Insulin-like growth factors and breast cancer risk in Chinese women.
H Yu,F Jin,B Li,XO Shu,Q Dia,H Berkel,W Zheng
2001-01-01
Cancer Epidemiology Biomarkers & Prevention
Abstract:Insulin-like growth factor (IGF)-I has mitogenic and antiapoptotic effects on breast cancer cells. Highcirculating IGF-I was found to be associated with increased risk of breast cancer in several previous epidemiological studies, mostly conducted in the Caucasian populations. Little is known about the association between IGF and breast cancer in Asian women whose dietary habits differ considerably from their Caucasian counterparts. A population-based casecontrol study was conducted to assess the associations of IGFs and IGF binding protein-3, a major IGF binding protein in the circulation, with breast cancer risk in Chinese women. The study included 300 incident breast cancer patients diagnosed between August 1996 and March 1998 in Shanghai and 300 ageand menopausematched controls selected randomly from the general population. Plasma levels of IGF-I, IGF-II, and IGFBP-3 were measured using commercial ELISA kits (Diagnostic Systems Laboratories, Webster, TX). Conditional logistic regression analysis was performed to examine the association between IGF and breast cancer risk after adjusting for potential confounding factors. Breast cancer patients had higher plasma levels of IGF-I and IGFBP-3. A dose-response relationship was observed between breast cancer risk and the level of IGF-I or IGFBP-3. The adjusted odds ratios were 2.01 (95% confidence interval, 1.26–3.19) or 3.01 (95% confidence interval, 1.81–4.99), respectively, for women with the highest tertile of IGF-I or IGFBP-3 compared with those with the lowest tertile of these molecules. These associations were more evident in premenopausal women or women with high body mass index or high waist-to-hip ratio. No significant association was found for IGF-II. The study confirms that high circulating levels of IGF-I are associated with elevated risk of breast cancer. In contrast to the findings from several studies conducted in Caucasian women, we found that IGFBP-3 was positively associated with breast cancer risk in Chinese women. Introduction Emerging evidence suggests that IGF-I plays an important role in breast cancer (1, 2). IGF-I is a single-chain polypeptide with 70 amino acids and is involved in the regulation of cell proliferation, differentiation, and apoptosis (3, 4). Laboratory studies demonstrate that IGF-I has strong mitogenic and antiapoptotic effects on breast cancer cells (5–9) and interacts synergistically with estrogen to stimulate breast cancer growth (10–12). Animal studies indicate that IGF-I levels in the circulation are positively correlated with total energy intake and that IGF-I administration can abolish the inhibitory effect of energy restriction on tumor growth (13, 14). Several epidemiological investigations, including both case-control and cohort studies, found that high circulating levels of IGF-I were associated with increased risk of breast cancer (15–19). High IGF-I was also correlated with the mammagraphic density of breast tissue, a strong indicator of breast cancer risk (20). These findings lend additional support to the role of IGF-I in breast cancer development. Thus far, all of the epidemiological studies reported have been conducted in the Caucasian populations. Little is known about the relationship between IGF-I and breast cancer in Asian women. Given the possible influence of diet and nutrition on circulating IGF-I, and the substantial differences in dietary habits between Asian and Caucasian populations, we conducted a population-based case-control study to examine the associations of IGF-I, IGF-II, and IGFBP-3 with breast cancer risk in Chinese women. Materials and Methods Study Subjects. A large population-based case-control study was conducted in the urban area of Shanghai, China, from August 1996 through March 1998. Detailed descriptions of the study have been given elsewhere (21). Briefly, the study enrolled 1459 incident breast cancer patients aged between 25 and 64 years, and 1556 healthy controls with frequency-match on age. The cases represented 91% of the breast cancer patients newly diagnosed in the region during the study time; the controls represented 90% of eligible women randomly selected, as potential controls, from the general population in Shanghai. The Shanghai Resident Registry, which keeps records for all of the permanent residents in urban Shanghai, was used to select potential controls randomly from female residents, frequencyReceived 4/27/01; revised 4/10/02; accepted 4/22/02. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Supported by Public Health Service Grants CA83050 and CA64277 from the National Cancer Institute. 2 To whom requests for reprints should be addressed, at Department of Epidemiology and Public Health, Yale University School of Medicine, 60 College Street, P.O. Box 208034, New Haven, CT 06520-8034. Phone: (203) 785-5688; Fax: (203) 785-6980; E-mail: herbert.yu@yale.edu. 3 The abbreviations used are: IGF, insulin-like growth factor; CV, coefficient of variation; OR, odds ratio; CI, confidence interval; BMI, body mass index; WHR, waist-to-hip ratio. 705 Vol. 11, 705–712, August 2002 Cancer Epidemiology, Biomarkers & Prevention on July 6, 2017. © 2002 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from matched to cases by age (5-year interval). The number of controls in each age-specific stratum was determined in advance according to the age distributions of the incident breast cancer cases reported to the Shanghai Cancer Registry in recent years. Only women who lived at the address identified during the study period were considered to be eligible for the study. Each potential control was visited by an interviewer according to the home address provided in the resident registry to additionally determine study eligibility and her interest in participating in the study. All of the study participants completed an in-person interview with the use of a structured questionnaire, which elicited information on demographic features, menstrual and reproductive history, hormone use, medical history, physical activity, alcohol and tobacco use, dietary habits, and family history of cancer. Eighty-two percent of the cases (1, 193) and 84% of the controls (1, 310) provided blood samples. The specimens were collected in a 10-ml vacutainer tube with either EDTA or heparin anticoagulant. Plasma was separated immediately after collection and stored at 70°C until the study. To enhance the comparability between the cases and controls in the current study, we selected 300 cases whose blood samples were collected before any radiotherapy or chemotherapy, and 300 individually matched controls. For each index case, a control was selected randomly from the pool of 1310 healthy women who completed the interview and provided a blood sample. The cases and controls were individually matched on age ( 5 years), date of blood collection ( 30 days), and menopausal status (except for 14 pairs). The ages of the 300 patients ranged between 28 and 64 years, and the mean was 48.5 years (SD 8.3), which was comparable with the mean age of the controls, 48.4 years (SD 8.4; range 29–64). Postmenopausal women accounted for 43% of the women in the study. Of the 600 women selected, there was only 1 subject whose plasma sample was not available for the study. Specimen Measurement. Plasma concentrations of IGF-I, IGF-II, and IGFBP-3 were determined with the use of commercially available ELISA kits (DSL, Inc., Webster, TX). These methods were used in most of previous epidemiological studies and had fairly good reproducibility (15, 22, 23). The calibrators used in the assays ranged between 4.5–640 ng/ml for IGF-I, 50–2000 ng/ml for IGF-II, and 2.5–100 ng/ml for IGFBP-3. For IGFBP-3 measurement, plasma samples were diluted at 1:100 in an assay buffer. The intraand interassay precisions are 1.5–3.4 and 1.5–8.5% of CV, respectively, for IGF-I, 4.2–7.2 and 6.3–10.7% of CV for IGF-II, and 0.5–1.9 and 1.8–3.9% of CV for IGFBP-3 (24). Each assay has no cross-reaction with other members of the IGF family. To improve the precision of case-control comparisons, samples from a matched case-control pair were assayed in the same plate. Statistical Analysis. Because it was known that the distributions of the IGF variables are positively skewed (24), levels of IGF-I, IGF-II, and IGFBP-3 were compared between the cases and controls using the Wilcoxon signed rank test. The Spearman correlation coefficients were used to evaluate the correlations between the IGF variables or between IGF and other variables. To measure the associations of IGF variables with breast cancer risk, we calculated the adjusted ORs and their 95% CI using the conditional logistic regression model. In the logistic regression analysis, IGFs and IGFBP-3 were analyzed both as continuous and categorical variables. Values of IGFs and IGFBP-3 were log transformed when analyzed as continuous variables and were classified into three categories, low, medium, and high, based on the tertile distributions in the control group, when analyzed as categorical variables. Potential confounding factors adjusted in the logistic regression analysis included BMI (kg/m), age at menarche (year), age at first live birth (year), total energy intake (kcal), WHR, history of fibroadenoma (yes versus no), and family history of breast cancer (yes versus no). The association of IGF with breast cancer was additionally analyzed in women stratified by menopausal status. Stratified analyses were also performed in subgroups of women, stratified by BMI, WHR, and total energy intake, to assess the joint effect of these variables with IGFs and IGFBP-3. Levels of these variables were classified into two groups using median values in the control group as a cutoff. To assess the possible interactions between the