Cases and controls: This cross-sectional study was performed at Kuwait Cancer Control Center (KCCC), Shuwaikh, Kuwait and included 144 consecutive female patients with breast cancer during June 2007 - June 2009. The diagnosis of cancer was made on histological basis. Only newly diagnosed patients with breast cancer were recruited immediately before treatment commencement. Cases who were not histologically proven, male patients, patients with previous history of breast or other cancers, patients with history of diabetes mellitus on treatment were excluded. Also subjects who reported a recent (within the previous 1-6 months) weight gain or loss of 5 per cent or more of their current weight were excluded. Control subjects (n=77) comprising apparently healthy females were enrolled from the Blood Bank, Safat, Kuwait. In the control group exclusion of breast cancer was made on the basis of questionnaire and clinical examination. At the time of the study, 57 of 144 patients (39.58%) and 27 of 77 (35.06%) women of the control group were post-menopausal. None of the subjects was on hormone replacement therapy. All study subjects signed an informed voluntary consent form. The study protocol was approved by the local ethics committee. Women having menstrual cycles at the time of study were asked to report after fasting for 12-14 h on the next 2nd to 4th day of the cycle.
Blood pressure of each patient and control subject was measured in the sitting position after resting for 5 to 10 min using a mercury sphygmomanometer. Body weight (kg) was measured in light clothing without shoes. Height (cm) was measured as the distance from the top of the head to the bottom of the feet using a fixed stadiometer. Body mass index (BMI, kg/m2) was calculated. Waist circumference (cm) at the level of the umbilicus was also taken.
Blood (10 ml) was collected from patients and controls after fasting for 12-14 h and plasma was separated for analysis.
Adiponectin assay: Fasting plasma adiponectin was measured using a commercially available enzyme-linked immunoassay (ELISA) kit (Linco Research, Missouri, USA) with a sensitivity of 0.39 μg/ml. The inter- and intra-assay coefficients of variation on pooled plasma specimen with adiponectin concentration of 8.2 μg/ml were 4.7 and 6.8 per cent, respectively.
Leptin assay: Plasma leptin concentration was determined with the DSL-10-23100 ACTIIVE ELISA kit (Diagnostics Systems Laboratories, Texas, USA) with an assay sensitivity of 0.5 ng/ml. The inter- and intra-assay coefficients of variation on pooled plasma specimen with leptin concentration of 23.6 ng/ml were 4.1 and 5.3 per cent, respectively.
Other laboratory methods
: All routine biochemistry tests were performed immediately after sample collection. Fasting plasma glucose (FPG), total cholesterol (TC), triglycerides (TG) and high density lipoprotein cholesterol (HDL-cholesterol) were analyzed on an automated analyzer (Beckman DXC 800, Beckman Corporation, USA). The low density lipoprotein cholesterol (LDL-cholesterol) was calculated using the Friedewald formula9
Fasting serum insulin was determined by an ELISA (DSL-10-1600 ACTIVE, Diagnostics Systems Laboratories, Texas, USA). Insulin resistance was calculated using the homeostasis model assessment (HOMA-IR) using a calculator downloaded from http://www.dtu.ox.ac.uk/index.html?maindoc=/publications/
(Diabetes Trials Unit, Oxford, 2004). We used HOMA-IR > 2 as the cut-off point for determination of insulin resistance10
Quantitative determinations of luteinizing hormone (LH), follicle stimulating hormone (FSH), estradiol (E2), total testosterone and sex hormone-binding globulin (SHBG) were performed on an automated analyzer (Immulite 1000, Siemens Healthcare Diagnostics, Deerfield, USA).
Statistical analysis: Statistical Package for the Social Sciences (SPSS) version 16.0 for windows software (SPSS Inc., Chicago, IL) was used for statistical analysis. Several variables (BMI, adiponectin, leptin, insulin, and HOMA-IR) were not normally distributed. These variables were log-transformed when parametric tests were used. Spearman correlation coefficients (r) were used to describe the association between variables and comparison between two groups was performed with the unpaired Student's t test. Binary logistic regression analysis was used to estimate the odds ratios (ORs) and 95% confidence intervals (CIs) for the association between the variables and breast cancer. In the regression models age and estradiol (except in the association with estradiol when only age was used) were included as potentially confounding variables.