Abstract
Purpose
Diabetes has been associated with increased risk of breast cancer in a number of epidemiologic studies, but its effects on survival among women diagnosed with breast cancer have been examined less frequently. Importantly, prior investigations have rarely considered the influence of factors associated with diabetes such as obesity, age at diabetes diagnosis, duration of diabetes, or diabetes treatments.
Methods
We evaluated the effect of self-reported diabetes on breast cancer incidence and mortality in the Long Island Breast Cancer Study Project, which includes 1,447 breast cancer cases and 1,453 controls. Follow-up data for all-cause (n = 395) and 5-year breast cancer-specific mortality (n = 104) through December 2005 were determined for case women from the National Death Index. Adjusted logistic regression and Cox proportional hazards models were used to estimate odds ratios (OR) and hazards ratios (HR), respectively.
Results
Postmenopausal women with diabetes were at increased risk of developing breast cancer [OR = 1.35; 95 % confidence interval (CI) = 0.99–1.85], as were those who were not of white race regardless of menopausal status [OR = 3.89; 95 % CI = 1.66–9.11]. Among case women, diabetes was associated with a modestly increased risk of death from all causes [HR = 1.65; 95 % CI = 1.18–2.29], an association that was stronger in women who were obese at breast cancer diagnosis [HR = 2.49; 94 % CI = 1.58–3.93]. In analyses restricted to diabetics, there was no statistically significant effect of duration of diabetes or type of treatment on breast cancer incidence or mortality.
Conclusions
Our findings suggest that diabetes may increase incidence of breast cancer in older women and non-whites, and mortality due to all causes.
Keywords: Breast cancer, Diabetes, Survival
Introduction
Diabetes and breast cancer are increasing health concerns for women worldwide, particularly in older women [1]. Approximately 90–95 % of all cases of diabetes diagnosed are classified at type 2 diabetes (T2D), and it is currently estimated that over 10 % of women in the United States over the age of 20 have T2D, including those with undiagnosed disease [2]. T2D is characterized by insulin resistance and hyperinsulinemia, is associated with high BMI that is a well-established risk factor for breast cancer in postmenopausal women [3], and is associated with poor prognosis regardless of menopausal status [4, 5]. It is thought that T2D affects risk of developing breast cancer through the direct effects of insulin on breast tissue, or indirectly through the increase in sex steroids due to the inhibition of sex hormone-binding globulin (SHBG), increased insulin-like growth factor-I (IGF-I) production, and disruption of adipokines [6, 7]. These changes in circulating hormone levels can lead to abnormalities in cellular growth and regulation [8].
T2D has primarily been shown to increase the risk of breast cancer incidence, with a recent large meta-analysis reporting about a 20 % increase in risk for both case–control and cohort studies [9]. However, few population-based studies of survival after a breast cancer diagnosis have reported on this potentially important pathway [10–12] and even fewer have reported breast cancer-specific mortality [11, 13]. The biologic plausibility of an association with mortality is strong, as many components of diabetes have been linked to breast cancer incidence and prognosis including centralized obesity, insulin resistance, and raised fasting plasma glucose [6]. Furthermore, hyperinsulinemia has been associated with risk of recurrence and mortality in breast cancer [14].
For women with diabetes, the risk of developing or dying from breast cancer may also be affected by variations in the management of their diabetes, including types and length of treatments. There is encouraging evidence that metformin, an insulin sensitizer and the most commonly used therapy for patients with T2D, may decrease breast cancer risk by reducing hepatic glucose output [15]. Other treatments, however, such as insulin, or secretogogues, that stimulate insulin production, may increase cancer risk and death [16, 17]. There have been few studies that have investigated diabetes treatment on breast cancer risk [18, 19], and we are not aware of any studies that have investigated the influence of diabetes treatment in terms of survival.
Because diabetes is increasingly becoming a worldwide health problem where the number of women at risk is growing, it is important to understand the impact of diabetes and diabetes treatments on risk of developing breast cancer and survival after a breast cancer diagnosis. To investigate the effects of diabetes and diabetes treatments on risk of breast cancer and mortality, we conducted a large population-based study.
Materials and methods
This study draws upon data that were collected from participants as part of the Long Island Breast Cancer Study Project (LIBCSP), a population-based study of English-speaking residents of Nassau and Suffolk counties of Long Island, NY [20]. The study reported here utilizes resources from both the case–control and the follow-up studies of the LIBCSP, as described below.
Study population
Case–control study
Eligible case participants were women newly diagnosed with a first, primary in situ or invasive breast cancer between August 1, 1996 and July 31, 1997. Cases were identified using a rapid reporting system established specifically for the LIBCSP and were confirmed by physicians’ and medical records. The attending physician was contacted to confirm study eligibility and to seek permission to contact the patient. Controls were women who were residents of the same two counties, frequency matched by 5-year age group to the expected age distribution of cases. Potentially eligible control women were identified by Waksberg’s method of random digit dialing (RDD) [21] for those under 65 years of age, and by Health Care Finance Administration (HCFA) rosters for those 65 years of age and older. Institutional review board (IRB) approval of the study protocol was obtained from each collaborating institution and participating hospital, and written informed consent was obtained from each participant prior to the baseline interview. A total of 1,508 women with breast cancer, of which 1,273 had invasive breast cancer, and 1,556 control women participated in the baseline case–control study interview. In the LIBCSP population, 93 % or participants reported their race as white, 4.8 % as black, and 2.2 % as other, which is consistent with the U.S. Census data for these two NY counties [20].
Follow-up study
The population-based cohort of women with breast cancer who participated in the baseline interview (n = 1,508) have been followed to determine complete first course of treatment for the first primary breast cancer diagnosis and vital status.
Data collection
Baseline, case–control data
Diabetes and most of the covariate data used in this analysis were collected as part of the LIBCSP baseline case–control interview, which for case women occurred about 2 months after the initial breast cancer diagnosis. The baseline structured questionnaire was administered in-home by a trained interviewer and took approximately 2 h to complete. Information obtained from the baseline questionnaire includes reproductive and menstrual history, exogenous hormone use (hormone replacement or oral contraceptives), family history of cancer, physical activity, smoking history, alcohol intake, demographic characteristics, and diabetes status. Descriptive characteristics for the entire LIBCSP study have been previously published [20]. As part of the baseline interview, a modified Block food frequency questionnaire was self-completed by 98 % of all LIBCSP respondents; these data were used to estimate intake of total fat and calories in the year prior to the baseline interview.
Additionally, as part of the baseline case–control study, medical records of the cases were abstracted for tumor stage, estrogen receptor (ER) status, progesterone receptor (PR) status, and initial course of treatment. Nearly two-thirds of the baseline interviews with cases occurred prior to the initiation of chemotherapy.
Diabetes status
Diabetes status was determined at the baseline, case–control interview. Participants were asked whether they had ever been told by a physician that they had diabetes, sugar diabetes, or high blood sugar. There were 7 participants (3 cases, 4 controls) with missing information on diabetes status. No distinction was indicated as to diabetes type, however, based on prior literature, in order to increase the probability that our population was limited to those with type 2 diabetes, women diagnosed with diabetes before the age of 30 were excluded from the analyses (n = 19) [22, 23], resulting in a total of 1,495 cases and 1,543 controls available for analysis. If the participant had reported having diabetes, they were asked when they were diagnosed and were asked about medication use. Medication use was determined from the questionnaire where women responded to a question asking whether they had taken medication for diabetes for 3 or more consecutive months. Women reported the names of the medications used, and the duration they used each medication. Reports of using insulin, hepatic glucose production inhibitors (metformin), and/or an insulin secretogogue (majority of which were sulfonylureas, some were meglitinides) were classified as having used a medication.
Follow-up data among women with breast cancer
For women with breast cancer who participated in the LIBCSP baseline interview, follow-up telephone interviews were conducted in 2002–2004 by trained interviewers using a structured questionnaire with 1,098 case participants (of which 8 % were completed with a proxy). The follow-up interview included ascertainment of information on completed course of treatment for the first primary breast cancer diagnosis. These self-reported treatment data were compared with updated information from the medical records, which were retrieved as part of the follow-up and abstracted for 598 breast cancer cases. Trained abstractors reviewed medical records to determine the complete course of treatment for the first primary breast cancer diagnosis, and these data were compared with the respondent’s self-reported treatment from the follow-up interview. A very high concordance was found between information abstracted from medical records and self-reported radiation therapy (Kappa = 0.97), chemotherapy, (Kappa = 0.96), and hormone therapy (Kappa = 0.92). Thus, self-reported breast cancer treatment was used for this analysis. At the time of the follow-up medical record review, nodal status for each woman’s first primary breast cancer diagnosis was also ascertained.
Study outcome for the follow-up analyses
For the LIBCSP follow-up, the National Death Index (NDI) was used to ascertain all-cause and breast cancer-specific mortality among case participants. Cases were followed from diagnosis until December 31, 2005 for a mean of 96.4 months (range, 2.7–113.0). Among the 1,495 women in this study diagnosed with breast cancer, 303 (20.3 %) deaths occurred. Based on International Classification of Diseases (ICD) codes 174.9 and C-50.9 listed as a primary or secondary code on the death certificate, 106 (35.0 %) deaths after 5 years of follow-up were due to breast cancer.
Statistical methods
Risk of developing breast cancer and demographic factors were compared between participants with a self-reported diabetes diagnosis and those without a diabetes diagnosis using t tests and chi-square tests. All tests of statistical significance are two-sided and considered significant at the 0.05 level. All analyses were carried out using the statistical software package SAS version 9.2 (SAS Institute Inc., Cary, NC).
Case–control analyses
For the assessment of the association between the risk of developing breast cancer and a history of diabetes, odds ratios (ORs) and 95 % confidence intervals (CI) were calculated using unconditional logistic regression models [24]. All models were adjusted for 5-year age group at diagnosis. Additional factors considered as potential confounders included: variables related to demographic factors (race, income, education, marital status, religion), reproduction (parity, age at first live birth, breast feeding), and menstruation (age at menarche, menopausal status). Exogenous hormone use was also considered (hormonal birth control, hormone replacement among postmenopausal women) as was medical history (benign breast disease, family history of breast cancer), and lifestyle factors (alcohol consumption, dietary fat and total caloric intake, cigarette smoking, physical activity, and body size measured as body mass index [BMI; weight in kilograms divided by height in meters squared]). Using manual backward elimination, potential confounders were removed from models. Variables remained in the final models if their exclusion changed the estimate of effect by ≥10 % [24]. Adjustment for most covariates did not alter the estimates of effect by more than 10 %, and therefore associations reported are only adjusted for 5-year age group, menopausal status (pre- vs. postmenopause), race (whites vs. blacks and others), and body size (BMI<30 vs. BMI ≥30). All case–control analyses were carried out on a dataset restricted to participants without missing values for menopausal status, race, or obesity resulting in a final dataset of 1,447 cases and 1,453 controls.
We also evaluated the effects of age at diabetes diagnosis, duration of diabetes, and diabetes medication use among those who reported having been diagnosed with diabetes. ORs and 95 % CI were calculated for the association between breast cancer and age at diabetes diagnosis (±55 years), median duration of diabetes (±7 years), whether they had ever received medication for diabetes for 3 or more consecutive months, and type of medication (insulin, metformin, secretogogues). Age at diabetes diagnosis and duration of diabetes were mutually adjusted for each other. Additionally, all types of diabetes medications we evaluated in the same model.
Effect measure modification on the multiplicative scale between categorical covariates was examined comparing the log likelihood statistic for logistic regression models with and without the cross-product terms [25]. We evaluated models stratified by age at breast cancer diagnosis (±65 years), menopausal status (pre- and postmenopause), BMI one year prior to breast cancer diagnosis (<25, ≥25–<30, ≥30), lifetime average physical activity (ever, never), lifetime average alcohol consumption (ever, never), median average daily caloric intake (±1,251.1 kcal/day), hormone replacement (ever, never), and race (white, black, other).
Survival analysis
Cox proportional hazards regression [25] was used to estimate hazard ratios (HR) and 95 % confidence intervals (CI) for all-cause and breast cancer-specific mortality in relation to a diabetes diagnosis reported at the time of baseline interview. Since most women who die as a result of their breast cancer diagnosis usually do so within 5 years, we presented only 5-year survival for breast cancer-specific deaths. Models re-run with follow-up time through 2005 were nearly identical to those limited to 5 years.
To investigate the differences in associations between diabetes and survival, analyses were stratified by selected covariates: age at breast cancer diagnosis (±65 years), menopausal status (pre- and postmenopause), BMI one year prior to breast cancer diagnosis (<25, ≥25–<30, ≥30), lifetime average physical activity (ever, never), lifetime average alcohol consumption (ever, never), median average daily caloric intake (±1,251.1 kcal/day), hormone replacement (ever, never), and race (white, black, other). Associations were also evaluated by stratification on the tumor characteristics, ER status (negative, positive), PR status (negative, positive), tumor stage (in situ, invasive), nodal status (node negative, node positive), and tumor size (<2 cm, ≥2 cm).
All models were adjusted for age at diagnosis. In addition to consideration of the covariates listed above for the case–control analyses, for the survival analyses, we also considered as potential confounders other factors including history of co-morbidities reported at the baseline interview (high cholesterol, history of blood clots, hypertension, previous myocardial infarction [MI], and stroke), tumor characteristics (tumor stage, tumor size, and nodal status), and treatment undergone for the original breast cancer diagnosis. Adjustment for most covariates did not alter the estimates of effect by more than 10 %, and therefore associations reported are adjusted for 5-year age group, menopausal status, race, body size, and MI only. All survival analyses were carried out on a dataset restricted to participants with complete data for menopausal status, race, obesity, and MI resulting in a final dataset of 1,444 breast cancer cases.
Results
In Table 1, we report the distribution of characteristics of the LIBCSP stratified by self-reported diabetes from questionnaire data recorded in 1996–1997. There were 219 (7.2 %) participants in our population who reported having a diabetes diagnosis. Compared to non-diabetics, those with diabetes tended to be postmenopausal at diagnosis, have a higher BMI at diagnosis, and were less likely to engage in physical activity, drink alcohol or take hormone replacement than women without diabetes. Mean follow-up for women with breast cancer was 86.8 months for those with diabetes and 97.4 months for non-diabetics. There were no differences between those with and without diabetes according to tumor characteristics (ER/PR positivity, nodal status) or treatment type.
Table 1.
Selected characteristics of women diagnosed with a first primary breast cancer in 1996–1997, Long Island Breast Cancer Study Project, by diabetes status prior to diagnosis
| Characteristics | Diabetes | No diabetes | p Value |
|---|---|---|---|
| Mean age (years) | 63.6 | 57.4 | <0.0001 |
| Breast cancer diagnosis (%) | 55.7 | 48.7 | 0.046 |
| 5-year death due to breast cancer (%among cases) | 11.5 | 6.7 | 0.049 |
| Menopausal status (%postmenopausal) | 86.2 | 65.6 | <0.0001 |
| Race (% Caucasian) | 84.9 | 93.4 | <0.0001 |
| BMI, mean | 30.9 | 26.1 | <0.0001 |
| Energy intake (Kcal/day), mean | 1,316 | 1,343 | 0.594 |
| Regular physical activity (≥3 h/week) (%) | 61.9 | 71.8 | 0.002 |
| Ever regular alcohol drinker (%) | 42.9 | 64.0 | <0.0001 |
| Ever take hormone replacement (%) | 20.4 | 33.5 | 0.0003 |
| Breast cancer stage (% invasive among cases) | 87.7 | 84.1 | 0.287 |
| ER + tumor (% among cases) | 73.6 | 73.6 | 0.997 |
| PR + tumor (% among cases) | 66.7 | 64.0 | 0.619 |
| Chemotherapy (% among cases) | 33.9 | 41.9 | 0.239 |
| Hormone therapy (% among cases) | 72.2 | 60.6 | 0.089 |
| Radiation therapy (% among cases) | 50.0 | 61.7 | 0.081 |
Case–control analysis
After adjustment for obesity, menopausal status, and race, the odds ratio for the association between diabetes and risk of developing breast cancer was slightly elevated, although not statistically significant (OR = 1.27; 95 % CI = 0.95–1.69) (Table 2). Estimates did not change appreciably with additional adjustment for other potential confounders such as physical activity, age at menarche, alcohol consumption, daily caloric intake, hormonal birth control, hormone replacement, education, or income (data not shown). We found a modest association of diabetes on the risk developing breast cancer among postmenopausal women, which was of borderline significance (OR = 1.35; 95 % CI = 0.99–1.85). The diabetes–breast cancer association was most pronounced when we limited the analysis to women over the age of 65 at breast cancer diagnosis (OR = 1.59; 95 % CI = 1.04–2.44).
Table 2.
Odds ratios (95% CI) for developing breast cancer according to diabetes diagnosis, stratified by breast cancer risk factors among women diagnosed with breast cancer in 1996–1997, Long Island Breast Cancer Study Project
| Diabetes
|
No diabetes
|
ORa | 95 % CI | ORb | 95 % CI | |||
|---|---|---|---|---|---|---|---|---|
| Cases | Controls | Cases | Controls | |||||
| All women | 121 | 95 | 1,326 | 1,358 | 1.24 | 0.93–1.64 | 1.27 | 0.95–1.69 |
| Menopausal status | ||||||||
| Premenopausal women | 12 | 18 | 456 | 474 | 0.68 | 0.32–1.43 | 0.86 | 0.40–1.86 |
| Postmenopausal women | 109 | 77 | 870 | 884 | 1.40 | 1.03–1.91 | 1.35 | 0.99–1.85 |
| Age at breast cancer diagnosis | ||||||||
| <65 years | 50 | 58 | 895 | 968 | 0.93 | 0.63–1.38 | 1.03 | 0.69–1.55 |
| ≥65 years | 71 | 37 | 431 | 390 | 1.74 | 1.14–2.64 | 1.59 | 1.04–2.44 |
| BMI (kg/m2) | ||||||||
| BMI >18.5 to <30 | 67 | 46 | 1,061 | 1,101 | 1.45 | 0.99–2.14 | 1.52 | 1.03–2.25 |
| BMI ≥30 | 54 | 49 | 265 | 257 | 0.98 | 0.64–1.51 | 0.99 | 0.65–1.53 |
| Ever regular lifetime physical activityc | ||||||||
| None | 52 | 31 | 373 | 371 | 1.58 | 0.99–2.53 | 1.60 | 0.99–2.59 |
| Any | 68 | 64 | 946 | 983 | 1.06 | 0.74–1.51 | 1.09 | 0.75–1.56 |
| Ever alcohol drinker | ||||||||
| Never drinker | 69 | 48 | 495 | 515 | 1.47 | 1.00–2.18 | 1.41 | 0.94–2.09 |
| Ever drinker | 52 | 47 | 831 | 843 | 1.03 | 0.68–1.55 | 1.14 | 0.75–1.73 |
| Average daily caloric intaked | ||||||||
| <1,275.1 kcal/day | 69 | 39 | 665 | 666 | 1.67 | 1.10–2.51 | 1.66 | 1.10–2.52 |
| ≥1,275.1 kcal/day | 43 | 50 | 630 | 651 | 0.85 | 0.55–1.30 | 0.89 | 0.58–1.38 |
| Ever take hormone replacemente | ||||||||
| No | 88 | 59 | 573 | 587 | 1.50 | 1.06–2.13 | 1.38 | 0.96–1.96 |
| Yes | 20 | 18 | 296 | 296 | 1.11 | 0.57–2.14 | 1.25 | 0.64–2.47 |
| Race | ||||||||
| White | 96 | 86 | 1,257 | 1,251 | 1.07 | 0.79–1.45 | 1.05 | 0.77–1.43 |
| Other | 24 | 9 | 103 | 66 | 3.14 | 1.25–7.87 | 3.89 | 1.66–9.11 |
Adjusted for 5-year age group
Additionally adjusted for menopausal status, race (white, other), and obesity (BMI <30, BMI ≥30)
12 missing responses for any lifetime physical activity
87 missing responses for average daily caloric intake
Among postmenopausal women
When evaluating the potential effect measure modification of lifestyle factors on the association between diabetes and breast cancer, we found no association between those who were obese at diagnosis (OR = 0.99; 95 % CI = 0.65–1.53), whereas those who were not obese had an increased association (OR = 1.52; 95 % CI = 1.03–2.25). The diabetes–breast cancer associations were strengthened with decreasing BMI and were over twofold for those with a BMI between 18.5 and 25 at diagnosis (OR = 2.13; 95 % CI = 1.10–4.13) (data not in table). Similarly, we saw stronger associations for those who gained less than 30 lbs since age 20 (OR = 1.72; 95 % CI = 0.92–3.20) and had either lost weight or gained less than 13.5 lbs after the age 50 (OR = 1.65; 95 % CI = 1.00–2.74; OR = 1.50; 95 % CI = 0.74–3.04, respectively; data not in table). We also observed increased associations among those who consumed fewer daily calories (OR = 1.66; 95 % CI = 1.10–2.52). Among women who did not engage in regular lifetime physical activity (OR = 1.60; 95 % CI = 0.99–2.59), we found that women with diabetes had increased risk of developing breast cancer, an association that approached statistical significance. We found no evidence of effect modification of the association between diabetes and breast cancer for use of postmenopausal hormones.
We observed modification of the effect of diabetes on breast cancer for race. After additional adjustment for both income and education, white women had no increased risk of developing breast cancer with diabetes; however, those of races other than white had over a threefold increase in the OR (OR = 3.89; 95 % CI = 1.66–9.11), but the estimate was unstable as reflected by the wide confidence intervals.
Survival analysis
Through 2005, there were 295 deaths overall and 148 deaths due to breast cancer, 104 of which occurred within 5 years of diagnosis. After adjusting for age, menopausal status, race, obesity, and history of MI, women who reported having a diabetes diagnosis had increased all-cause mortality compared to women who did not have diabetes (HR = 1.65; 95 % CI = 1.18–2.29) (Table 3). This association was stronger among postmenopausal women. Additionally, women who were older age at diagnosis (≥65 years) also had higher mortality than those who were younger.
Table 3.
Hazard ratios (95 % CI) for all-cause mortality through 2005 and 5-year breast cancer mortality according to diabetes diagnosis, stratified by breast cancer risk factors among women diagnosed with breast cancer in 1996–1997, Long Island Breast Cancer Study Project
| All cause survival
|
Breast cancer 5-year survival
|
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Diabetes
|
No diabetes
|
HRa | 95 % CI | Diabetes
|
No diabetes
|
HRa | 95 % CI | |||||
| Deaths | Total pop | Deaths | Total pop | Deaths | Total pop | Deaths | Total pop | |||||
| All women | 49 | 121 | 246 | 1,326 | 1.65 | 1.18–2.29 | 14 | 121 | 90 | 1,326 | 1.17 | 0.63–2.19 |
| Menopausal status | ||||||||||||
| Premenopausal women | 4 | 12 | 63 | 456 | 1.30 | 0.43–3.89 | 3 | 12 | 32 | 457 | 1.76 | 0.48–6.49 |
| Postmenopausal women | 45 | 109 | 183 | 870 | 1.74 | 1.23–2.45 | 11 | 109 | 58 | 870 | 1.12 | 0.56–2.25 |
| Age at breast cancer diagnosis | ||||||||||||
| <65 years | 15 | 50 | 123 | 895 | 1.47 | 0.81–2.68 | 5 | 50 | 62 | 895 | 0.78 | 0.28–2.16 |
| ≥65 years | 34 | 71 | 123 | 431 | 1.63 | 1.09–2.43 | 9 | 71 | 28 | 431 | 1.60 | 0.72–3.54 |
| BMI (kg/m2) | ||||||||||||
| BMI >18.5 to <30 | 19 | 67 | 186 | 1,060 | 1.20 | 0.73–1.99 | 7 | 67 | 64 | 1,060 | 1.49 | 0.64–3.48 |
| BMI ≥30 | 30 | 54 | 60 | 263 | 2.49 | 1.58–3.93 | 7 | 54 | 26 | 263 | 1.22 | 0.51–2.91 |
| Ever regular lifetime physical activityb | ||||||||||||
| None | 24 | 52 | 96 | 373 | 1.55 | 0.95–2.53 | 8 | 52 | 36 | 373 | 1.22 | 0.51–2.88 |
| Any | 25 | 68 | 149 | 946 | 1.76 | 1.12–2.78 | 6 | 68 | 54 | 946 | 1.04 | 0.40–2.72 |
| Ever regular alcohol drinker | ||||||||||||
| Never drinker | 27 | 69 | 98 | 495 | 1.68 | 1.07–2.63 | 6 | 69 | 38 | 495 | 0.73 | 0.29–1.82 |
| Ever drinker | 22 | 52 | 148 | 831 | 1.61 | 0.98–2.64 | 8 | 52 | 52 | 831 | 2.14 | 0.90–5.10 |
| Average daily caloric intakec | ||||||||||||
| <1,275.1 KCAL | 26 | 69 | 118 | 665 | 1.60 | 1.01–2.53 | 4 | 69 | 42 | 665 | 0.73 | 0.24–2.15 |
| ≥1,275.1 KCAL | 18 | 43 | 119 | 630 | 1.81 | 1.07–3.08 | 8 | 43 | 45 | 630 | 2.05 | 0.89–4.69 |
| Ever take hormone replacementd | ||||||||||||
| No | 38 | 88 | 141 | 586 | 1.70 | 1.17–2.48 | 9 | 88 | 46 | 586 | 1.05 | 0.49–2.24 |
| Yes | 6 | 20 | 45 | 299 | 1.86 | 0.77–4.50 | 1 | 20 | 12 | 299 | 1.01 | 0.13–7.92 |
| Estrogen–Progesterone Receptor Statuse | ||||||||||||
| ER−PR− | 12 | 17 | 54 | 187 | 1.53 | 0.76–3.10 | 5 | 17 | 32 | 187 | 0.79 | 0.27–2.31 |
| ER−PR + breast cancer | 2 | 6 | 13 | 44 | 0.69 | 0.11–4.22 | 1 | 6 | 4 | 44 | 0.32 | 0.01–12.5 |
| ER + PR− breast cancer | 5 | 12 | 33 | 127 | 2.33 | 0.83–6.55 | 2 | 12 | 14 | 127 | 1.82 | 0.36–9.25 |
| ER + PR+ | 21 | 52 | 87 | 510 | 1.70 | 1.02–2.85 | 4 | 52 | 22 | 510 | 1.32 | 0.41–4.24 |
| Estrogen Receptor Statuse | ||||||||||||
| ER− | 14 | 23 | 67 | 231 | 1.48 | 0.79–2.77 | 6 | 23 | 36 | 231 | 0.83 | 0.32–2.14 |
| ER+ | 26 | 64 | 120 | 637 | 1.74 | 1.11–2.74 | 6 | 64 | 36 | 637 | 1.44 | 0.56–3.67) |
| Progesterone Receptor Statuse | ||||||||||||
| PR− | 17 | 29 | 87 | 314 | 1.93 | 1.10–3.37 | 7 | 29 | 46 | 314 | 1.25 | 0.54–2.93 |
| PR+ | 23 | 58 | 100 | 554 | 1.64 | 1.01–2.85 | 5 | 58 | 26 | 554 | 1.35 | 0.48–3.81 |
Adjusted for menopausal status, race (white, other), obesity (BMI <30, BMI ≥30), and history of myocardial infarction
8 missing responses for any lifetime physical activity
40 missing responses for average daily caloric intake
Among postmenopausal women
Not all breast cancer cases had data for receptor status (n = 955 with receptor status)
When stratified by body size, we found a higher risk of all-cause mortality in association with diabetes among those with a BMI of 30 or greater (HR = 2.49; 95 % CI = 1.58–3.93) than that seen in those with a BMI of less than 30 (HR = 1.20; 95 % CI = 0.73–1.99). Additionally, among women who were ever regular drinkers, although the association was not statistically significant, we found a stronger association for diabetes on breast cancer-specific mortality (HR = 2.14; 95 % CI = 0.90–4.69), whereas there was no increased risk of breast cancer death in association with diabetes among never drinkers.
Disease duration and medication use
Among diabetics, there was no consistent association between age at diabetes diagnosis, years since diabetes diagnosis, or diabetes medication use on the risk of developing breast cancer or mortality among cases (Table 4). After adjustment for duration of diabetes, modest non-significant increases in risk of developing breast cancer and all-cause mortality and diabetes were observed for women who were diagnosed with diabetes age after the age of 55 (OR = 1.71; 95 % CI = 0.71–4.11), as well as non-significantly increased all-cause mortality among those who took secretogogues (HR = 1.90; 95 % CI = 0.54–6.71). We also observed non-significant decreases in the association between diabetes and breast cancer risk and mortality among those who took metformin as treatment for their diabetes. We did not observe any associations for duration of diabetes, receiving treatment, or other treatment types.
Table 4.
Odds ratios (95 % CI) and hazard ratios (95 % CI) for the associations of select diabetes characteristics and breast cancer risk and overall survival among diabetic women, Long Island Breast Cancer Study Project, 1996–1997
| Characteristics | Controls | Cases | Deaths | ORa | 95 % CI | Overall death HRb | 95 % CI |
|---|---|---|---|---|---|---|---|
| Age at diabetes diagnosisc | |||||||
| <55 years | 50 | 45 | 15 | 1.00 | 1.00 | ||
| ≥55 years | 38 | 56 | 31 | 1.71 | 0.71–4.11 | 1.74 | 0.67–4.54 |
| Duration of diabetesd | |||||||
| <7 years | 45 | 42 | 19 | 1.00 | 1.00 | ||
| ≥7 years | 43 | 59 | 27 | 1.70 | 0.81–3.57 | 1.31 | 0.61–2.83 |
| Received treatment | |||||||
| No | 23 | 31 | 8 | 1.00 | 1.00 | ||
| Yes | 65 | 70 | 29 | 0.67 | 0.34–1.34 | 1.43 | 0.62–3.28 |
| Types of treatmente | |||||||
| No insulin | 49 | 50 | 21 | 1.00 | 1.00 | ||
| Took insulin | 16 | 20 | 8 | 1.15 | 0.40–3.40 | 0.91 | 0.29–2.82 |
| No metformin | 47 | 57 | 25 | 1.00 | 1.00 | ||
| Took metformin | 18 | 13 | 4 | 0.68 | 0.28–1.66 | 0.53 | 0.16–1.82 |
| No secretogogue | 14 | 15 | 4 | 1.00 | 1.00 | ||
| Took secretogoguesf | 51 | 55 | 25 | 1.29 | 0.42–3.99 | 1.90 | 0.54–6.71 |
Adjusted for menopausal status, obesity, and race
Adjusted for menopausal status, obesity, race, and history of BMI
Additionally adjusted for duration of diabetes
Additionally adjusted for age at diabetes diagnosis
Any treatment versus no treatment; mutually adjusted for other treatments
Majority of secretogogues were sulfonylureas
Discussion
This large population-based study suggests a moderate and independently increased risk of all-cause mortality among women with a breast cancer diagnosis. After adjustment for menopausal status, race, obesity, and history of MI, breast cancer patients with diabetes had more than a 60 % increased risk of all-cause mortality than those without diabetes; this association was more than twofold for women with a BMI of 30 or greater at diagnosis. We also found that self-reported diabetes was associated with risk of developing breast cancer among older women and those with a lower BMI as well as those who had gained less weight during adulthood. In light of the fact that the prevalence of T2D is rapidly increasing, the results of our study have strong clinical implications for breast cancer prevention and improving survival.
Our findings of an 65 % increased risk of death due to any cause confirm those found in population-based studies that also show increased all-cause mortality after breast cancer diagnosis, finding excess deaths due to diabetes ranging from 35 to 76 % [11, 12, 26, 27]. Although all-cause mortality was significantly increased with diabetes, the association with mortality due to breast cancer specifically is less clear in our study. Some studies have found increased associations with breast cancer death [12, 28, 29]; however, other studies have not found this association [30]. It has been suggested that women with diabetes are less likely to receive mammography screening [31] and tend to present with more advanced disease at diagnosis [27]. However, in our study, we saw no difference in mammography use or tumor stage between women with and without diabetes. It is unclear why the association with all-cause mortality observed for diabetes is not seen for breast cancer-specific death. Some of the association could be due in part to medication use reducing insulin and other circulating hormone levels associated with breast cancer, as nearly three-quarters of diabetics in our study regularly took medication for their condition. It is also possible that those with diabetes are more likely to die of other diseases that share the same risk factors as diabetes including renal disease, liver disease, and infections [32].
The risk of breast cancer in relation to T2D is thought to be due to increased circulating levels of insulin that are a direct result of insulin resistance. Biologic mechanisms of how diabetes can lead to breast cancer and affect prognosis include direct effects of high insulin levels, which have been shown to promote tumor proliferation, and insulin receptors are often overexpressed in breast cancer cells [33]. Insulin resistance can also indirectly affect breast cancer outcomes through increased sex steroid availability through decreases in sex hormone-binding globulin, increased IGF-I production, and disruption of adipokines [6]. In addition, T2D has been associated with chronic, low-grade inflammation. It has been shown that inflammatory molecules produced by adipose tissue as well as macrophages may lead to insulin resistance [34, 35].
As a result of the different mechanisms involved, medications taken for T2D may affect breast cancer depending on its mode of action, and there has been a recent effort toward studying how these medications affect cancer risk. Several studies have shown that medications that increase insulin levels, including use of insulin or insulin secretogogues, are associated with increased risk of cancer [16, 17, 36], while it has been suggested that treatment with insulin sensitizers, including metformin, may reduce risk of developing breast cancer and recurrence by killing the stem cells that are thought to be responsible for the spread of breast cancer [37]. We are not aware of any studies that have looked at medication use and survival after a breast cancer diagnosis; however, there is evidence that sensitizer medications, specifically metformin, may reduce cancer mortality [38]. While we were not able to adequately assess diabetes treatment on survival due to sparse data, there was a suggestion that metformin may reduce mortality and use of secretogogues increased mortality, although these associations may be biased due to unmeasured factors associated with medication compliance including regimen complexity, emotional factors, and medication cost [39]. Our results show no association with breast cancer development for any type of treatment, which are similar to those of another study, which looked at type of diabetes treatment and risk of developing breast cancer in women over the age of 65 [40], and the Nurses’ Health Study that found no increased risk of breast cancer for diabetes among those taking diabetes medication [22]. A second study of Hispanic women found an increased risk of developing breast cancer with use of insulin using a control group consisting of women who had received a diagnostic mammogram due to either inconclusive or abnormal results [19]. A recent meta-analysis on metformin use also found no association with breast cancer [41]. Further research is needed to adequately assess the impact of types of diabetes treatments, including sulfonylureas for which there are no supporting data, on breast cancer outcomes.
Increased body size is an established risk factor for developing breast cancer in postmenopausal women and affects survival after breast cancer diagnosis [4, 42]. Increased body size is also a well-known risk factor for diabetes. Our estimates of association between diabetes and risk of developing breast cancer did not change with adjustment for obesity. However, while we found no increased risk of developing breast cancer overall in relation to diabetes, when we stratified this association by body size, we found an increased risk of developing breast cancer among women who with a BMI less than 30 (OR = 1.52, 95 % CI = 1.03–2.25); an association that was even stronger when restricted to those with a BMI<25 (OR = 2.13, 95 % CI = 1.11–4.10). The reasons for these observations are unclear. Perhaps women with increased body size are at increased risk of postmenopausal breast cancer due to factors associated with their obesity other than diabetes, such as elevated estrogen levels [43] and increased adipocytokines [44] that have been shown to increase breast cancer cell proliferation and have involvement with angiogenesis. It is possible that women with lower BMI have fewer risk factors in general, and therefore their diabetes and hyperinsulinemia would have more of an impact on risk of developing breast cancer. Thus, it is possible that as is the case for exogenous hormone use, where effects are only evident among women without an increased body size [45, 46], the risk of breast cancer in association with diabetes is evident among women without an increased body size. Some researchers have attributed the interaction between body size and exogenous hormone use as a threshold effect; namely, the effects of hormone use are not evident among women with increased body size who are already estrogen-swamped [47]. Clearly, further research is warranted on modulating effects of body size on the association between diabetes and the risk of developing breast cancer.
We examined the association between diabetes and cancer while controlling for risk factors that are common to both, including age, obesity, physical activity, dietary factors, and alcohol consumption. Specifically, we assessed the association of diabetes and breast cancer with dietary factors often associated with diabetes such as high carbohydrate and high calorie intake. Our results do not suggest a difference in association when stratified by these factors. This is not surprising as the associations between these factors and breast cancer have been mixed [48–50]. However, low physical activity and obesity are strongly linked to both T2D and breast cancer. Because these factors are interrelated, it makes it difficult to identify the contribution of each on the relation of diabetes on breast cancer outcomes. Therefore, we cannot rule out residual confounding as an explanation for our findings.
We found more than a threefold increase in risk of developing breast cancer associated with diabetes among those of non-white race, although the number of non-white women in the LIBCSP is low and so cannot rule out chance for these findings. Because of the high prevalence of diabetes in African American communities, it is widely thought that race and ethnicity are major contributors to diabetes risk. More recent research, however, has shown that socioeconomic factors have a stronger association with prevalence of diabetes than race or ethnicity. Two recent studies report that after considering socioeconomic status, African Americans [51, 52] and Hispanics [51] have similar risks of diabetes as those found in Caucasians. In our study, however, after considering menopausal status, obesity, other comorbidities, income, and education, the association between diabetes and risk of developing breast cancer remained for non-whites. This may imply that diabetes has a differential effect on risk of developing breast cancer according to race or may simply imply that there is an additional unmeasured factor in our study that is driving the relationship. One reason for this association may be due to waist circumference (WC) that was not assessed in the LIBCSP. WC, a measure upper body obesity, correlates strongly with hyperinsulinemia [53] and is greater in African–American women than in white women with similar BMI [54]. Further research into this association including studies on environmental, behavioral, and genetic factors is needed.
There are a few limitations of this study that warrant mention. Although diabetes in our study was self-reported and did not distinguish between type 1 and type 2 diabetes, the majority of women who reported taking diabetes medications listed medications that are used to treat type 2 diabetes (85 %). Additionally, it is estimated that only 2.7 % of the population ages 20–44 has either undiagnosed or diagnosed diabetes [55]. Of the women in our study who reported having diabetes, 18 (7.8 %) women had a diabetes diagnosis before the age of 30, only 4 of whom were diagnosed before the age of 20. We excluded from analyses all women who reported having been diagnosed with diabetes before age 30 to increase the probability that the diabetes under investigation was adult-onset. However, we were not able to assess laboratory measurements that would confirm a diabetes diagnosis, nor were we able to adequately assess certain aspects of diabetes in relation to risk of developing breast cancer, such as types and duration of therapies due to the low numbers of participants with diabetes reporting use of specific types of therapies.
Our findings show that diabetes increases breast cancer risk and mortality in older women, regardless of hormone receptor status of the tumor. This has strong clinical implications as the prevalence of diabetes in the United States continues to increase, which could result in a large number of women who could be at risk of excess death after a breast cancer diagnosis. Identification of factors that affect breast cancer risk and survival could help health care providers better counsel to their patients by offering screening for diabetes as well as developing interventions aimed at preventing T2D and better maintenance of diabetes, including coordination of diabetes and breast cancer treatments.
Acknowledgments
This work supported by Susan G. Komen for the Cure Grant no. #KG081373; Marilyn Gentry Fellowship Program in Nutrition and Cancer; National Cancer Institute and the National Institutes of Environmental Health and Sciences Grant nos. UO1CA/ES66572 and P30ES10126.
Footnotes
Conflict of interest The authors declare they have no conflict of interest.
Contributor Information
Rebecca J. Cleveland, Email: becki@unc.edu, Department of Medicine, University of North Carolina, CB# 7280, Chapel Hill, NC 27599-7280, USA
Kari E. North, Department of Epidemiology, University of North Carolina, Chapel Hill, NC, USA
June Stevens, Department of Nutrition, University of North Carolina, Chapel Hill, NC, USA.
Susan L. Teitelbaum, Department of Preventive Medicine, Mt. Sinai School of Medicine, New York, NY, USA
Alfred I. Neugut, Departments of Medicine and Epidemiology, Columbia University, New York, NY, USA
Marilie D. Gammon, Department of Epidemiology, University of North Carolina, Chapel Hill, NC, USA
References
- 1.Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care. 2004;27:1047–1053. doi: 10.2337/diacare.27.5.1047. [DOI] [PubMed] [Google Scholar]
- 2.Wild S, Roglic G, Green A, Sicree R, King H. Centers for Disease Control and Prevention. National diabetes fact sheet: national estimates and general information on diabetes and pre-diabetes in the United States, 2011. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; Atlanta: 2011. [Google Scholar]
- 3.Carmichael AR. Obesity as a risk factor for development and poor prognosis of breast cancer. BJOG. 2006;113:1160–1166. doi: 10.1111/j.1471-0528.2006.01021.x. [DOI] [PubMed] [Google Scholar]
- 4.Cleveland RJ, Eng SM, Abrahamson PE, et al. Weight gain prior to diagnosis and survival from breast cancer. Cancer Epidemiol Biomarkers Prev. 2007;16:1803–1811. doi: 10.1158/1055-9965.EPI-06-0889. [DOI] [PubMed] [Google Scholar]
- 5.Loi S, Milne RL, Friedlander ML, et al. Obesity and outcomes in premenopausal and postmenopausal breast cancer. Cancer Epidemiol Biomarkers Prev. 2005;14:1686–1691. doi: 10.1158/1055-9965.EPI-05-0042. [DOI] [PubMed] [Google Scholar]
- 6.Vona-Davis L, Howard-McNatt M, Rose DP. Adiposity, type 2 diabetes and the metabolic syndrome in breast cancer. Obes Rev. 2007;8:395–408. doi: 10.1111/j.1467-789X.2007.00396.x. [DOI] [PubMed] [Google Scholar]
- 7.Kaaks R. Nutrition, hormones, and breast cancer: is insulin the missing link? Cancer Causes Control. 1996;7:605–625. doi: 10.1007/BF00051703. [DOI] [PubMed] [Google Scholar]
- 8.Lawlor DA, Smith GD, Ebrahim S. Hyperinsulinaemia and increased risk of breast cancer: findings from the British Women’s Heart and Health Study. Cancer Causes Control. 2004;15:267–275. doi: 10.1023/B:CACO.0000024225.14618.a8. [DOI] [PubMed] [Google Scholar]
- 9.Larsson SC, Mantzoros CS, Wolk A. Diabetes mellitus and risk of breast cancer: a meta-analysis. Int J Cancer. 2007;121:856–862. doi: 10.1002/ijc.22717. [DOI] [PubMed] [Google Scholar]
- 10.van de Poll-Franse LV, Houterman S, Janssen-Heijnen ML, Dercksen MW, Coebergh JW, Haak HR. Less aggressive treatment and worse overall survival in cancer patients with diabetes: a large population based analysis. International journal of cancer Journal International du Cancer. 2007;120:1986–1992. doi: 10.1002/ijc.22532. [DOI] [PubMed] [Google Scholar]
- 11.Srokowski TP, Fang S, Hortobagyi GN, Giordano SH. Impact of diabetes mellitus on complications and outcomes of adjuvant chemotherapy in older patients with breast cancer. Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology. 2009;27:2170–2176. doi: 10.1200/JCO.2008.17.5935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lipscombe LL, Goodwin PJ, Zinman B, McLaughlin JR, Hux JE. The impact of diabetes on survival following breast cancer. Breast Cancer Res Treat. 2008;109:389–395. doi: 10.1007/s10549-007-9654-0. [DOI] [PubMed] [Google Scholar]
- 13.Fleming ST, Rastogi A, Dmitrienko A, Johnson KD. A comprehensive prognostic index to predict survival based on multiple comorbidities: a focus on breast cancer. Med Care. 1999;37:601–614. doi: 10.1097/00005650-199906000-00009. [DOI] [PubMed] [Google Scholar]
- 14.Goodwin PJ, Ennis M, Pritchard KI, et al. Fasting insulin and outcome in early-stage breast cancer: results of a prospective cohort study. J Clin Oncol. 2002;20:42–51. doi: 10.1200/JCO.2002.20.1.42. [DOI] [PubMed] [Google Scholar]
- 15.Bodmer M, Meier C, Krahenbuhl S, Jick SS, Meier CR. Long-term metformin use is associated with decreased risk of breast cancer. Diabetes Care. 2010;33:1304–1308. doi: 10.2337/dc09-1791. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bowker SL, Majumdar SR, Veugelers P, Johnson JA. Increased cancer-related mortality for patients with type 2 diabetes who use sulfonylureas or insulin. Diabetes Care. 2006;29:254–258. doi: 10.2337/diacare.29.02.06.dc05-1558. [DOI] [PubMed] [Google Scholar]
- 17.Monami M, Lamanna C, Balzi D, Marchionni N, Mannucci E. Sulphonylureas and cancer: a case-control study. Acta Diabetol. 2009;46:279–284. doi: 10.1007/s00592-008-0083-2. [DOI] [PubMed] [Google Scholar]
- 18.Lipscombe LL, Goodwin PJ, Zinman B, McLaughlin JR, Hux JE. Diabetes mellitus and breast cancer: a retrospective population-based cohort study. Breast Cancer Res Treat. 2006;98:349–356. doi: 10.1007/s10549-006-9172-5. [DOI] [PubMed] [Google Scholar]
- 19.Sanderson M, Peltz G, Perez A, et al. Diabetes, physical activity and breast cancer among Hispanic women. Cancer Epidemiol. 2010;34:556–561. doi: 10.1016/j.canep.2010.06.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gammon MD, Neugut AI, Santella RM, et al. The Long Island Breast Cancer Study Project: description of a multi-institutional collaboration to identify environmental risk factors for breast cancer. Breast Cancer Res Treat. 2002;74:235–254. doi: 10.1023/a:1016387020854. [DOI] [PubMed] [Google Scholar]
- 21.Waksberg J. Sampling methods for random digit dialing. J Am Statistic Assoc. 1978;73:40–46. [Google Scholar]
- 22.Michels KB, Solomon CG, Hu FB, et al. Type 2 diabetes and subsequent incidence of breast cancer in the Nurses’ Health Study. Diabetes Care. 2003;26:1752–1758. doi: 10.2337/diacare.26.6.1752. [DOI] [PubMed] [Google Scholar]
- 23.Sellers TA, Jensen LE, Vierkant RA, et al. Association of diabetes with mammographic breast density and breast cancer in the Minnesota breast cancer family study. Cancer causes & Control: CCC. 2007;18:505–515. doi: 10.1007/s10552-007-0128-9. [DOI] [PubMed] [Google Scholar]
- 24.Hosmer DW, Lemeshow S. Applied logistic regression. Wiley; New York: 2000. [Google Scholar]
- 25.Allison P. Survival analysis using SAS: a practical guide. SAS Publishing; Cary: 1995. [Google Scholar]
- 26.Yancik R, Wesley MN, Ries LA, Havlik RJ, Edwards BK, Yates JW. Effect of age and comorbidity in postmenopausal breast cancer patients aged 55 years and older. JAMA. 2001;285:885–892. doi: 10.1001/jama.285.7.885. [DOI] [PubMed] [Google Scholar]
- 27.Schrauder MG, Fasching PA, Haberle L, et al. Diabetes and prognosis in a breast cancer cohort. J Cancer Res Clin Oncol. 2011;137:975–983. doi: 10.1007/s00432-010-0960-2. [DOI] [PubMed] [Google Scholar]
- 28.Coughlin SS, Calle EE, Teras LR, Petrelli J, Thun MJ. Diabetes mellitus as a predictor of cancer mortality in a large cohort of US adults. Am J Epidemiol. 2004;159:1160–1167. doi: 10.1093/aje/kwh161. [DOI] [PubMed] [Google Scholar]
- 29.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. doi: 10.2337/diacare.26.4.1047. [DOI] [PubMed] [Google Scholar]
- 30.Du W, Simon MS. Racial disparities in treatment and survival of women with stage I–III breast cancer at a large academic medical center in metropolitan Detroit. Breast Cancer Res Treat. 2005;91:243–248. doi: 10.1007/s10549-005-0324-9. [DOI] [PubMed] [Google Scholar]
- 31.Lipscombe LL, Hux JE, Booth GL. Reduced screening mammography among women with diabetes. Arch Intern Med. 2005;165:2090–2095. doi: 10.1001/archinte.165.18.2090. [DOI] [PubMed] [Google Scholar]
- 32.Seshasai SR, Kaptoge S, Thompson A, et al. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med. 2011;364:829–841. doi: 10.1056/NEJMoa1008862. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Papa V, Belfiore A. Insulin receptors in breast cancer: biological and clinical role. J Endocrinol Invest. 1996;19:324–333. doi: 10.1007/BF03347871. [DOI] [PubMed] [Google Scholar]
- 34.Hanley AJ, Festa A, D’Agostino RB, Jr, et al. Metabolic and inflammation variable clusters and prediction of type 2 diabetes: factor analysis using directly measured insulin sensitivity. Diabetes. 2004;53:1773–1781. doi: 10.2337/diabetes.53.7.1773. [DOI] [PubMed] [Google Scholar]
- 35.Xu H, Barnes GT, Yang Q, et al. Chronic inflammation in fat plays a crucial role in the development of obesity-related insulin resistance. J Clin Invest. 2003;112:1821–1830. doi: 10.1172/JCI19451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Wilson C. Diabetes: long-term use of insulin glargine might increase the risk of breast cancer. Nat Rev Endocrinol. 2011 doi: 10.1038/nrendo.2011.112. [DOI] [PubMed] [Google Scholar]
- 37.Hirsch HA, Iliopoulos D, Tsichlis PN, Struhl K. Metformin selectively targets cancer stem cells, and acts together with chemotherapy to block tumor growth and prolong remission. Cancer Res. 2009;69:7507–7511. doi: 10.1158/0008-5472.CAN-09-2994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Landman GW, Kleefstra N, van Hateren KJ, Groenier KH, Gans RO, Bilo HJ. Metformin associated with lower cancer mortality in type 2 diabetes: ZODIAC-16. Diabetes Care. 2010;33:322–326. doi: 10.2337/dc09-1380. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Odegard PS, Capoccia K. Medication taking and diabetes: a systematic review of the literature. Diabetes Educ. 2007;33:1014–1029. doi: 10.1177/0145721707308407. discussion 30-1. [DOI] [PubMed] [Google Scholar]
- 40.Lipscombe LL, Goodwin PJ, Zinman B, McLaughlin JR, Hux JE. Increased prevalence of prior breast cancer in women with newly diagnosed diabetes. Breast Cancer Res Treat. 2006;98:303–309. doi: 10.1007/s10549-006-9166-3. [DOI] [PubMed] [Google Scholar]
- 41.Decensi A, Puntoni M, Goodwin P, et al. Metformin and cancer risk in diabetic patients: a systematic review and meta-analysis. Cancer Prev Res (Phila) 2010;3:1451–1461. doi: 10.1158/1940-6207.CAPR-10-0157. [DOI] [PubMed] [Google Scholar]
- 42.Renehan AG, Tyson M, Egger M, Heller RF, Zwahlen M. Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet. 2008;371:569–578. doi: 10.1016/S0140-6736(08)60269-X. [DOI] [PubMed] [Google Scholar]
- 43.Key T, Appleby P, Barnes I, Reeves G. Endogenous sex hormones and breast cancer in postmenopausal women: reanalysis of nine prospective studies. J Natl Cancer Inst. 2002;94:606–616. doi: 10.1093/jnci/94.8.606. [DOI] [PubMed] [Google Scholar]
- 44.Simard J, Gingras S. Crucial role of cytokines in sex steroid formation in normal and tumoral tissues. Mol Cell Endocrinol. 2001;171:25–40. doi: 10.1016/s0303-7207(00)00387-7. [DOI] [PubMed] [Google Scholar]
- 45.Eng SM, Gammon MD, Terry MB, et al. Body size changes in relation to postmenopausal breast cancer among women on Long Island. New York. American Journal of Epidemiology. 2005;162:229–237. doi: 10.1093/aje/kwi195. [DOI] [PubMed] [Google Scholar]
- 46.Huang Z, Hankinson SE, Colditz GA, et al. Dual effects of weight and weight gain on breast cancer risk. JAMA. 1997;278:1407–1411. [PubMed] [Google Scholar]
- 47.Jurgens RW, Jr, Downey LJ, Abernethy WD, Cutler NR, Conrad J. A comparison of circulating hormone levels in postmenopausal women receiving hormone replacement therapy. Am J Obstet Gynecol. 1992;167:459–460. doi: 10.1016/s0002-9378(11)91429-x. [DOI] [PubMed] [Google Scholar]
- 48.Barclay AW, Petocz P, McMillan-Price J, et al. Glycemic index, glycemic load, and chronic disease risk–a meta-analysis of observational studies. Am J Clin Nutr. 2008;87:627–637. doi: 10.1093/ajcn/87.3.627. [DOI] [PubMed] [Google Scholar]
- 49.George SM, Mayne ST, Leitzmann MF, et al. Dietary glycemic index, glycemic load, and risk of cancer: a prospective cohort study. Am J Epidemiol. 2009;169:462–472. doi: 10.1093/aje/kwn347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Kabat GC, Shikany JM, Beresford SA, et al. Dietary carbohydrate, glycemic index, and glycemic load in relation to colorectal cancer risk in the Women’s Health Initiative. Cancer Causes Control. 2008;19:1291–1298. doi: 10.1007/s10552-008-9200-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Link CL, McKinlay JB. Disparities in the prevalence of diabetes: is it race/ethnicity or socioeconomic status? Results from the Boston Area Community Health (BACH) survey. Ethn Dis. 2009;19:288–292. [PMC free article] [PubMed] [Google Scholar]
- 52.Signorello LB, Schlundt DG, Cohen SS, et al. Comparing diabetes prevalence between African Americans and Whites of similar socioeconomic status. Am J Public Health. 2007;97:2260–2267. doi: 10.2105/AJPH.2006.094482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Okosun IS, Liao Y, Rotimi CN, Prewitt TE, Cooper RS. Abdominal adiposity and clustering of multiple metabolic syndrome in White. Black and Hispanic americans. Ann Epidemiol. 2000;10:263–270. doi: 10.1016/s1047-2797(00)00045-4. [DOI] [PubMed] [Google Scholar]
- 54.Berman DM, Rodrigues LM, Nicklas BJ, Ryan AS, Dennis KE, Goldberg AP. Racial disparities in metabolism, central obesity, and sex hormone-binding globulin in postmenopausal women. J Clin Endocrinol Metab. 2001;86:97–103. doi: 10.1210/jcem.86.1.7147. [DOI] [PubMed] [Google Scholar]
- 55.Berman DM, Rodrigues LM, Nicklas BJ, Ryan AS, Dennis KE, Goldberg AP. Early Release of Selected Estimates Based on Data From the 2010 National Health Interview Survey. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; Atlanta: 2011. [Google Scholar]