.J.W.) carried out the data abstraction by a predefined database. The disagreements of these two databases were resolved by joint reevaluation. The following key study characteristics were abstracted from each included study: the first author’s name, publication year, study region, study design, numbers of cases and controls, assay method, categories of exposure with corresponding risk estimates, the factors matched between cases and controls and potential confounders adjusted in the analysis. Statistical analysis
As the absolute risk of endometrial cancer is low and all included studies were nested case–control or case–control studies, and therefore we interpreted all risk estimates as odds ratio (OR) for simplicity. We converted the reported risk estimates into a standard scale of effect to compare persons with the levels of exposure in the top tertile with persons with those in the bottom tertile. This scaling method was first used by Danesh et al.,31 in essence giving an estimate per 2.18 standard deviation units of exposure, where 2.18 is the difference in the means of the highest and lowest tertile of the standard normal distribution. When the log OR for the comparison of the top and bottom tertiles of adiponectin or leptin was not directly available from the published report,14,20,23,26, it was estimated as (i) a scaling factor of 2.18 divided by 1.59 times the log OR for comparison of the top and bottom halves, (ii) a scaling factor of 2.18 divided by 2.54 times the log OR for comparison of the top and bottom quartiles or (iii) a scaling factor of 2.18 times the log OR for a 1 standard deviation difference in the baseline levels of adiponectin or leptin. More details about this method could be obtained elsewhere. Additionally, for the studies that did not use the lowest category of adiponectin as the reference,14,31,32 as well as reported the L/A ratio rather than the A/L ratio as the exposure,21,24 we used the effective-count method proposed by Hamling et al.33 to recalculate the ORs and 95% CIs. To examine the associations between the exposure of interest (i.e., adiponectin, leptin and A/L ratio) and endometrial cancer risk, the summary odds ratio (SOR) with 95% CIs were estimated by summarizing the risk estimates of each study using the random effect models, which considered both within- and between-study variation.34 We first C 2015 UICC Int. J. Cancer: 137, 1967–1978 (2015) V
1969
Gong et al.
C 2015 UICC Int. J. Cancer: 137, 1967–1978 (2015) V
Heterogeneity among studies was assessed with I2 statistics, and the results were defined as heterogeneous for an I2 value of >50%.40 Small-study bias, such as publication bias, was tested by visual inspection of funnel plot and testing by using Egger’s test41 and Begg’s test.42 Two-sided p-values of
