Hospitalitaliano.org.ar

E p i d e m i o l o g y / H e a l t h S e r v i c e s / P s y c h o s o c i a l R e s e a r c h
O R I G I N A L
Type 2 Diabetes and Subsequent
Incidence of Breast Cancer in the
Nurses’ Health Study
ARIN B. MICHELS, SCD
SUSAN E. HANKINSON, SCD
AREN G. SOLOMON, MD
GRAHAM A. COLDITZ, MD
RANK B. HU, MD
OANN E. MANSON, MD
ERNARD A. ROSNER, PHD
hormone with mitogenic effects in bothnormal and malignant breast tissue (8,9).
Insulin suppresses IGF binding protein-1and thus increases bioavailable IGF-1 OBJECTIVE — Hyperinsulinemia may promote mammary carcinogenesis. Insulin resistance
(10). The effect of estradiol on hormone- has been linked to an increased risk of breast cancer and is also characteristic of type 2 diabetes.
dependent breast cancer cell proliferation We prospectively evaluated the association between type 2 diabetes and invasive breast cancer incidence in the Nurses’ Health Study.
RESEARCH DESIGN AND METHODS — A total of 116,488 female nurses who were
insulin secretory defect causes type 2 di- 30 –55 years old and free of cancer in 1976 were followed through 1996 for the occurrence of abetes. Hyperinsulinemia with insulin re- type 2 diabetes and through 1998 for incident invasive breast cancer, verified by medical recordsand pathology reports.
sistance also has been postulated toincrease the risk of breast cancer (12–14).
RESULTS — During 2.3 million person-years of follow-up, we identified 6,220 women with
Obesity is associated with type 2 diabetes type 2 diabetes and 5,189 incident cases of invasive breast cancer. Women with type 2 diabetes and leads to a rise in endogenous estrogen had a modestly elevated incidence of breast cancer (hazard ratio [HR] ϭ 1.17; 95% CI 1.01– 1.35) compared with women without diabetes, independent of age, obesity, family history of breast cancer, history of benign breast disease, reproductive factors, physical activity, and alco- hol consumption. This association was apparent among postmenopausal women (1.16; 0.98 –1.62) but not premenopausal women (0.83; 0.48 –1.42). The association was predominant between type 2 diabetes and breast cancer among women with estrogen receptor–positive breast cancer (1.22; 1.01–1.47).
might have public health implications.
We used data from the large ongoing CONCLUSIONS — Women with type 2 diabetes may have a slightly increased risk of breast
Nurses’ Health Study cohort to investigate whether type 2 diabetes is associated withsubsequent incidence of breast cancer in- Diabetes Care 26:1752–1758, 2003
Breastcancerincidenceishigherin ars,andanimalfatsmayalsoplayanim- RESEARCHDESIGNAND
portant role. This Western lifestyle often results in insulin resistance, a condition characterized by a decreased sensitivity of target tissues to circulating insulin and The Nurses’ Health Study was established sulin inhibits the production of sex hor- observation (2), a lifestyle characterized which results in an increase in free steroid mailed questionnaire on their health sta- tus and on various potential risk factorsfor cancer, cardiovascular disease, and ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● other major illnesses. Participants receive From the 1Obstetrics and Gynecology Epidemiology Center, Brigham and Women’s Hospital, HarvardMedical School, Boston, Massachusetts; the 2Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts; the 3Department of Epidemiology, Harvard School of Public Health, Boston, Mas- sachusetts; the 4Department of Nutrition, Harvard School of Public Health, Boston, Massachusetts; and the thropometric, and lifestyle factors and on 5Department of Biostatistics, Harvard School of Public Health, Boston, Massachusetts.
Address correspondence and reprint requests to Karin B. Michels, Department of Epidemiology, Harvard School of Public Health, 677 Huntington Ave., Boston, MA 02115. E-mail: kmichels@rics.bwh.harvard.edu.
betes and breast cancer. The response rate Received for publication 7 September 2002 and accepted in revised form 21 February 2003.
Abbreviations: HR, hazard ratio; SHBG, sex hormone– binding globulin.
A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion 2003 by the American Diabetes Association.
DIABETES CARE, VOLUME 26, NUMBER 6, JUNE 2003 Michels and Associates
medical records to confirm the diagnosis.
considered, self-reported date of diagno- Ͼ99% of women whose reports were re- low-up were calculated as the time from not be obtained for 7% of the cases, anal- baseline to the date of return of the 1998 questionnaire or to the date of diagnosis gree of accuracy of the participants’ re- noma in situ, other cancer, death, or loss sidered to have “definite” type 2 (non– to follow-up, whichever occurred first.
the National Diabetes Data Group criteria breast carcinoma in situ (n ϭ 612) were betic subjects (exposed) from the time of for diabetes (15), did not meet criteria for censored from this analysis because we do reached one of the above-listed censoring rence of invasive breast cancer in our co- cancers progresses to become invasive.
betic subjects from the start of follow-up.
end point because differential use of pre- ventive services could lead to higher de- the time they were free of type 2 diabetes.
classified as probable or unlikely on the used to calculate the hazard of developing invasive breast cancer given a history of tionnaires to confirm and characterize di- models were adjusted for age (in months), family history of breast cancer in first- degree relative(s) (dichotomous), history ticipants with self-reported type 2 diabe- (n ϭ 3,302), as were those who reported type 1 diabetes (n ϭ 497) or type 2 dia- plementary questionnaire (n ϭ 1,091) or Self-reports of type 2 diabetes were con- if their date of diagnosis of diabetes was missing (n ϭ 4). Women who reported parity (0, 1, 2, 3, Ն4 children), age at endocrinologist for 61 (98%) of the cases.
onset of diabetes before age 30 years were birth of first child (Ͻ25, 25–29.9, 30 – they were more likely to have type 1 dia- betes (n ϭ 112). Participants also were excluded if their date of birth was missing analysis, the change in criteria did not af- (n ϭ 27), if they died shortly after agree- past user for Ͻ5 years, past user for Ն5 ing to participate in the study (n ϭ 2), if years, current user for Ͻ5 years, current they did not report their height (n ϭ 149), or if they developed breast cancer during ous physical activity (Ͻ1, 1–1.9, 2–3.9, follow-up but their date of diagnosis was 4 – 6.9, Ն7 h per week), and alcohol con- not available (n ϭ 29). This left a study the analysis if they developed breast can- cer or any other cancer, if they died, or if questionnaire between 1976 and 1996.
ing the previous 2 years, and if so, what Death Index is also routinely searched for variable as a covariate in all our analytic who reported breast cancer (or the next of abetes in 1976 were included in the anal- DIABETES CARE, VOLUME 26, NUMBER 6, JUNE 2003 Type 2 diabetes and breast cancer
Table 1—Age-standardized characteristics of 116,488 participants of the Nurses’ Health
Study according to diabetes status
for developing invasive breast cancer was 1.11 (95% CI 0.96 –1.28) for women with for potential confounders, family history of breast cancer, a history of benign breast disease, height, BMI at age 18 years, cur- rent BMI, age at menarche, parity, age at first child’s birth, menopausal status, age Mean number of children (among parous women) Mean age at first birth (among parous women) (years) icant (HR ϭ 1.17; 95% CI 1.01–1.35).
entered into the regression model as con- Use of postmenopausal hormones (among all women) sity possible given the observed data.
Physical activity (mean number of hours per week) model. The slight increase in the HR after Mean grams per day (among women who drink alcohol) adjustment for covariates was due to neg- *All values presented represent average values during follow-up; †a family history of breast cancer is defined as ever reporting breast cancer in a first-degree relative; ‡a history of benign breast disease is defined as ever reporting a history of fibrocystic or other benign breast disease.
age at menopause, and alcohol consump-tion. This negative confounding was estimates. Analyses were stratified by BMI stronger than the positive confounding by current BMI. A model adjusting for all co- variates except current BMI resulted in an mone use, family history of breast cancer, ing for all covariates, including current BMI, but not for the menopausal variables those who remained free of the disease.
cation), and by duration of diabetes (Յ5, duced an HR for the association of diabe- 5.1–10, 10.1–15, 15.1–20, Ն20 years) before a diagnosis of breast cancer. Sepa- 0.98 –1.31). A model, fully adjusted ex- cept for alcohol consumption, resulted in the association between a history of type 2 an HR ϭ 1.15 (95% CI 0.99 –1.33).
diabetes and estrogen receptor–positive ratio, were less physically active, drank hip ratio did not materially alter the esti- All tests of statistical significance are Table 2—History of type 2 diabetes and HRs of invasive breast cancer among 116,488 par-
ticipants of the Nurses’ Health Study, 1976 –1998
RESULTS — During 2.3 million per-
son-years of follow-up over 22 years,
*Hazard ratios and 95% CI adjusted for age, family history of breast cancer, history of benign breast disease, was known, 2,915 had estrogen receptor– height, BMI at age 18 years, current BMI, age at menarche, parity, age at first child’s birth, menopausal status, positive and 989 had estrogen receptor– age at menopause, use of postmenopausal hormones, physical activity, and alcohol consumption.
DIABETES CARE, VOLUME 26, NUMBER 6, JUNE 2003 Michels and Associates
Table 3—History of type 2 diabetes and HRs of invasive breast cancer subtypes among participants of the Nurses’ Health Study, 1976 –1998
*HR and 95% CI adjusted for age, family history of breast cancer, history of benign breast disease, height, BMI at age 18 years, current BMI, age at menarche, parity,age at first child’s birth, menopausal status, age at menopause, use of postmenopausal hormones, physical activity, and alcohol consumption; †receptor status wasnot known for all breast cancer cases.
mate (HR ϭ 1.16; 95% CI 1.00 –1.34).
association between a history of diabetes 4). There was no statistically significant not available at baseline and not available for a considerable number of participants P ϭ 0.16). The association between type 2 without considering waist-to-hip ratio.
appreciably modified by BMI (Table 4).
not substantially modified by a family his- cancer risk was slightly lower for diabetic tory of breast cancer or the use of insulin Table 4—History of type 2 diabetes and HRs of invasive breast cancer among different subgroups of the Nurses’ Health Study, 1976 –1998
*The reference group for calculation of each HR is the women without diabetes in the same stratum for the given covariates; †HR and 95% CI adjusted for age, familyhistory of breast cancer, history of benign breast disease, height, BMI at age 18 years, current BMI, age at menarche, parity, age at first child’s birth, menopausal status,age at menopause, use of postmenopausal hormones, physical activity, and alcohol consumption; ‡information on menopausal status was not available for allparticipants; §includes insulin and sulfonylureas; ¶additionally adjusted for duration of diabetes.
DIABETES CARE, VOLUME 26, NUMBER 6, JUNE 2003 Type 2 diabetes and breast cancer
Table 5—Years since diagnosis of type 2 diabetes and HRs of invasive breast cancer among
participants of the Nurses’ Health Study
tering height, BMI at age 18 years, andcurrent BMI as continuous variables intothe regression model (37). Neither cur- the diabetes– breast cancer association.
Central obesity, which is an even stronger predictor of insulin resistance and type 2 crease breast cancer risk among both pre- ciation between type 2 diabetes and breast *Hazard ratios and 95% CI adjusted for age, family history of breast cancer, history of benign breast disease, height, BMI at age 18 years, current BMI, age at menarche, parity, age at first child’s birth, menopausal status, age at menopause, use of postmenopausal hormones, physical activity, and alcohol consumption.
risk for breast cancer was nonsignificantlylower in premenopausal women with decades ago to have higher rates of diabe- studies that followed, some reported ele- vated risk ratios for breast cancer among diabetes between 10 and 15 years before a women with type 2 diabetes (24 –27), but diagnosis of breast cancer than for women and obesity is associated with reduced en- evate free plasma estrogen levels through ered to reveal a modest association. Fur- were few cases in this group. The relation accounted for general and/or central obe- sity, which is associated with both type 2 quent breast cancer risk was similar when diabetes and breast cancer. Weiss et al.
control study in the U.S., found no signif- diabetes and breast cancer by a family his- CONCLUSIONS — These data from
tio ϭ 1.13; 95% CI 0.7–1.90). The Iowa hort (41). In that analysis, however, the slightly but significantly higher risk of de- Women’s Study did not reveal an overall no diabetes. The elevated risk was appar- both BMI and waist-to-hip ratio (36).
though the association between type 2 di- Study provide the largest population with ined a possible detection bias by restrict- abetes and breast cancer did not appear to trolled for obesity in our analyses as well several previous reports of a slightly ele- vated risk (21–27,29,33). Most previous study, we have to consider the possibility studies, however, did not have sufficient that the observed association was a result of such biases. Furthermore, waist-to-hip ratio measurements were not available for described in patients with cancer (19). In the entire cohort, and thus, our final re- largely independent of self-reported adi- sults were not adjusted for central obesity.
strated that tumor slices sustained higher rates of glucose utilization and lactic acid production than normal tissue sections.
obesity as closely as possible using the ob- potential confounders, indicating that the DIABETES CARE, VOLUME 26, NUMBER 6, JUNE 2003 Michels and Associates
4. Kaaks R: Nutrition, hormones, and breast 17. American Diabetes Association: Clinical cancer: is insulin the missing link? Cancer ing for type 2 diabetes. Diabetes Care 21: to be considered as a possible explanation for the modest increase in risk observed.
18. Cox DR: Regression models and life ta- globulin production by growth factors.
bles. J R Stat Soc 32:187–220, 1972 19. Freund E: Zur Diagnose des Carcinoms.
6. Singh A, Hamilton-Fairley D, Koistinen R, Wiener Medizinische Bl 8:268 –269, 1885 are therefore misclassified as nondiabetic 20. Warburg O: The Metabolism of Tumors.
Effect of insulin-like growth factor-type I 21. Glicksman AS, Rawson RW: Diabetes and women classified as free of type 2 diabetes tients with cancer. Cancer 9:1127, 1956 would be classified as having type 2 dia- toma cells. J Endocrinol 124:R1–R3, 1990 7. Rosner W: The functions of corticoste- duction in nondifferential misclassification glucose tolerance. Arch Gyna¨kol 220:73– binding globulin: recent advances. Endocr diabetes and breast cancer observed.
mammary slices in organ culture. I. Meth- tions. Arch Gyna¨kol 221:83–91, 1976 33% in the U.S.(43), reflecting the rapid tions on the effects of insulin. Exp Cell Res 24. de Waard F, Baanders-van Halewijn EA: A increase in the prevalence of obesity (44).
9. Van der Burg BB, Rutterman GR, Blanken- women. Int J Cancer 14:153–160, 1974 25. O’Mara BA, Byers T, Schoenfeld E: Diabe- tes mellitus and cancer risk: a multisite synergistic action of insulin and estrogen.
case-control study. J Chronic Dis 38:435– diabetes was estimated at 16 million (43).
J Cell Physiol 134:101–108, 1988 Thus, even a slight increase in breast can- 10. Conover CA, Lee PD, Kanaley JA, Clark- 26. Unger C, Rageth JC, Wyss P, Spillmann of public health concern. If hyperinsulin- faktoren. Schweiz Med Wochenschr 121: emia plays a role in breast cancer patho- tein-1 in obese and nonobese humans.
J Clin Endocrinol Metab 74:1355–1360, insulin sensitivity such as exercise and di- 11. Lippman ME, Bolan G: Oestrogen respon- endometrial and breast cancer in patients with diabetes mellitus. Int J Cancer 71: (45), but this assumption requires further term tissue culture. Nature 256:592–594, 28. Adami HO, Rimsten A: Prevalence of hy- pertension and diabetes in breast cancer: jen WJ: Insulin resistance and breast-cancer age-matched, non-hospitalized controls.
Acknowledgments — This study was sup-
risk. Int J Cancer 52:511–516, 1992 ported by Public Health Service Grant CA- 13. Yang G, Lu G, Jin F, Dai Q, Best R, Shu 29. Rgozzino M, Melton LJ, Chu CP, Palumbo 87969 from the National Cancer Institute, National Institutes of Health, Department of study of blood C-peptide level and breast residents with diabetes mellitus. J Chronic cancer risk. Cancer Epidemiol Biomarkers assistance and Maureen Ireland for help with 30. Kopp S, Tanneberger S, Mohner M, Kieser R: Diabetes and breast cancer risk. Int J References
menopausal breast cancer risk. Breast Cancer Res Treat 47:111–120, 1998 and history of selected medical conditions Institute for Cancer Research: Food, Nutri- 15. National Diabetes Data Group: Classifica- linked with female hormones. Eur J Can- tion and the Prevention of Cancer: A Global Perspective. Washington, DC, American ance. Diabetes 28:1039 –1057, 1979 2. Hulka BS, Stark AT: Breast cancer: cause and prevention. Lancet 346:883– 887, mellitus. Cancer Causes Control 2:307– 3. Rockhill B, Willett WC, Hunter DJ, Man- D’Avanzo B, Boyle P: A case-control study of diabetes mellitus and cancer risk. Br J activity and breast cancer risk. Arch Intern women. Arch Intern Med 151:1141–1147, DIABETES CARE, VOLUME 26, NUMBER 6, JUNE 2003 Type 2 diabetes and breast cancer
38. Huang Z, Willett WC, Colditz GA, Hunter tion-based cohorts of diabetic patients. waist:hip ratio, and risk of breast cancer in the Nurses’ Health Study. Am J Epide- Examination Survey, 1988 –1994. Diabe- lone KE, Schoenberg JB: Breast cancer risk medical conditions. Int J Epidemiol 28: JS: Diabetes trends in the U.S. 1990 –1998.
Diabetes Care 23:1278 –1283, 2000 36. Sellers TA, Sprafka JM, Gapstur SM, Rich gain on breast cancer risk. JAMA 278: 44. Flegal KM, Carroll MD, Kuczmarski RJ, SS, Potter JD, Ross JA, McGovern PG, Nel- 40. Gerber M: Re: Body size and breast cancer the United States: prevalence and trends, (Letter). Am J Epidemiol 145:669, 1997 1960 –1994. Int J Obes 22:39 – 47, 1998 menopausal breast cancer? Epidemiology 41. Sellers TA, Anderson KE, Olson JE, Fol- som AR: Family histories of diabetes mel- postmenopausal breast cancer. Epidemiol- incidence of type 2 diabetes with lifestyle body size to health outcomes? Am J Epide- 42. Harris MI, Flegal KM, Cowie CC, Eber- intervention or metformin. N Engl J Med DIABETES CARE, VOLUME 26, NUMBER 6, JUNE 2003

Source: http://www.hospitalitaliano.org.ar/archivos/cursos_attachs/307.pdf