© 2007 European Society for Medical Oncology
epidemiology |
Association between glycosylated hemoglobin and cancer risk: a New Zealand linkage study
1 Centre for Public Health Research, Massey University, Wellington
2 Public Health Intelligence, Ministry of Health, Wellington
3 Research Centre for M
ori Health and Development, Massey University, Wellington
4 Hepatitis Foundation of New Zealand, Whakatane, New Zealand
* Correspondence to: N. Travier, Centre for Public Health Research, Massey University, Private Bag 756, Wellington, New Zealand. Tel: +64-4-801-2794 ext. 6088; Fax: +64-4-380-0600; E-mail: n.travier{at}massey.ac.nz
 |
Abstract |
|---|
 Top Abstract introductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Background: The purpose of this study was to examine the relationship between glycosylated hemoglobin (HbA1c) level and subsequent cancer risk.
Material and methods: HbA1c measurements were made on blood samples of participants in a hepatitis B (HB) screening program (1999–2001). Cancer incidence was determined by linkage to cancer registrations and hospitalization records to the end of 2004. Participants previously diagnosed with diabetes or cancer were excluded. Hazard ratios (HR) and 95% confidence intervals (CIs) were estimated using Cox regression.
Results: Among the 46 575 participants (70% Mori, 12% Pacific, 5% Asian and 12% Other), 634 cancer cases were observed. For all cancers combined, a significant increased risk was found in persons with moderately elevated HbA1c levels (6%–6.9%) (HR 1.40, 95% CI: 1.11–1.76), with a smaller increased risk in persons with highly elevated levels (
7%) (HR 1.09, 95% CI: 0.80–1.48) as compared with persons having low HbA1c levels (<6%). The HRs for respiratory cancers were 2.27 (95% CI: 1.34–3.86) for the moderate HbA1c category and 1.58 (95% CI: 0.77–3.26) for the upper HbA1c category. For endometrial cancers, the HRs were 4.05 (95% CI: 1.10–14.88) and 5.07 (95% CI: 1.20–21.31), respectively. For other cancer sites, no significantly increased risks were found.
Conclusions: These findings are consistent with other evidence that abnormal glucose metabolism may be associated with an increased risk of some cancers.
Key words: cancer incidence, diabetes, glycosylated hemoglobin, New Zealand
|
introduction |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
A number of epidemiological studies have investigated the association between type 2 diabetes and cancer and increased risks have been found, although not entirely consistently, for a range of cancers. The most consistent increased risks have been found for cancers of the endometrium [1, 2], colon/rectum [3, 4], pancreas [5, 6], liver [2, 7, 8] and breast [1, 9]. Although the findings for prostate cancer have been relatively inconsistent, a recent meta-analysis reported that men with diabetes have a significant decreased risk of developing prostate cancer [10].
The mechanisms by which diabetes could cause cancer are uncertain and the relationship between diabetes and site-specific malignancies is difficult to analyze for several reasons. First, cancers from certain sites such as colon, breast (in postmenopausal women), endometrium, esophagus and kidney share some important risk factors with diabetes, such as obesity and a sedentary lifestyle [11]. Secondly, for some cancers, such as pancreatic cancer, the critical question is whether diabetes is a true causal factor or merely a consequence of cancer during a prediagnostic stage, particularly since the mechanisms by which chronic diabetes may cause pancreatic cancer are uncertain [6]. Thirdly, as most types of cancer are rare, they are usually studied through case–control studies that may be prone to information bias if information on diabetic status is collected retrospectively. Nevertheless, it is plausible that hyperglycemia and hyperinsulinemia may contribute to the association between type 2 diabetes and cancer; in particular, glucose through its action on production of insulin and insulin-like growth factor (IGF)-I may enhance tumor development by stimulating cell proliferation and by inhibiting apoptosis [12, 13].
The study reported here examined the association between glycosylated hemoglobin (HbA1c) and subsequent cancer risk among subjects initially free of diabetes. HbA1c reflects overall glucose levels for a period of 2–3 months, the life span of red blood cells, and is used to monitor diabetes treatment [14, 15]. As HbA1c levels are not affected by recent meals and does not require a fasting blood sample, it is regarded as an alternative to fasting plasma glucose and 2-h postload plasma glucose for preliminary screening for undiagnosed diabetes [14, 16].
|
materials and methods |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
From 1999 to 2001, free HB screening was offered in the lower half of the North Island of New Zealand. Although this program was open to everyone, it was targeted at non-European (predominantly M
ori, Pacific and Asian) adults. The program was publicized through community groups and media, and participants were recruited through caravans located in public places such as shopping malls. All participants without diabetes were offered an HbA1c measurement as an additional test at the same time as the HB test. We have examined the subsequent cancer risk to the end of 2004 in these participants by linkage to routinely collected health data using the National Health Index (NHI), a unique identifier used in the New Zealand Public Health system. Cancer diagnoses were obtained from the Cancer Registry as well as from public hospital discharges. Cancer registration has been compulsory in New Zealand since 1994, and the Cancer Registry has been virtually 100% complete since that date [17], although there are delays in the registration process for some tumor sites. The Cancer Registry was used to identify all incident cancer cases during 1999–2001, the year to which it is currently complete, as well as ‘priority cancers’ (breast, cervix, prostate, colorectal, melanoma, respiratory, hematological and lymphoid) diagnosed in 2002–2003. For 2002, 2003 and 2004, public hospital discharges were also used to identify people with a first admission for cancer that had not yet been registered with the Cancer Registry.
We excluded the participants who had had a cancer registered or a diabetes diagnosis before their HbA1c test. Diagnoses of diabetes were based on hospitalizations or general practice prescriptions for insulin, an oral hypoglycemic or a home glucose monitoring kit. Ethnicity was self-identified, as is standard in New Zealand health research [18], using the categories of Mori, Pacific, Asian and ‘other’; the other category is predominantly of European origin. Participants who reported more than one ethnicity were classified using the standard system of prioritization in which participants who identify as Mori are classified as Mori (even if they also recorded other ethnicities); of the remainder, those who identify as Pacific are classified as Pacific ethnicity, and so on. When ethnicity was not reported in the HB program records, the participant was assigned their ethnicity as recorded in the NHI database. Smoking data were self-reported at the time of the blood sampling for HbA1c measure using the categories of never smoker, ex-smoker and current smoker.
HbA1c was categorized into three groups, namely <6%, 6.0%–6.9% and 7% or above. Cox proportional hazards regression was used to model the risk of subsequent cancer following an individual's HbA1c measurement in each of these three categories. Follow-up extended from the date of the HbA1c measure to the first of the following: date of cancer registration/hospitalization, date of death or 31st December 2004. The timescale underlying the Cox model was age. The proportional hazards assumption was assessed graphically for each model. All models were adjusted for age (continuous variable), gender (except for sex-specific cancers) and ethnicity (Mori, Pacific, Asian and Other). For the site-specific analyses, when there were too few cancer cases, some of the ethnic groups were combined to allow multivariate analyses.
To investigate the possibility of confounding by smoking, we ran models adjusted for self-reported smoking (current, ex-smoker, nonsmoker) at baseline, as well as analyses restricted to nonsmokers. We presented the unadjusted (and unrestricted) findings as the main results, since 13% of participants had missing values for this variable, and we did not want to exclude them. All analyses were conducted using Stata, version 8.0 [19].
|
results |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
HbA1c measurements were obtained for 48 673 subjects. Of these, 764 (1.6%) did not appear to be correctly matched (based on dates of birth of >366 days difference between the hepatitis and NHI databases), three participants were recorded as having died before having the HbA1c test, 791 (1.6%) were under age 18 or had had a diabetes diagnosis at the time of the HbA1c test and 540 (1.1%) people had had a diagnosis of cancer before the HbA1c test. All analyses are therefore based on 46 575 subjects. As shown in Table 1, there was a slight preponderance of females compared with males in the total population but an excess of males in the high HbA1c category. Of the full sample, 70% were M
ori, 12% were Pacific, 5% were Asian and 12% had other ethnicities. Pacific and Asian people were overrepresented in the high HbA1c category. The mean age was 38 years. The median HbA1c level was 5.2%; 91.1% of the subjects had normal levels of HbA1c (<6%), while 5.6% had moderately elevated levels (6.0%–6.9%) and 3.4% had highly elevated levels (7% or above). As expected, age was inversely associated with HbA1c. In these crude analyses, current smokers were less likely to have high HbA1c levels than never smokers.
|
The median follow-up period was 4.4 years (range: 2 days to 5.3 years). During the follow-up period, 634 new cases of cancer were diagnosed, including 254 men and 380 women. The most frequently occurring cancers were female breast cancer (n = 132), oral and digestive system cancers (n = 112), respiratory cancers (n = 78), male genital cancers (n = 63) and hematopoietic cancers (n = 59).
Cancer incidence was greater in persons with HbA1c levels of 6% or above than in persons with normal levels of HbA1c (Table 2). For all cancers combined, we observed a significantly higher rate of cancer incidence in persons with moderately elevated HbA1c levels [hazard ratio (HR) 1.40, 95% confidence interval (CI): 1.11–1.76], while a nonsignificant 9% increase was observed in persons with highly elevated HbA1c levels (HR 1.09, 95% CI: 0.80–1.48) as compared with persons having normal HbA1c levels.
|
In particular, increased risks were observed for respiratory cancers, for which we found a significantly increased risk of 2.27 (95% CI: 1.34–3.86) for the participants having moderately elevated HbA1c levels and a nonsignificant increased risk of 1.58 (95% CI: 0.77–3.26) for the participants having highly elevated HbA1c levels as compared with persons having normal HbA1c levels. Increased risks were also observed for female genital cancers and more particularly for endometrial cancer. We did not observe any increased risk for breast cancer in general, and the pattern was similar when women were considered as pre- or postmenopausal (defined as under or over age 50 at the time of the blood test). For the other cancer sites, and in particular for colorectal cancers, no significant increased risks were found.
When the analyses were restricted to those with smoking data, the association between HbA1c levels and all cancers appeared slightly attenuated (HR 1.29, 95% CI: 0.99–1.67 and HR 0.99, 95% CI: 0.70–1.41 for moderately and highly elevated levels, respectively, compared with normal HbA1c levels). When these analyses were then adjusted for smoking, the association between HbA1c levels and all cancers did not change; for the participants having moderately elevated HbA1c levels, the HR changed from 1.29 to 1.28 while for the participants having highly elevated HbA1c levels the HR did not change at all. Similarly, adjusting for tobacco smoking did not change the association between HbA1c levels and respiratory cancers; the risk experienced by the persons having moderately elevated HbA1c levels changed from 1.83 (95% CI: 1.01–3.34) to 1.78 (95% CI: 0.97–3.24), while the risk experienced by the persons having highly elevated HbA1c levels changed from 1.36 (95% CI: 0.61–3.06) to 1.31 (95% CI: 0.58–2.93) as compared with persons having normal HbA1c levels.
When the analyses were restricted to never smokers, the association between HbA1c levels and all cancers combined was partially attenuated (HR 1.35, 95% CI: 0.91–1.99 and HR 1.18, 95% CI: 0.72–0.91 for moderately and highly elevated levels, respectively, compared with normal HbA1c levels). For respiratory cancers, a strong effect of HbA1c levels persisted in the moderate but not the raised HbA1c category (HR 5.73, 95% CI: 1.77–18.61 and HR 1.42, 95% CI: 0.17–12.09).
|
discussion |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
This study has found evidence to indicate that there may be an increased cancer risk associated with HbA1c levels
6%, particularly in those with moderately elevated HbA1c levels (6%–6.9%), i.e. at HbA1c levels below the accepted threshold indicative of diabetes. The most consistent increases were found for all cancers, and were noted in site-specific analyses for respiratory cancers and female genital cancers. While subjects with highly elevated HbA1c levels also experienced increased cancer risks, increased risks were generally highest among the subjects having moderately elevated HbA1c. The findings presented here should be interpreted cautiously, since this study has several limitations. The first and probably main limitation of this study is the lack of anthropometric data. Indeed, as overweight is a known risk factor for impaired glucose tolerance and diabetes as well as for several cancers, the increased risks observed for these cancers may be due to excess body weight or obesity rather than elevated glucose levels per se. Excess body weight is, in particular, associated with cancers of the colon, breast (in postmenopausal women), endometrium, kidney, esophagus and gastric cardia [11]. Therefore, we cannot exclude that the increase in endometrial cancer experienced by both subjects with moderately and highly elevated HbA1c levels might be due to excess weight. The short follow-up time is a second limitation of this study, as with such a short time between blood donation and outcome, we could not restrict the analyses to the cases diagnosed at least 2 years after the HbA1c test to eliminate subjects with undetected disease at the time of the blood test. Therefore, we cannot exclude the possibility that undiagnosed cancer at the time of the HbA1c test might have led to elevated glucose levels. Finally, the small numbers of some specific cancers limited the power for analyses by cancer types, so the HRs obtained for specific types of cancer should be interpreted with caution.
On the other hand, the use of HbA1c levels to assess blood glucose concentrations over time is one of the major strengths of this study. As these levels are not subjected to day-to-day variations, as fasting and 2-h oral glucose tolerance tests can be, they are likely to more accurately indicate blood glucose levels over time.
Bearing in mind these limitations, the findings of this study are of considerable interest. Several studies have previously reported increased cancer risks in patients with elevated glucose levels and/or diabetes, indicating that these increases may be related to compensatory hyperinsulinemia [20, 21]. Insulin may favor cancer development acting by itself as a growth factor, via insulin receptors, or by increasing IGF-I activity. Insulin enhances the stimulatory effects of growth hormone on IGF-I synthesis, and can also increase IGF-I bioactivity by down-regulating the synthesis of IGF-binding proteins-1 and -2 [22]. Several prospective studies support the role of IGF-I on the development of prostate [23, 24], colorectal [25, 26] and premenopausal breast [27, 28] cancers in particular. Furthermore, as both insulin and IGF-I are also involved in the synthesis and circulating levels of sex steroids and sex hormone-binding globulin, insulin and IGF-I may also specifically increase the risk of hormone-responsive cancers such as breast and endometrial cancers [22].
In a recent publication, Wolf et al. [9] reported that a number of cohort and case–control studies indicate that type 2 diabetes may be associated with a 10%–20% excess relative risk for breast cancer. This finding is not supported by the present study which found a slight increased risk among moderately exposed subjects and no increased risk among highly exposed subjects. Nevertheless, in the analyses focused on women who were 50 years old or younger at the time of their blood test, moderately elevated glucose levels were associated with an increased risk for breast cancer. This finding is in agreement with a recent meta-analysis that showed a positive association between circulating IGF-I and breast cancer limited to premenopausal cancer [29].
The present study also indicated that women with elevated blood glucose levels have an increased risk for genital cancers and in particular endometrial cancer. Given the small number of endometrial cancer cases, this association should be interpreted with caution. Nevertheless, this finding is in agreement with previous research indicating that women with diabetes experience a higher incidence of endometrial cancer [1, 2]. These increases in endometrial and premenopausal breast cancers associated with elevated glucose levels tend therefore to support the hypothesis that glucose may increase the risk for cancer through its action on the production of insulin and IGF-I.
The significantly increased risk for respiratory cancers observed among the subjects having elevated glucose levels was unexpected, since previous studies reporting results on diabetes or impaired glucose tolerance and lung cancer have indicated only nonsignificant weakly positive associations [30–32] or inverse associations [33, 34]. Our findings are unlikely to be explained by tobacco smoking, as the analyses based on nonsmokers showed similar results. The increased risk observed may be due to elevated levels of IGF-I, as Yu et al. [35] previously reported a strong positive association between circulating IGF-I and lung cancer. However, more recent studies have not confirmed this finding [29, 36].
The present study did not indicate any increased risk for colorectal cancer among the subjects who had elevated levels of HbA1c. These results therefore do not support the positive associations previously reported between diabetes or elevated glucose levels and colorectal cancer [3, 20, 37]. Most previous studies were based on Caucasian populations, while this study, involving 70% Mori and 12% Pacific, mainly involved non-Caucasian people. As Mori and Pacific people have low rates of colorectal cancer [38], ethnic differences in the study populations might explain the observed differences.
The present study showed no association between blood glucose levels and prostate cancer risk. This is in agreement with the absence of association between diabetes and prostate cancer reported by Tavani et al. [39] but in contradiction with the inverse association reported in a recent meta-analysis [10]. If an inverse association does exist between diabetes and prostate cancer, this may be explained by decreased levels of testosterone, known to control cell division in the prostate gland, in diabetic men [10].
Studies of predominantly type 2 diabetes have generally shown significant raised risks for liver cancer [2, 7, 8, 40, 41] but this was not observed in the present study. However, the causal mechanism for an elevated risk for liver cancer in diabetic patients is unclear. While insulin resistance and hyperinsulinemia might explain the elevated risks observed in some studies, hepatitis, cirrhosis and alcohol consumption remains possible confounders [40, 41].
In light of these findings and the hypothesized mechanisms, it remains unclear why moderately elevated HbA1c levels should be more strongly related to cancer incidence than highly elevated levels. In fact, this finding is consistent with the results of Saydah et al. [30] indicating that impaired glucose tolerance was more strongly related to cancer death than diabetes itself. Competing causes of death as well as a protective effect of diabetes giving longer survival rates for individuals with malignant tumors were the possible explanations given for this finding. As the present study involved incidence, rather than mortality, neither of these hypotheses could explain our findings.
Increased cancer risks were found among subjects having relatively low levels of blood glucose (HbA1c of 6% or more). These increased risks were therefore present throughout the whole range of HbA1c levels, and in particular below the threshold commonly accepted for diagnosis of diabetes. Therefore, people should not be falsely reassured that because they have not been diagnosed with diabetes they do not have a significant health risk. Furthermore, as elevated HbA1c levels were quite common in the study population (9% of the study population had HbA1c levels of 6% or more), slightly to moderately elevated HbA1c levels should result in close monitoring.
The main implication of this study might be that elevated HbA1c levels may be associated with an increased risk for cancer, in particular respiratory and endometrial cancers. These results are also consistent with previous findings that identified elevated glucose levels, leading to elevated insulin levels and related physiologic determinants, as potential mechanisms and risk factors for cancer. These excess risks are unlikely to be due to confounding by smoking, since the results changed little when the analyses were adjusted for smoking or restricted to never smokers. Nevertheless, as the results presented here are not adjusted for anthropometric measures (which were not available), we cannot exclude the possibility that the increased risks observed for cancers known to be associated with excess body weight (e.g. endometrial cancer) might be due in part to excess weight rather than to elevated HbA1c levels.
|
Acknowledgements |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
The Centre for Public Health Research is supported by a Program Grant from the Health Research Council of New Zealand. We thank the staff at the New Zealand Health Information Service which is responsible for the datasets (on hospital admissions and mortality) that were used in these analyses.
Received for publication July 12, 2006. Revision received January 17, 2007. Revision received March 16, 2007. Accepted for publication March 23, 2007.
|
References |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
1. Weiderpass E, Gridley G, Persson I, et al. Risk of endometrial and breast cancer in patients with diabetes mellitus. Int J Cancer (1997) 71:360–363.[CrossRef][Web of Science][Medline]
2. Wideroff L, Gridley G, Mellemkjaer L, et al. Cancer incidence in a population-based cohort of patients hospitalized with diabetes mellitus in Denmark. J Natl Cancer Inst (1997) 89:1360–1365.
3. Larsson SC, Giovannucci E, Wolk A. Diabetes and colorectal cancer incidence in the cohort of Swedish men. Diabetes Care (2005) 28:1805–1807.
4. Hu FB, Manson JE, Liu S, et al. Prospective study of adult onset diabetes mellitus (type 2) and risk of colorectal cancer in women. J Natl Cancer Inst (1999) 91:542–547.
5. Everhart J, Wright D. Diabetes mellitus as a risk factor for pancreatic cancer. A meta-analysis. JAMA (1995) 273:1605–1609.
6. Silverman DT, Schiffman M, Everhart J, et al. Diabetes mellitus, other medical conditions and familial history of cancer as risk factors for pancreatic cancer. Br J Cancer (1999) 80:1830–1837.[CrossRef][Web of Science][Medline]
7. Davila JA, Morgan RO, Shaib Y, et al. Diabetes increases the risk of hepatocellular carcinoma in the United States: a population based case control study. Gut (2005) 54:533–539.
8. La Vecchia C, Negri E, Decarli A, Franceschi S. Diabetes mellitus and the risk of primary liver cancer. Int J Cancer (1997) 73:204–207.[CrossRef][Web of Science][Medline]
9. Wolf I, Sadetzki S, Catane R, et al. Diabetes mellitus and breast cancer. Lancet Oncol (2005) 6:103–111.[CrossRef][Web of Science][Medline]
10. Bonovas S, Filioussi K, Tsantes A. Diabetes mellitus and risk of prostate cancer: a meta-analysis. Diabetologia (2004) 47:1071–1078.[Web of Science][Medline]
11. Bianchini F, Kaaks R, Vainio H. Overweight, obesity, and cancer risk. Lancet Oncol (2002) 3:565–574.[CrossRef][Web of Science][Medline]
12. Kaaks R, Lukanova A. Energy balance and cancer: the role of insulin and insulin-like growth factor-I. Proc Nutr Soc (2001) 60:91–106.[Web of Science][Medline]
13. LeRoith D, Roberts CT Jr. The insulin-like growth factor system and cancer. Cancer Lett (2003) 195:127–137.[Web of Science][Medline]
14. Sacks DB, Bruns DE, Goldstein DE, et al. Guidelines and recommendations for laboratory analysis in the diagnosis and management of diabetes mellitus. Clin Chem (2002) 48:436–472.
15. Steinbrook R. Facing the diabetes epidemic—mandatory reporting of glycosylated hemoglobin values in New York City. N Engl J Med (2006) 354:545–548.
16. Miyazaki M, Kubo M, Kiyohara Y, et al. Comparison of diagnostic methods for diabetes mellitus based on prevalence of retinopathy in a Japanese population: the Hisayama Study. Diabetologia (2004) 47:1411–1415.[Web of Science][Medline]
17. New Zealand Health Information Service. In: Cancer: New Registrations and Deaths 1999 (2002) Wellington: New Zealand Ministry of Health.
18. Anonymous. Ethnicity in New Zealand: Recommendations for a Standard Classification. In: Discussion Paper (1990) Wellington: New Zealand Department of Statistics.
19. StataCorp. Stata Statistical Software: Release 8.0 (2003) College Station, TX: Stata Corporation.
20. Saydah SH, Platz EA, Rifai N, et al. Association of markers of insulin and glucose control with subsequent colorectal cancer risk. Cancer Epidemiol Biomarkers Prev (2003) 12:412–418.
21. Czyzyk A, Szczepanik Z. Diabetes mellitus and cancer. Eur J Intern Med (2000) 11:245–252.[CrossRef][Medline]
22. Kaaks R. Nutrition, hormones, and breast cancer: is insulin the missing link? Cancer Causes and Control (1996) 7:605–625.[CrossRef][Web of Science][Medline]
23. Stattin P, Bylund A, Rinaldi S, et al. Plasma insulin-like growth factor-I, insulin-like growth factor-binding proteins, and prostate cancer risk: a prospective study. J Natl Cancer Inst (2000) 92:1910–1917.
24. Chan JM, Stampfer MJ, Giovannucci E, et al. Plasma insulin-like growth factor-I and prostate cancer risk: a prospective study. Science (1998) 279:563–566.
25. Kaaks R, Toniolo P, Akhmedkhanov A, et al. Serum C-peptide, insulin-like growth factor (IGF)-I, IGF-binding proteins, and colorectal cancer risk in women. J Natl Cancer Inst (2000) 92:1592–1600.
26. Palmqvist R, Hallmans G, Rinaldi S, et al. Plasma insulin-like growth factor 1, insulin-like growth factor binding protein 3, and risk of colorectal cancer: a prospective study in northern Sweden. Gut (2002) 50:642–646.
27. Hankinson SE, Willett WC, Colditz GA, et al. Circulating concentrations of insulin-like growth factor-I and risk of breast cancer. Lancet (1998) 351:1393–1396.[CrossRef][Web of Science][Medline]
28. Muti P, Quattrin T, Grant BJ, et al. Fasting glucose is a risk factor for breast cancer: a prospective study. Cancer Epidemiol Biomarkers Prev (2002) 11:1361–1368.
29. Renehan AG, Zwahlen M, Minder C, et al. Insulin-like growth factor (IGF)-I, IGF binding protein-3, and cancer risk: systematic review and meta-regression analysis. Lancet (2004) 363:1346–1353.[CrossRef][Web of Science][Medline]
30. Saydah SH, Loria CM, Eberhardt MS, Brancati FL. Abnormal glucose tolerance and the risk of cancer death in the United States. Am J Epidemiol (2003) 157:1092–1100.
31. Coughlin SS, Calle EE, Teras LR, et al. Diabetes mellitus as a predictor of cancer mortality in a large cohort of US adults. Am J Epidemiol (2004) 159:1160–1167.
32. Jee SH, Ohrr H, Sull JW, et al. Fasting serum glucose level and cancer risk in Korean men and women. JAMA (2005) 293:194–202.
33. Batty GD, Shipley MJ, Marmot M, Smith GD. Diabetes status and post-load plasma glucose concentration in relation to site-specific cancer mortality: findings from the original Whitehall study. Cancer Causes Control (2004) 15:873–881.[CrossRef][Web of Science][Medline]
34. Hall GC, Roberts CM, Boulis M, et al. Diabetes and the risk of lung cancer. Diabetes Care (2005) 28:590–594.
35. Yu H, Spitz MR, Mistry J, et al. Plasma levels of insulin-like growth factor-I and lung cancer risk: a case-control analysis. J Natl Cancer Inst (1999) 91:151–156.
36. Lukanova A, Toniolo P, Akhmedkhanov A, et al. A prospective study of insulin-like growth factor-I, IGF-binding proteins-1, -2 and -3 and lung cancer risk in women. Int J Cancer (2001) 92:888–892.[CrossRef][Web of Science][Medline]
37. Khaw KT, Wareham N, Bingham S, et al. Preliminary communication: glycated hemoglobin, diabetes, and incident colorectal cancer in men and women: a prospective analysis from the European prospective investigation into cancer-Norfolk study. Cancer Epidemiol Biomarkers Prev (2004) 13:915–919.
38. New Zealand Health Information Service. Cancer: New Registrations and Deaths 2000 (2004) Wellington: New Zealand Ministry of Health.
39. Tavani A, Gallus S, Bertuzzi M, et al. Diabetes mellitus and the risk of prostate cancer in Italy. Eur Urol (2005) 47:313–317.[CrossRef][Web of Science][Medline]
40. Verlato G, Zoppini G, Bonora E, Muggeo M. Mortality from site-specific malignancies in type 2 diabetic patients from Verona. Diabetes Care (2003) 26:1047–1051.
41. Fujino Y, Mizoue T, Tokui N, Yoshimura T. Prospective study of diabetes mellitus and liver cancer in Japan. Diabetes Metab Res Rev (2001) 17:374–379.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  N. Brewer, C. S. Wright, N. Travier, C. W. Cunningham, J. Hornell, N. Pearce, and M. Jeffreys A New Zealand Linkage Study Examining the Associations Between A1C Concentration and Mortality Diabetes Care, June 1, 2008; 31(6): 1144 - 1149. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||