|Ahead of print publication
The association between diabetes and cancer in Mexico: Analysis using death certificate databases, 2009-2017
Administration and Personnel Development Department, Social Security Institute of the State of Mexico and Municipalities, Toluca de Lerdo, Estado de México, México
|Date of Submission||17-Oct-2019|
|Date of Decision||16-Dec-2019|
|Date of Acceptance||27-Jan-2020|
|Date of Web Publication||03-Oct-2020|
Administration and Personnel Development Department, Social Security Institute of the State of Mexico and Municipalities, Constituyentes #703, Col. Barrio De La Merced, CP 50080, Toluca De Lerdo, Estado De México
Source of Support: None, Conflict of Interest: None
Context: Previous studies have shown that diabetes mellitus (DM) is a risk factor of some type-specific cancers. However, no data are available on the association between cancer and DM in Latin America.
Aims: The aim of this study is to determine which type-specific cancers are associated with DM using multiple cause of death data.
Settings and Design: Whole country of Mexico, cross-sectional design.
Materials and Methods: Analysis of all cancer deaths (2009–2017) using death certificate databases of Mexican adults aged ≥20 years.
Statistical Analysis Used: Multivariable logistic regression.
Results: There were 710,292 total cancer deaths. DM increased the risk of pancreatic (adjusted odds ratio [aOR] = 1.7), liver (aOR = 1.6), kidney (aOR = 1.4), gallbladder (aOR = 1.2) and endometrial (aOR = 1.1) cancers, all P < 0.05. Type 2 or unknown-type DM were associated with the same cancer types with little variation of estimates. Higher estimates were found in males than females (except for kidney cancer). Type 1 DM was associated with pancreatic cancer only (aOR = 1.9).
Conclusions: DM in Mexico is associated with gastrointestinal (pancreatic, liver, gallbladder), kidney and endometrial cancers. Dissemination of knowledge to both health-care workers and diabetics regarding potential cancer risks including adequate diet, regular exercise, weight reduction if obese/overweight, cessation of smoking, and good glucose control and medication compliance should be reinforced. Specific cancer preventative measures should be implemented for patients with DM.
Keywords: Diabetes, death certificate, type-specific cancers
| > Introduction|| |
Diabetes and cancer are conditions whose incidence is increasing globally. An overwhelming number of studies on the association between diabetes mellitus (DM) and cancer have been published since 1959. Case–control or cohort studies have consistently shown that DM (primarily type 2 DM [T2DM]) increases the risk of liver, pancreatic, endometrial,, colorectal and breast ,, cancers. The evidence for other types of cancer is less consistent and is controversial for bladder cancer., Patients with DM seem to have a reduced risk for prostate cancer.,
Gender differences must be considered when analyzing DM' impact on cancer as gender-specific risks exist. Furthermore, the strength of the association between DM and specific cancer types varies by race.
Determining what specific cancer types are more closely associated with DM locally is important for targeting specific preventive measures for priority cancers. No previous population-based studies from Latin-America have determined the association between DM and type-specific cancer by analyzing multiple cause-of-death data using death certificates. The objective of this study was to determine which type-specific cancers are associated with DM among Mexicans.
| > Materials and Methods|| |
This population-based study analyzed the Mexican de-identified death certificate databases of patients with cancer aged ≥20 years who died between 2009 and 2017. The “Instituto Nacional de Estadística y Geografía” provided the death certificates' electronic files. Mexican guidelines are followed for the completion of death certificates. In the causes of death section, the diagnosis listed includes the cause of death and any other significant pathological states that contributed to death or were unrelated to the direct or underlying causes of death. Trained codifiers enter the International Classification of Diseases 10th edition (ICD-10th) codes for the cause of death and all other diagnoses/comorbidities listed in death certificates. The study population included only those with malignant neoplasms (ICD-10th codes from C00 to C97).
During the study period, the most frequent cancer types in Mexico included in this study (any mention in death certificate's diagnosis section and not only as cause of death) were in order of frequency: lung (ICD-10 code C34), prostate (C61), breast (C50), liver (C22), stomach (C16), colorectal (C18-C20), cervix (C53), pancreas (C25), leukemia (C91-C95), and ovary (C56). Other cancer types already known to be associated with DM  but not among the top ten types in Mexico were also analyzed and included kidney (C64) bladder (C67), oral cavity (C00-C10), gallbladder (C23, C24), and endometrial (C54, C55). Separated indicator variables were obtained for DM defined when any diagnosis was coded E10-E14 (for All-DM), E10 (for T1DM), E11 (for T2DM), and E12-E14 (for all other/unspecified DM type or unknown-type-DM) to determine their independent effect on each type-specific cancer in the multivariable analysis. The objective of this study was to establish the association between specific cancer types and DM and not whether cancer or DM were the underlying or direct cause of death.
Other variables in the database included in the multivariable analysis were gender (in both sexes analysis), age, year, and place of death (home, hospital/clinic, public places, other placer vs. other), level of education, type of Mexican public health-care coverage: Instituto Mexicano del Seguro Social (IMSS), Instituto de Seguridad y Servicios Sociales de los Trabajadores del Estado (ISSSTE), Petróleos Mexicanos (PEMEX), Secretaría de la Defensa Nacional y Marina (SEDENA/Marina), Seguro Popular (SP), other type of health insurance and a group with no health insurance.
Descriptive results were obtained for specific characteristics by gender and by cancer type. Multivariable logistic regression models were then built using the backward elimination procedure with the significance level set at 0.2 to be allowed variables in the final model. Logistic models were built separately for each cancer type, in females and males and jointly with the outcome variable being cancer-specific type versus all other cancers grouped together (reference group).
To determine the effect of DM, multivariable models were adjusted by the following covariates: Age, gender, level of education (grouped into “none” vs. “at least some elementary school”), occupation (working vs. not working), place of death (separate indicator variables for hospital, home vs. other places as the reference group), zone of residence (urban vs. rural) type of public health-care coverage (indicator variables for IMSS, ISSSTE, SP, PEMEX, SEDENA/Marina and other types of health insurance vs. no health insurance). When a specific cancer type was associated with DM, that type was removed from the next multivariable logistic regression model of the next associated cancer type. The statistical analysis was performed with SAS version 9.4 (Statistical Analysis System, Cary, NC, USA). Ethics approval was not required as the study was based on deidentified routine data.
| > Results|| |
A total of 710,292 with cancer recorded in any of the diagnoses listed in the death certificate represented 13.8% of all deaths (data not shown). About half of cancer deaths were in females (50.8%). The mean age (standard deviation) at the time of death was 65.7 ± 15.9, 64.1 ± 15.8, 67.4 ± 15.8 for both sexes, females and males, respectively. Most were between the age range 56–82 years at the time of death (63.4% males, 58% females). Females with cervical, ovarian cancers or leukemia died at younger age (mean age of 59.7, 59.7, and 55.2 years, respectively) as did males with leukemia (mean age 54.8 years), colorectal, kidney (64.8 years), stomach and oral cavity (65.8 years) or “other types” cancers grouped together (62.8%), data not shown. Most people had some level of education (74.8%), females were mostly unemployed (87.5%) compared to males (31.8%) and most lived in urban areas (80.8%) or died at home (56.3%), [Table 1], [Table 2], [Table 3] with the exception of those with leukemia or endometrial cancer who were more likely to die in hospitals/clinics (data not shown). Diabetes prevalence at death in the whole sample statistically increased over time from 11.7% in 2009 to 13.6% in 2017 (p-by trend <0.0001, data not shown). Diabetes prevalence for the whole period in females and males was, respectively, 13.6% and 11.9% for All-DM, 0.16% and 0.13% for T1DM, 9.5% and 8.2% for T2DM, and 4.8% and 4.2% for unknown-type-DM. Cancer types with DM rates higher than the overall DM rate (12.7%) included pancreas (19.2%), liver (17.4%), kidney (15.5%), gallbladder (15%), and endometrial (15.1%), [Table 1].
|Table 1: Percentages of specific characteristics for each cancer type in both sexes aged ≥20, years 2009-2017*|
Click here to view
|Table 2: Percentages of specific characteristics for each cancer type in females aged ≥20, years 2009-2017*|
Click here to view
|Table 3: Percentages of specific characteristics for each cancer type in males aged ≥20, years 2009-2017*|
Click here to view
After controlling by age, gender, level of education, zone of residence, place of death and type of health insurance, the multivariable logistic regression analysis showed All-DM to be an independent risk factor of pancreatic (adjusted odds ratio [aOR] = 1.74, 95% confidence interval [CI] 1.69–1.79), liver (aORboth = 1.58, 95% CI 1.54–1.62), kidney (aORboth = 1.37, 95% CI 1.31–1.43), and gallbladder (aORboth = 1.23, 95% CI 1.15–1.33) and endometrial cancers (aOR = 1.15, 95% CI 1.06–1.25). T2DM and unknown-type-DM were associated with the same cancer types with little variation on estimates. T1DM was associated with pancreatic cancer only (aORboth = 1.93, 95% CI 1.52–2.45). Analysis by gender showed mostly stronger effect of DM in males than females (except for kidney cancer), all P < 0.001, [Table 4].
|Table 4: Odds ratio (95% confidence interval)* of diabetes with each cancer type in females, males and jointly|
Click here to view
No association was found between DM and other cancer types.
| > Discussion|| |
To our knowledge, this is the first Mexican study exploring the association between DM and type-specific cancers. This population-based study analyzed all cancer deaths occurring in Mexico from 2009 to 2017, found that All-DM, T2DM or unspecified DM were associated with pancreatic, liver, kidney, and gallbladder cancers in the whole sample and both sexes. Diabetes was associated with endometrial cancer in females.
Diabetes conferred greater risk for pancreatic, liver, and gallbladder cancers in males than females and slightly lower risk for kidney cancer. Findings by gender for kidney and liver cancers show the same trend as those of a recent review where DM had a significantly greater risk in women than men for kidney cancer but a lower relative risk for liver cancer.
Increasing trend in DM prevalence found in this study is in agreement with previous national health surveys in Mexico. The overall prevalence of All-DM in the study population of cancer patients at the time of death was 12.7% which is lower than the 14.4% estimate among Mexican Americans based on hemoglobin A1c level, fasting plasma glucose level, or 2-h plasma glucose level in 2011–2012 or than the 14.8% IDF estimate in those aged 20–79 years in 2017 for all of Mexico. Our lower DM prevalence estimate may be explained by the inclusion of those older than 79 years and the bigger sample size when compared to the aforementioned studies. Comparison of DM prevalence estimates and the consistency of the association between DM type and type-specific cancers by gender indicate that DM underreporting in our study is minimal and that it is systematically searched/asked by physicians and/or relatives of the deceased at the time of death for completion of death certificates.
Type 1 DM was associated with pancreatic cancer only with higher estimates in the whole sample and males when compared to All-DM's estimates. A previous study showed T1DM to be associated with pancreatic, liver, esophagus, colon and rectum, stomach, thyroid, brain, lung, ovary, and endometrium cancers in females. In males, the risk generally increased for the same cancers, but fewer were significant, and a decreased risk for prostate cancer was also observed  due in part to reduced levels of circulating testosterone.
The small sample size for T1DM in our study likely represents a limitation for determining the association with other cancer types. Some patients labeled as “Unknown-type-DM” may have had T1DM, but it is likely that many had T2DM as it is the most common type accounting for about 90%–95% of prevalent cases. T2DM represented 67% and the “Unknown-type-DM” group, 32%, of our cases, suggesting that many patients in latter group were T2DM also because both groups (T2DM and unknown DM type) were associated with the same cancer types with the exception of endometrial cancer which was not associated with the unknown DM type.
Consistent with previous studies, there was a higher prevalence of DM among pancreatic cancer patients., The finding of DM conferring the highest risk for pancreatic cancer as compared to other cancer types has also been previously reported. Our overall OR estimate for All-DM (1.74) is just below the 1.82–1.94 range found previously,, but the estimates in males with All-DM (1.85), T1DM in both sexes and males (1.93 and 2.39, respectively) or unknown-type-DM (both sexes: 1.85 or males: 1.95) are within this range.
DM has been proposed to be a risk factor for and a consequence of pancreatic cancer (reverse causality). Unfortunately, the time between DM diagnosis and the development of cancer was not determined, but a recent study showed individuals with recent-onset DM (within three or fewer years of pancreatic cancer diagnosis) have a greater risk compared with those with long-term DM across all ages. The risk was also higher in Latinos as compared to other races suggesting that Latinos/Mexicans are more susceptible for pancreatic cancer from recent-onset DM than Whites or African Americans.
Currently, there are no adequate population biomarkers for pancreatic cancer, but candidates include CA19-9 and CancerSEEK  both of which have limitations. If T2DM could be physiologically distinguished from pancreatic cancer, related pancreatogenic DM or T3cDM, might provide some degree of cancer risk prediction. If chronic pancreatitis is ruled out as a risk factor for newly diagnosed DM, then low circulating pancreatic polypeptide (PP) levels or lack of PP response to a mixed-food meal could suggest a pancreatic cancer etiology of the DM  which would likely be worth further screening with CA19-9.
Mediators proposed as having the pro-carcinogenic effect of DM on pancreatic cancer development include insulin resistance, compensatory increased levels of insulin, insulin-like growth factor-1 (IGF-1), and sustained hyperglycemia. A recent study showed that weight loss and normal blood glucose levels before DM onset may be predictive of pancreatic cancer-related DM.
Liver cancer represented the second cancer type more associated with DM in this study. Review of published epidemiologic studies indicates that T2DM is associated 2.5-fold increase in the risk for hepatocellular carcinoma. Our estimates ranged from 1.37 to 1.74, with a stronger effect in males. Alcohol consumption or history of hepatitis are among the risk factors for liver cancer and were not available in this study. The lack of adjustment with these variables may have explained the difference in estimates. The stronger association in males may be associated with gender differences in DM-related factors that increase susceptibility to liver cancer, i.e., steatosis, nonalcoholic fatty liver disease, and cirrhosis  and it deserves further investigation.
Contrary to the association between recent DM diagnosis and pancreatic cancer, two studies found that a longer history of DM of greater than 10 years significantly strengthened the association between DM and hepatocellular carcinoma.,
A study from Taiwan of about 1 million Chinese found T2DM increasing the risk of kidney cancer by 1.7 times after adjusting for comorbidities, medication use, tobacco abuse, and alcohol-related diagnosis. The risk also increased with longer duration of DM. Our lower estimate in the whole sample (OR = 1.37) might be explained by the lack of adjustment of other risk factors associated with kidney cancer including, among others, smoking. However, studies from Italy  and two independent meta-analyses derived from nine cohort studies  and 24 studies  found a pooled relative risk of 1.4 which is similar to our estimate. Diabetes may affect the risk of kidney cancer by increasing insulin resistance and blood levels of insulin and IGF-I.
A recent systematic review found patients with T2DM have an increased risk of gallbladder cancer of 1.56 (summary relative risk) after controlling by sex, smoking, BMI, and a history of gallstones. The lack of adjustment by all of these variables may explain in part our lower OR estimate (1.26). In the recent systematic review, diabetic women and men had a similarly increased risk of gallbladder cancer. In the present study, there was a higher effect in males than females (OR = 1.35 vs. 1.22, respectively); this difference might be related to hormonal effects among Mexicans which deserves further investigation. The difference in risk may be explained in part by gender difference of other risk factors for gallbladder cancer including gallstone diseases, gallbladder polyps, obesity, and tobacco smoking., Hyperinsulinemia and up-regulated levels of IGF-1 may stimulate growth through cellular proliferation and the inhibition of apoptosis within the cholangiocytes, a plausible biological mechanism of DM increasing the risk of gallbladder cancer.,
Endometrial cancer was the type less associated with DM in this study with an OR = 1.15 (1.06–1.27). A systematic review found a summary relative risk of 1.89 (95%CI, 1.46–2.45; P < 0.001) for the incidence of endometrial cancer among women with versus without DM. Another cohort study found a weaker association (hazard ratio [HR] =1.44, 95% CI: 1.13–1.85) but without adjusting by body mass index (BMI), however, after BMI adjustment, the associations became nonsignificant. Elevated risk was noted when considering combining DM diagnosed at baseline and during follow-up as time-dependent exposure (HR = 1.31, 95% CI: 1.08–1.59) even after adjusting for BMI. The lack of adjustment with other variables having an effect on the association between DM and endometrial cancer including BMI, may have caused our underestimate; therefore, our findings should be interpreted with caution. High insulin levels and growth factors may also increase the risk of endometrial cancer in women with obesity.,
Diabetes may influence the neoplastic process by several mechanisms, including hyperinsulinemia (either endogenous due to insulin resistance or exogenous due to administered insulin or insulin secretogogues), hyperglycemia, or chronic inflammation.
The inclusion of all deaths, the large sample size and the ability to analyze other cancers for comparison in this population-based study represents the main strengths. The major limitation is that DM identification in death certificates is based on hospital discharge records or interview with relatives/friends which may have led to under-reporting in some degree for some type-specific cancers. As previously mentioned, history of smoking, obesity/high BMI, uncontrolled DM, lifestyle, diet among others, are all known risk factors of type-specific cancers and unfortunately, they were not available in our study. Therefore, adjusting for these variables would likely result in effect modification on estimates and deserve further investigation.
| > Conclusions|| |
Cancers types that were associated with DM among Mexicans include pancreatic, liver, kidney, gallbladder, and endometrial cancers. The association was stronger in males than females (except for kidney cancer). Type 2 or unknown-type DM were associated with the same cancer-types with some variation on estimates. The association between T1DM and pancreatic cancer only may be explained by small sample size and further research is needed. Diabetes education and prevention including an adequate diet, regular exercise, weight reduction if obese/overweight, cessation of smoking, good glucose control and medication compliance among others should be encouraged among Mexican diabetics. This study also demonstrates the use of death certificate databases for monitoring disease outcomes.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Giovannucci E, Harlan DM, Archer MC, Bergenstal RM, Gapstur SM, Habel LA, et al
. Diabetes and cancer: A consensus report. Diabetes Care 2010;33:1674-85.
Joslin EP, Lombard HL, Burrows RE, Manning MD. Diabetes and cancer. N Engl J Med 1959;260:486-8.
Tanaka K, Tsuji I, Tamakoshi A, Matsuo K, Wakai K, Nagata C, et al
. Diabetes mellitus and liver cancer risk: An evaluation based on a systematic review of epidemiologic evidence among the Japanese population. Jpn J Clin Oncol 2014;44:986-99.
Bosetti C, Rosato V, Li D, Silverman D, Petersen GM, Bracci PM, et al
. Diabetes, antidiabetic medications, and pancreatic cancer risk: An analysis from the international pancreatic cancer case-control consortium. Ann Oncol 2014;25:2065-72.
Liao C, Zhang D, Mungo C, Tompkins DA, Zeidan AM. Is diabetes mellitus associated with increased incidence and disease-specific mortality in endometrial cancer? A systematic review and meta-analysis of cohort studies. Gynecol Oncol 2014;135:163-71.
Chen HF, Liu MD, Chen P, Chen LH, Chang YH, Wen PC, et al
. Risks of breast and endometrial cancer in women with diabetes: A population-based cohort study. PLoS One 2013;8:e67420.
Jarvandi S, Davidson NO, Schootman M. Increased risk of colorectal cancer in type 2 diabetes is independent of diet quality. PLoS One 2013;8:e74616.
De Bruijn KM, Arends LR, Hansen BE, Leeflang S, Ruiter R, van Eijck CH. Systematic review and meta-analysis of the association between diabetes mellitus and incidence and mortality in breast and colorectal cancer. Br J Surg 2013;100:1421-9.
Turati F, Polesel J, Di Maso M, Montella M, Libra M, Grimaldi M, et al
. Diabetes mellitus and the risk of bladder cancer: An Italian case–control study. Br J Cancer 2015;113:127-30.
Goossens ME, Zeegers MP, Bazelier MT, De Bruin ML, Buntinx F, de Vries F. Risk of bladder cancer in patients with diabetes: A retrospective cohort study. BMJ Open 2015;5:e007470.
Tsilidis KK, Allen NE, Appleby PN, Rohrmann S, Nöthlings U, Arriola L, et al
. Diabetes mellitus and risk of prostate cancer in the European prospective investigation into cancer and nutrition. Int J Cancer 2015;136:372-81.
Dankner R, Boffetta P, Balicer RD, Boker LK, Sadeh M, Berlin A, et al
. Time-dependent risk of cancer after a diabetes diagnosis in a cohort of 2.3 million adults. Am J Epidemiol 2016;183:1098-106.
Ohkuma T, Peters SA, Woodward M. Sex differences in the association between diabetes and cancer: A systematic review and meta-analysis of 121 cohorts including 20 million individuals and one million events. Diabetologia 2018;61:2140-54.
Setiawan VW, Stram DO, Porcel J, Chari ST, Maskarinec G, Le Marchand L, et al
. Pancreatic cancer following incident diabetes in African Americans and Latinos: The multiethnic cohort. J Natl Cancer Inst 2019;111:27-33.
Guide for the Filling of the Death Certificate and the Fetal Death Certificate. Mexican Center for disease classification. General Directorate of Health Information. Ministry of Health 3rd
Edition; 2007. Available from: http://www.salud.gob.mx/unidades/cdi/documentos/DOCSAL7761.pdf
. [Last retrieved on 2019 Oct 15].
Rojas-Martínez R, Basto-Abreu A, Aguilar-Salinas CA, Zárate-Rojas E, Villalpando S, Barrientos-Gutiérrez T. Prevalence of previously diagnosed diabetes mellitus in Mexico. Salud Publica Mex 2018;60:224-32.
Menke A, Casagrande S, Geiss L, Cowie CC. Prevalence of and trends in diabetes among adults in the United States, 1988-2012. JAMA 2015;314:1021-9.
Harding JL, Shaw JE, Peeters A, Cartensen B, Magliano DJ. Cancer risk among people with type 1 and type 2 diabetes: Disentangling true associations, detection bias, and reverse causation. Diabetes Care 2015;38:264-70.
Kasper JS, Giovannucci E. A meta-analysis of diabetes mellitus and the risk of prostate cancer. Cancer Epidemiol Biomarkers Prev 2006;15:2056-62.
Aggarwal G, Kamada P, Chari ST. Prevalence of diabetes mellitus in pancreatic cancer compared to common cancers. Pancreas 2013;42:198-201.
Huxley R, Ansary-Moghaddam A, Berrington de González A, Barzi F, Woodward M. Type-II diabetes and pancreatic cancer: A meta-analysis of 36 studies. Br J Cancer 2005;92:2076-83.
Ben Q, Xu M, Ning X, Liu J, Hong S, Huang W, et al
. Diabetes mellitus and risk of pancreatic cancer: A meta-analysis of cohort studies. Eur J Cancer 2011;47:1928-37.
Cohen JD, Li L, Wang Y, Thoburn C, Afsari B, Danilova L, et al
. Detection and localization of surgically resectable cancers with a multi-analyte blood test. Science 2018;359:926-30.
Andersen DK, Korc M, Petersen GM, Eibl G, Li D, Rickels MR, et al
. Diabetes, pancreatogenic diabetes, and pancreatic cancer. Diabetes 2017;66:1103-10.
Eibl G, Cruz-Monserrate Z, Korc M, Petrov MS, Goodarzi MO, Fisher WE, et al
. Diabetes mellitus and obesity as risk factors for pancreatic cancer. J Acad Nutr Diet 2018;118:555-67.
Mueller AM, Meier CR, Jick SS, Schneider C. Weight change and blood glucose concentration as markers for pancreatic cancer in subjects with new-onset diabetes mellitus: A matched case-control study. Pancreatology 2019;19:578-86.
El-Serag HB, Hampel H, Javadi F. The association between diabetes and hepatocellular carcinoma: A systematic review of epidemiologic evidence. Clin Gastroenterol Hepatol 2006;4:369-80.
Yu MC, Tong MJ, Govindarajan S, Henderson BE. Nonviral risk factors for hepatocellular carcinoma in a low-risk population, the non-Asians of Los Angeles County, California. J Natl Cancer Inst 1991;83:1820-6.
El-Serag HB, Tran T, Everhart JE. Diabetes increases the risk of chronic liver disease and hepatocellular carcinoma. Gastroenterology 2004;126:460-8.
Tseng CH. Type 2 diabetes mellitus and kidney cancer risk: A retrospective cohort analysis of the national health insurance. PLoS One 2015;10:e0142480.
Zucchetto A, Dal Maso L, Tavani A, Montella M, Ramazzotti V, Talamini R, et al
. History of treated hypertension and diabetes mellitus and risk of renal cell cancer. Ann Oncol 2007;18:596-600.
Larsson SC, Wolk A. Diabetes mellitus and incidence of kidney cancer: A meta-analysis of cohort studies. Diabetologia 2011;54:1013-8.
Bao C, Yang X, Xu W, Luo H, Xu Z, Su C, et al
. Diabetes mellitus and incidence and mortality of kidney cancer: A meta-analysis. J Diabetes Complications 2013;27:357-64.
Gu J, Yan S, Wang B, Shen F, Cao H, Fan J, et al
. Type 2 diabetes mellitus and risk of gallbladder cancer: A systematic review and meta-analysis of observational studies. Diabetes Metab Res Rev 2016;32:63-72.
Hundal R, Shaffer EA. Gallbladder cancer: Epidemiology and outcome. Clin Epidemiol 2014;6:99-109.
Park M, Song DY, Je Y, Lee JE. Body mass index and biliary tract disease: A systematic review and meta-analysis of prospective studies. Prev Med 2014;65:13-22.
Cai HH, Sun YM, Bai JF, Shi Y, Zhao HL, Miao Y. Relationship between the GH-IGFs axis and the proliferation of bile duct cancer cell line QBC939 in vitro
. Hepatobiliary Pancreat Dis Int 2008;7:76-81.
Alvaro D, Barbaro B, Franchitto A, Onori P, Glaser SS, Alpini G, et al
. Estrogens and insulin-like growth factor 1 modulate neoplastic cell growth in human cholangiocarcinoma. Am J Pathol 2006;169:877-88.
Luo J, Beresford S, Chen C, Chlebowski R, Garcia L, Kuller L, et al
. Association between diabetes, diabetes treatment and risk of developing endometrial cancer. Br J Cancer 2014;111:1432-9.
Ordener C, Cypriani B, Vuillermoz C, Adessi GL. Epidermal growth factor and insulin induce the proliferation of guinea pig endometrial stromal cells in serum-free culture, whereas estradiol and progesterone do not. Biol Reprod 1993;49:1032-44.
[Table 1], [Table 2], [Table 3], [Table 4]