Epidemiology of cancer and diabetes

There is a growing body of evidence to support a relationship between diabetes and cancer. Epidemiologic studies have observed a higher incidence of various types of cancer among individuals with diabetes, particularly type 2 diabetes, as compared with people without diabetes. People with diabetes also have poorer survival than non-diabetics after a diagnosis of cancer. The relationship between diabetes and cancer incidence and/or mortality is however complex, and potential sources of bias need to be considered with care when interpreting the data. Furthermore, the association with diabetes varies in important respects between cancer types, which need to be considered on an individual basis rather than collectively when assessing the relationship between diabetes and cancer.

Introduction

Studies from the early part of the last century suggested that people with diabetes were less likely to die of cancer than non-diabetic individuals, an observation that can be explained by ‘competing risks’ (i.e., causes of death) - in this case the high mortality from accelerated arterial disease in the diabetic population. As cardiovascular risk management in diabetes becomes more effective, the morbidity and mortality associated with cancer will inevitably increase in the diabetic population.

Many recent studies have in contrast confirmed a strong positive association between diabetes and many types of tumour[1][2]. Furthermore, people with diabetes have a shorter survival after a diagnosis of cancer than non-diabetic individuals, although this might once again be due in part to competing risks (higher diabetes-associated mortality).[3]

Diabetes type and cancer

Since 90-95% of people with diabetes in the general population have type 2 diabetes[4], and belong to an age group more commonly affected by cancer, much more is known about the relationship between cancer and type 2 diabetes than about type 1. The comparison between the conditions is of particular interest in that a similar pattern of cancer risk in the type 2 and type 1 populations would support a causal role for hyperglycemia, whereas a different pattern would point to non-glycemic factors.

Risk for Individual Cancers in type 2 diabetes

Summary data from meta-analyses of cancer risks in type 2 diabetes are shown in the figure.

The strongest association between diabetes and incident cancer is for Pancreatic cancer[5] and Liver cancer and diabetes[6] ; however, ‘reverse causality’ (where the cancer itself leads to diabetes) may in part overestimate the magnitude of this relationship. The risk for Endometrial cancer is approximately twice as high in the diabetic population.[7]

Individuals with diabetes also have a moderately elevated risk (20-40%) of incident Breast Cancer, Colorectal cancer, Bladder cancer and diabetes, thyroid and kidney. There is no consistent evidence for an increased risk of lung or ovarian cancers in the diabetic population.[8]

A notable exception is Prostate cancer and diabetes, where men with diabetes have a 15% lower risk of prostate cancer than non-diabetic men, possibly due to lower endogenous testosterone levels in men with diabetes.[9][10]

Cancer Mortality and Diabetes

Individuals with diabetes also have a higher risk of dying from various cancers, including prostate cancer, compared with individuals who do not have diabetes. In general, findings for increased cancer mortality reflect cancer incidence. Among cancer patients, those with diabetes have a higher risk of mortality if they have colorectal, liver, pancreatic and bladder cancers.

There is less consistent evidence around the risk of mortality among diabetic individuals with breast and endometrial cancer.[3] Competing risks may potentially confound comparisons between survival in the diabetic and non-diabetic populations, since people with diabetes have a higher cardiovascular mortality than those without.

Biologic plausibility

Associations at the epidemiologic level correlate with results from biologic experiments. The leading hypothesis for the association between diabetes and cancer is that hyperinsulinemia (i.e., high insulin levels), a common feature of diabetes, also promotes the growth of cancer cells, which typically express the insulin receptor.[1]. See Hyperinsulinemia and cancer. Although hyperinsulinemia is the leading biologic hypothesis, hyperglycemia (see Hyperglycemia and cancer) and chronic inflammation may also play a role in promoting cancer growth.[1][2]

Shared risk factors for cancer

The relationship between diabetes and cancer is unique to each cancer site and differs by sex and the type of diabetes (i.e., type 1 vs. type 2). Both diabetes and cancer are chronic diseases associated with an aging population. Aside from age and sex, modifiable risk factors such as obesity, smoking, alcohol use, diet, physical inactivity and other lifestyle choices are factors that increase the risk for both diseases. This combination of common non-modifiable (i.e., age, sex, genetics, etc) and modifiable (i.e., smoking, diet, etc) risk factors increases the likelihood for an individual with diabetes to also develop cancer and adds to the complexity of describing excess cancer risk in the diabetic population.

Methodological considerations

Given the indolent and slowly-progressing nature of both cancer and diabetes, ascertaining cancer rates in the diabetic population is challenging. Ideally, studies should have two cancer-free cohorts of individuals: one with newly diagnosed diabetes and another without diabetes, both of which should be followed for a long enough duration for cancer to develop (i.e., several years).[2]

Some epidemiologic methods used to describe cancer rates in individuals with diabetes may be prone to ‘detection bias’ (where a visit to a doctor at the time of diabetes diagnosis makes it more likely for the doctor to find an undiagnosed cancer), which may overestimate the risk of cancer in people with newly diagnosed diabetes. [11][12] Other studies may be prone to ‘lead time bias’ (where individuals must survive a certain amount of time after a diagnosis of diabetes to be included in a study), which may underestimate the rates of cancer in the diabetes population.[13] Other methodologic considerations surrounding epidemiologic research of diabetes and cancer were highlighted in two recent consensus papers published by members of the international Diabetes and Cancer Research Consortium.[2][3]

References

  1. ^ Giovannucci E et al. Diabetes and a consensus report. Diabetes Care. 2010; 1674–85.

  2. ^ Johnson JA et al. Diabetes and cancer (1): evaluating the temporal relationship between type 2 diabetes and cancer incidence. Diabetologia. 2012;55(6):1607–18.

  3. ^ Renehan AG et al. Diabetes and cancer (2): evaluating the impact of diabetes on mortality in patients with cancer. Diabetologia. 2012 Jun;55(6):1619–32.

  4. ^ World Health Organization. Diabetes: Fact Sheet No.312 [Internet]. who.int. 2013 [cited 2013 Jun 13].

  5. ^ Huxley R et al. Type-II diabetes and pancreatic cancer: a meta-analysis of 36 studies. Br J Cancer. 2005 Jun 6;92(11):2076–83.

  6. ^ El-Serag HB et al. The association between diabetes and hepatocellular carcinoma: a systematic review of epidemiologic evidence. Clin Gastroenterol Hepatol. 2006 Mar 1;4(3):369–80.

  7. ^ Friberg E et al. Diabetes mellitus and risk of endometrial cancer: a meta-analysis. Diabetologia. 2007 Jul 1;50(7):1365–74.

  8. ^ Hall GC et al. Diabetes and the risk of lung cancer. Diabetes Care. 2005 Mar;28(3):590–4.

  9. ^ Weiderpass E et al. Reduced risk of prostate cancer among patients with diabetes mellitus. Int J Cancer. 2002;102(3):258–61.

  10. ^ Kasper JS, Giovannucci E. A meta-analysis of diabetes mellitus and the risk of prostate cancer. Cancer Epidemiol Biomarkers Prev. 2006 Nov 1;15(11):2056–62.

  11. ^ Colmers IN et al.Evidence of detection bias and overestimated risk of bladder cancer in type 2 diabetes. Diabetes Care (in press). 2013.

  12. ^ Johnson JA et al. Time-varying incidence of cancer after the onset of type 2 evidence of potential detection bias. Diabetologia. 2011 Sep;54(9):2263–71.

  13. ^ Suissa S. Immortal time bias in pharmaco-epidemiology. Am J Epidemiol. 2008 Feb 15;167(4):492–9.

Comments

Nobody has commented on this article

Commenting is only available for registered Diapedia users. Please log in or register first.