|
nif:isString
|
-
The study is secondary data analysis and was approved by IRB Office, University of Maryland, College Park, Lee Building, Room 2100, Zip 5125, College Park, Maryland 20742. IRB APPLICATION # 09–0527. Written informed consent was given by participants in the Health ABC study for their clinical records to be used. The Health ABC study is a longitudinal cohort study of 3,075 community-dwelling, well-functioning white and black men and women aged 70 to 79 years at the commencement of the study. This study was described in details elsewhere [32]. Briefly, participants were recruited for the study from a random sample of white residents receiving Medicare benefits and all age-eligible black residents of Pittsburgh, Pennsylvania, and Memphis, Tennessee. The subjects were considered eligible to participate in the study if they reported no difficulty walking a quarter of a mile, climbing 10 steps without resting, or performing basic activities of daily living, were free of life-threatening illness, and planned to remain in the geographic area for at least 3 years. Those who reported active treatment for cancer or participation in diet or exercise intervention were excluded from the study. Participants gave written informed consent, and protocols were approved by the Institutional Review Boards at the two study sites. The participants’ information was anonymized and de-identified prior to analysis. Enrollment in the cohort and baseline clinic visits, which included collection of serum and measures of body composition, weight-related health conditions, and physical function, were conducted between April 1997 and June 1998. Thereafter, clinical examinations to obtain follow-up body composition measures and clinical outcome data were conducted annually for six years, and then every other year through year 10. Semi-annual phone interviews were also conducted between clinic visits to ascertain outcome data, In the present analysis, data from baseline through year 10 of the Health ABC study were used. The sample size for this analysis was 2,919 (men = 1,419; women = 1,500) after excluding those with incomplete information for vital status (n = 101) and missing serum leptin (n = 55). One hundred thirteen (n = 113) participants were missing data for total percent body fat, 119 for abdominal visceral fat, 209 for abdominal subcutaneous fat, 203 for TNF-α, 38 for C-reactive protein (CRP), 65 for plasminogen activator inhibitor-1(PAI-1) and 162 for IL-6. The number of participants missing information for these variables did not differ by the outcome variable.
Blood samples were collected in the morning when participants underwent venipuncture at the baseline visit after an overnight fast, and serum samples were frozen at -70°C. Both IL-6 and TNF- α were measured in duplicate using an ultrasensitive enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, MN, USA). The limit of detection was 0.10 pg/ml for IL-6 and 0.18 pg/ml for TNF- α. The level of PAI-1 was measured in citrated plasma samples using a 2-site enzyme-linked immunosorbent assay. Serum levels of CRP were also measured in duplicate by enzyme-linked immunosorbent assay based on purified protein and polyclonal anti-CRP antibodies (Calbiochem, San Diego, CA, USA) with a coefficient of variation of 8.0%. Serum concentrations of leptin were measured in duplicate by means of radioimmunoassay (Linco Research Inc, St Charles, MO, USA). The assay range is 0.05 to 100 ng/ml leptin in serum. Intra-assay CVs ranged from 3.7% to 7.5%, and inter-assay CVs ranged from 3.2% to 8.9%.
Weight was measured in kilograms using a standard balance beam scale. Height was measured twice in centimeters using a Harpenden stadiometer (Holtain Ltd., Crosswell, U.K.) and the average of the two measurements was used. BMI [weight (kg)/height (m2)] was calculated. Dual energy x-ray absorptiometry (DXA) (Hologic QDR 4500A, software version 8.21, Hologic, Waltham, MA, USA) was used to assess total fat mass and total percentage of body fat was calculated.
A standardized questionnaire was administered at baseline to collect information on socio-demographic variables including age, gender, self-identified race, years of education, and lifestyle variables including smoking status and average alcohol consumption during the past year. Information on use of estrogen or female hormone pills (Hormonal Replacement Therapy, HRT) since menopause was also collected from women. Cigarette packs smoked per day were multiplied by the number of years of smoking to calculate the pack-years over a lifetime of cigarette smoking. Level of physical activity was ascertained by a standardized questionnaire specifically designed for the Health ABC study [33]. The frequency, duration, and intensity level of activities such as self-report of walking and exercise were recorded, and approximate values of metabolic equivalent unit (MET) were assigned to each activity category to estimate weekly energy expenditure in kcal/kg/week. Dietary calorie intake was measured during Year 2 of the Health ABC Study, at the first annual follow-up visit using a 108-item food frequency questionnaire (FFQ) that was designed to assess intake in the year prior. Nutrient and food group intakes were determined by Block Dietary Data Systems [34].
All deaths were adjudicated by a central committee. The outcomes included in this analysis were all-cause mortality, and CVD-related mortality events occurring between baseline through November 26, 2007. Every 6 months, participants either received an in-person examination or were contacted by telephone. Deaths were also ascertained through review of death certificates, hospital records, proxy interviews, and Social Security Death Index data. All deaths in Health ABC were adjudicated by a central committee. CVD-related death was defined as any death where the underlying cause was confirmed as atherosclerotic cardiovascular disease (definite fatal myocardial infarction, or definite or possible fatal coronary artery disease), cerebrovascular disease, atherosclerotic disease other than cardiovascular, and other cardiovascular disease.
Serum leptin values and body fat vary by gender; therefore, the analysis was performed separately by gender [15, 35]. Baseline characteristics of men and women were examined by quartiles of serum leptin. For continuous variables, least square means were computed with Dunnett’s test option, and for categorical variables, the chi-square test was used to compare the means of quartile 2 through 4 to those of quartile 1. The inflammatory markers and leptin was log transformed because they were not normally distributed. The correlations between explanatory variables (age, smoking status, exercise, total percent fat, leptin, CRP, IL-6, TNF- α, and PAI-1) were examined and multicollinearity was assessed by using variance inflation factors (VIF). The VIF remained below the suggested cut off value of four for all explanatory variables. Interactions between serum leptin with race, inflammatory markers (CRP, IL-6, TNF- α, and PAI-1), adiposity measures (total percent fat) and CVD diagnosis were tested; none were statistically significant based on alpha level of 0.05. Survival time was defined as the time from the date of the baseline visit until the date of death and/or date of last contact. Cox proportional hazard regression analysis was used to assess the risk of all-cause mortality and mortality from CVD for persons in quartiles 2 through 4 of serum leptin with those in quartile 1. The regression models were sequentially adjusted for potential confounders (age, gender, race, study site, education, smoking status, alcohol use, physical activity, HRT use, and number of hours fasted), total percent fat and markers of systemic inflammation (CRP, TNF- α, IL-6 and PAI-1). Final model included the main predictor variable (leptin) and covariates that were significantly associated with the outcome. Log transformed leptin was used in linear regression model for trend analysis. All the covariates met the assumptions for proportional hazard regression. Statistical significance was set at p ≤ 0.05, and analysis was performed using SAS software program (version 9.1; SAS Institute Inc., Cary, NC).
|
![[This page as RDF]](http://data.gesis.org/somesci/static/rdf-icon.gif){kind=icon}