The Cox proportional hazards model is routinely used to analyze time-to-event data. To use this model requires the definition of a unique well-defined time scale. Most often, observation time is used as the time scale for both clinical and observational studies. Recently after a suggestion that it may be a more appropriate scale, chronological age has begun to appear as the time scale used in some reports. There appears to be no general consensus about which time scale is appropriate for any given analysis. It has been suggested that if the baseline hazard is exponential or if the age-at-entry is independent of covariates used in the model, then the two time scales provide similar results. In this report we provide an empirical examination of the results using the two different time scales using a large collection of data sets to examine the relationship between systolic blood pressure and coronary heart disease death (CHD death). We demonstrate, in this empirical example that the two time-scales can lead to differing results even when these two conditions appear to hold.

This article looks at the association of maternal blood pressure with the blood pressure of the offspring from birth to childhood. The Barker hypothesis states that maternal and "in utero" attributes during pregnancy affect a child's cardiovascular health throughout life. We present an analysis of a unique dataset that consists of three distinct developmental processes: maternal cardiovascular health during pregnancy; fetal development; and child's cardiovascular health from birth to 14 years. This study explored whether a mother's blood pressure reading in pregnancy predicts fetal development and determines if this in turn is related to the future cardiovascular health of the child. This article uses data that have been collected prospectively from a Jamaican cohort which involves the following three developmental processes: (1) maternal cardiovascular health during pregnancy which is the blood pressure and anthropometric measurements at seven time-points on the mother during pregnancy; (2) fetal development which consists of ultrasound measurements of the fetus taken at six time-points during pregnancy; and (3) child's cardiovascular health which consists of the child's blood pressure measurements at 24 time-points from birth to 14 years. The inter-relationship of these three processes was examined using linear mixed effects models. Our analyses indicated that attributes later in childhood development, such as child's weight, child's baseline systolic blood pressure (SBP), age and sex, predict the future cardiovascular health of children. The results also indicated that maternal attributes in pregnancy, such as mother's baseline SBP and SBP change, predicted significantly child's SBP over time.

We determined the prevalence of osteosarcopenic obesity (loss of bone and muscle coexistent with increased adiposity) in overweight/obese postmenopausal women and compared their functionality to obese-only women. Results showed that osteosarcopenic obese women were outperformed by obese-only women in handgrip strength and walking/balance abilities indicating their higher risk for mobility impairments.

The Cox proportional hazards model is widely used for analyzing associations between risk factors and occurrences of events. One of the essential requirements of defining Cox proportional hazards model is the choice of a unique and well-defined time scale. Two time scales are generally used in epidemiological studies: time-on-study and chronological age. The former is the most frequently used time scale, both in clinical studies and longitudinal observation studies. However, there is no general consensus on which time scale is the most appropriate for a given question or study. In this article, we address the question of robustness of the results using one time scale when the other is actually the correct one. We use three criteria to measure the performances of these models through simulations: magnitude of the bias of the regression coefficients, mean square errors, and the measure of overall predictive discrimination of the models. We conclude that the time-on-study models are more robust to misspecification of the underlying time scale.

Correspondence: Daniel Lee McGee Department of Mathematical Sciences, University of Puerto Rico, Mayaguez, Puerto Rico 00681-5000 Tel +1 787 263 3828 Fax + 1 787 265 5454 Email daniel.mcgee@upr.edu Background: This article presents cohort studies that use data from the National Health Information Survey from 1986 to 1994 and compares the effectiveness of Cox proportional hazards models that assume a linear relationship between body mass index (BMI) and the risk of prostate cancer with models that assume a J-shaped relationship. Methods and results: Our study found that for black males over 40 years of age, neither a linear nor a J-shaped relationship yielded a statistically significant model. With white males over 40 years, assuming a linear relationship did not yield a statistically significant model (P = 0.582). When we assume a J-shaped relationship, the optimal change point where the risk of prostate cancer death is minimized occurs when the BMI is 25.5. Among white males over 40 years with BMI , 25.5, an inverse relationship was found (P = 0.009). Among white males over 40 years with BMI . 25.5, a direct relationship was found (P = 0.017). Conclusion: With this data set, we found that for white males over 40 years, Cox proportional hazards models that assume a J-shaped relationship between BMI and prostate cancer death provide a much better fit than models assuming a linear relationship.

Given a prognostic model based on one population, one may ask: Can this model be used to accurately predict disease in a different population? When the underlying rate of disease differs in the new population, the model must be calibrated. van Houwelingen (2000) considered this calibration problem focusing on proportional hazards models. We extend the validation by calibration to the log-logistic accelerated failure time model. We use calibration of proportional hazards models and log-logistic accelerated failure time models to examine whether a survival model based on the Framingham Heart Study can be applied to diverse studies around the world.

The Cox proportional hazards model is widely used in time-to-event analysis. Two time scales are used in practice: time-on-study and chronological age. The former is the most frequently used time scale in clinical studies and longitudinal observation studies. However, there is no general consensus about which time scale is the best. It has been asserted that if the cumulative baseline hazard is exponential or if the age-at-entry is independent of the covariate, then the two models are equivalent. We show that neither of these conditions leads to equivalency. Variability in the age-at-entry of individuals in the study causes the models to differ significantly. This is shown both analytically and through a simulation study. Additionally, we show that the time-on-study model is more robust to changes in age-at-entry than the chronological age model.

OBJECTIVES: To investigate the effects of testosterone supplementation on bone, body composition, muscle, physical function, and safety in older men. DESIGN: Double‐blind, randomized, placebo‐controlled trial. SETTING: A major medical institution. PARTICIPANTS: One hundred thirty‐one men (mean age 77.1 ± 7.6) with low testosterone, history of fracture, or bone mineral density (BMD) T ‐score less than −2.0 and frailty. INTERVENTION: Participants received 5 mg/d of testosterone or placebo for 12 to 24 months; all received calcium (1500 mg/d diet and supplement) and cholecalciferol (1,000 IU/d). MEASUREMENTS: BMD of hip, lumbar spine, and mid‐radius; body composition; sex hormones, calcium‐regulating hormones; bone turnover markers; strength; physical performance; and safety parameters. RESULTS: Ninety‐nine men (75.6%) completed 12 months, and 62 (47.3%) completed end therapy (mean 23 months; range 16–24 months for 62 who completed therapy). Study adherence was 54%, with 40% of subjects maintaining 70% or greater adherence. Testosterone and bioavailable testosterone levels at 12 months were 583 ng/dL and 157 ng/dL, respectively, in the treatment group. BMD on testosterone increased 1.4% at the femoral neck and 3.2% at the lumbar spine ( P =.005) and decreased 1.3% at the mid‐radius ( P <.001). There was an increase in lean mass and a decrease in fat mass in the testosterone group but no differences in strength or physical performance. There were no differences in safety parameters. CONCLUSION: Older, frail men receiving testosterone replacement increased testosterone levels and had favorable changes in body composition, modest changes in axial BMD, and no substantial changes in physical function.

The Bayesian dynamic survival model (BDSM), a time‐varying coefficient survival model from the Bayesian prospective, was proposed in early 1990s but has not been widely used or discussed. In this paper, we describe the model structure of the BDSM and introduce two estimation approaches for BDSMs: the Markov Chain Monte Carlo (MCMC) approach and the linear Bayesian (LB) method. The MCMC approach estimates model parameters through sampling and is computationally intensive. With the newly developed geoadditive survival models and software BayesX, the BDSM is available for general applications. The LB approach is easier in terms of computations but it requires the prespecification of some unknown smoothing parameters. In a simulation study, we use the LB approach to show the effects of smoothing parameters on the performance of the BDSM and propose an ad hoc method for identifying appropriate values for those parameters. We also demonstrate the performance of the MCMC approach compared with the LB approach and a penalized partial likelihood method available in software R packages. A gastric cancer trial is utilized to illustrate the application of the BDSM. Copyright © 2009 John Wiley & Sons, Ltd.

In 1981, Dr. Jerome Sullivan proposed the hypothesis that the risk of coronary heart disease (CHD) increases in a positive fashion as body iron stores increase.Serum ferritin and other less precise measures of body iron stores have been used in those studies to test the hypothesis.Serum ferritin was not significantly related to risk of developing CHD in the vast majority of the observational cohort studies, case-control, or cross-sectional studies.In an underpowered clinical trial, those receiving phlebotomy to lower body stores of iron did not have a significantly lower risk of death from all causes (primary endpoint) or of death plus non-fatal heart attack or stroke compared to controls.The presence of the Cys282Tyr mutation, which accounts for most of the cases of hemochromatosis, was not found to be associated with CHD risk in two meta-analysis studies.At present, the vast majority of the epidemiological data does not support the hypothesis that body iron stores are directly related to the risk of developing CHD.

The Cox proportional hazards model is widely used in time-to-event analysis. Two time scales are used in practice: time-on-study and chronological age. The former is the most frequently used time scale both in clinical studies and longitudinal observation studies. There is no general consensus about which time scale is the best. It has been asserted that if the cumulative baseline hazard is exponential or if the age-at-entry is independent of the covariate, then the two models are equivalent. We show that neither of these conditions leads to equivalency. Variability in the age-at-entry of individuals in the study causes the models to differ significantly. This is shown both analytically and through a simulation study.

Based on the 40-year follow-up of the Framingham Heart Study (FHS), we used logistic regression models to demonstrate that different designs of an observational study may lead to different results about the association between BMI and all-cause mortality. We also used dynamic survival models to capture the time-varying relationships between BMI and mortality in FHS. The results consistently show that the association between BMI and mortality is dynamic, especially for men. Our analysis suggests that the dynamic property may explain part of the heterogeneity observed in the literature about the association of BMI and mortality.

PURPOSE AND METHODS: We evaluated whether hypertension control differs by ethnicity after accounting for patient characteristics, treatment, and adherence to treatment using the third National Health and Nutrition Examination Survey (US population estimate, 42,511,379). Outcome measures were prescribed treatment, treatment adherence, hypertension control (blood pressure [BP] < 140/90 mm Hg). Multivariate logistic regression was per-formed with non-Hispanic whites (NHW) as the comparison group.RESULTS: Non-Hispanic blacks (NHB) were more likely to report medication prescription (odds ratio [OR] 1.6, 95% confidence interval [CI] 1.1-2.5) and being advised to restrict salt (OR 1.5, CI: 1.2-2.0). Among those advised, NHB were more likely to report salt restriction (OR 1.5, CI: 1.1-2.1) and weight-loss attempts (OR 1.7, CI: 1.3-2.3). Among persons advised to follow exercise, alcohol restriction, smoking cessation, tension reduction, or diet modification, NHB (OR 2.2, CI: 1.6-3.0) and Mexican Americans (OR 2.0, CI: 1.1-3.9) were more likely to report adherence. The likelihood of uncontrolled hypertension was higher in NHB (OR 1.4, CI: 1.1-1.7) and Mexican Americans (OR 1.5, CI 1.1-2.0) despite medication adherence.CONCLUSIONS: Even after adjustment for treatment and adherence, substantial ethnic differences in hypertension control were found. Initiating treatment, while crucial, is not sufficient and future guidelines should emphasize aggressive treatment escalation to achieve hypertension control.

Studies on the association between physical activity and fatal prostate cancer have produced inconclusive results. The Puerto Rico Heart Health Program was a cohort study of a randomly selected sample of 9824 men age 35 to 79 years at baseline who were followed for mortality until 2002. Multiple examinations collected information on lifestyle, diet, body composition, exercise, urban-rural residence, and smoking habits. Physical activity status was measured using the Framingham Physical Activity Index, an assessment of occupational, leisure-time, and other physical activities measured as usual activity over the course of a 24-hour day. Physical activity was stratified into quartiles. Multivariate logistic regression analysis was used to assess the association of physical activity with prostate cancer mortality. Other covariates included age, education, urban-rural residence, smoking, and body mass index. Compared with the lowest level of physical activity (Q1), the risk of prostate cancer mortality was OR = 0.99 (95% CI = 0.64-1.55) for Q2, OR = 1.34 (95% CI = 0.88-2.05) for Q3, and OR = 1.19 (95% CI = 0.75-1.90) for Q4. Further analyses by age group, overweight status, or vigorous physical activity also did not show a significant association between physical activity and prostate cancer mortality. Physical activity did not predict prostate cancer mortality in this group of Puerto Rican men.

Metabolic syndrome is a combination of risk factors linked to type 2 diabetes and cardiovascular disease. Such clustering carries a greater risk for adverse clinical outcomes, consequently increasing the risk for all cause mortality. The National Cholesterol Education Program's Adult Treatment Panel III and the World Health Organization have adopted different criteria for defining metabolic syndrome, but there is agreement that lifestyle changes constitute the first line of therapy. PURPOSE: To examine the independent effect of physical activity on all cause mortality among Puerto Rican men with metabolic syndrome. METHODS: We used data from the Puerto Rico Heart Health Program, a longitudinal study of 9824 men aged 35-74 years. A physical examination at baseline collected data on body weight, height, smoking, age, education, urine albumin levels, and blood pressure. Blood samples were also collected to measure serum triglycerides, total blood cholesterol, and fasting blood glucose. Using these data and WHO clinical criteria, we identified 1705 participants with metabolic syndrome at baseline. Physical activity was assessed using the Framingham Physical Activity questionnaire, and participants were classified into quartiles based on metabolic equivalents and time spent in different activities. Multivariate logistic regression analysis was used to estimate the risk of mortality by physical activity categories after adjusting for age, education, smoking, and BMI. RESULTS: Among the 1705 men who met the WHO criteria for metabolic syndrome, 498 deaths were recorded and 1207 were alive after 12 years, with only 9 participants lost to follow up. Using physical inactivity (quartile 1) as the reference, participation in moderate amounts of physical activity (quartile 2) conferred significant protective benefit from all cause mortality (OR=0.58; 95% CI 0.43, 0.77). Moreover, those who participated in more physical activity (quartiles 3 and 4) had significantly lower risk (OR=0.49 and OR 0.43, respectively, P<.001) than men who were inactive. CONCLUSIONS: Our results support the hypothesis that physical activity protects against all cause mortality among Puerto Rican men with metabolic syndrome. Supported by Department of Defense Grant DAMD17-02-1-0252 and NIH Grant P20 CA96256-01A1, R03 CA103475-01.

Objective: To determine the relationships between C-reactive protein (CRP) levels and features of Type 1 diabetes. Research Design and Methods: Serum CRP was measured by nephelometry in a cross-sectional study of the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) cohort (n=983) and nondiabetic subjects (n=71). Results: CRP levels [geometric mean (95% CI)] were higher in diabetic than in control subjects, 1.6 (1.5-1.7) vs. 1.2 (1.1-1.5) mg/l, P=.019. CRP was higher in diabetic women (n=438) than in men (n=545) [2.0 (1.8-2.3) vs. 1.3 (1.2-1.5), P<.001]. Diabetic subjects formerly in the DCCT intensive treatment group had higher CRP levels than those who were randomized to the conventional treatment group [1.8 (1.6-1.9), n=479 vs. 1.5 (1.3-1.6), n=456, P=.010], attributable to greater BMI in the prior intensive group. In diabetes, CRP corTelated with HbA(1c) (r=0.13, P<.0001) and with insulin resistance traits: BMI (r=0.34, P<.0001), waist-to-hip ratio (WHR; males: r=0.35, P<.0001; females: r=0.22, P<.0001), diastolic blood pressure (r=0.07, P=.025), triglycerides (r=0.19, P<.0001), apoB (r=0.22, P<.0001), LDL particle concentration (r=0.26, P<.0001), and LDL particle size (r=-0.22, P<.0001). CRP was not associated with complications. Significant independent predictors of CRP in diabetes were gender, BMI, WHR, concurrent HbA(1c), and oral contraceptive pill use. Conclusions: CRP was elevated relative to nondiabetic subjects, and in diabetes was higher in females. Elevated CRP in Type 1 diabetes was associated with poor glycemic control, larger body habitus, and other factors that comprise the insulin resistance syndrome. Nevertheless, CRP levels were not associated with complications. Longitudinal studies are warranted. (c) 2008 Elsevier Inc. All rights reserved.

The Evans County Heart Study (ECHS), initiated in 1960, was one of the first major studies to document cardiovascular disease (CVD) risks for African Americans and Caucasians with elevated blood pressures. In the early 1970s, the Hypertension Detection and Follow-up Program (HDFP), with a site in Georgia (HDFP-GA), was one of the first major studies to demonstrate that treating hypertension with stepped care (SC), vs. referred care (RC), has better short-term outcomes. With this background, study objectives were to evaluate 30-year survival and cardiovascular outcomes of the HDFP-GA and to compare outcomes of these patients with 1,619 hypertensive individuals (30 to 69 years of age) from the ECHS. The HDFP-GA patients included 688 individuals (Black [n = 267]; White [n = 421]) randomized to RC (n = 341) and SC (n = 347). The ECHS was comprised of 733 Black and 886 White hypertensives. All-cause mortality and CVD mortality were assessed in the HDFP-GA and compared with those in the ECHS hypertensives. After 30 years of follow-up, 65.7% of the HDFP-GA cohort had died compared with a similar 65.8% of the ECHS hypertensives. However, CVD mortality rates, while similar for the SC and RC arms, were lower than in the HDFP-GA total study group than the hypertensive participants of ECHS (32.6% vs. 40.3%; P < .001). CVD survival rates for both SC and RC HDFP-GA arms were significantly better than population-based hypertensive individuals in the ECHS, with consistent benefits in all four race-gender groups. These results identify the importance of long-term follow-up of individuals in hypertension studies and trials that include CVD outcomes.