Biomedical subjects
Stuart R Lipsitz
Publications and source records attributed to Stuart R Lipsitz.
Impact of nutrients on insulin-like growth factor-I, insulin-like growth factor binding protein-3 and their ratio in African American and white males.
OBJECTIVE: Higher levels of insulin-like growth factor-I (IGF-I) and lower levels of IGF binding protein-3 (IGFBP-3) have been associated with an increased risk of prostate cancer. Nutrition is known to partially regulate IGF levels and it is possible that nutritional factors mediate the impact of IGF levels on prostate cancer risk. DESIGN: A cross-sectional analysis of the impact of nutritional factors measured by a dietary questionnaire on plasma levels of IGF-I, IGFBP-3 and their molar ratio. Multiple linear regression analysis was used to test for effects of nutrients on IGF levels. SETTING: Prostate cancer screening at the Hollings Cancer Center in Charleston, South Carolina. SUBJECTS: Ninety-five African American and 138 white males aged 33-83 years attending the screening. RESULTS: In whites, intakes of total, saturated and monounsaturated fats were positively associated with an increase in the molar ratio, while there was no association in African Americans. In African Americans, we found that increasing intake of calcium and dairy servings was positively associated with IGF-I levels. Increased vegetable intake was positively associated with IGFBP-3 in African Americans, while there was no effect in whites. A higher percentage of alcohol in the total diet was significantly associated with a decrease in the molar ratio and an increase in IGFBP-3 in both groups. CONCLUSIONS: Our results confirm previous findings of nutritional determinants of IGF levels. Additionally, we found the impact of several nutrients on IGF levels to be different in whites and African Americans, which warrants further investigation.
Pseudo-likelihood methods for longitudinal binary data with non-ignorable missing responses and covariates.
In this paper we consider longitudinal studies in which the outcome to be measured over time is binary, and the covariates of interest are categorical. In longitudinal studies it is common for the outcomes and any time-varying covariates to be missing due to missed study visits, resulting in non-monotone patterns of missingness. Moreover, the reasons for missed visits may be related to the specific values of the response and/or covariates that should have been obtained, i.e. missingness is non-ignorable. With non-monotone non-ignorable missing response and covariate data, a full likelihood approach is quite complicated, and maximum likelihood estimation can be computationally prohibitive when there are many occasions of follow-up. Furthermore, the full likelihood must be correctly specified to obtain consistent parameter estimates. We propose a pseudo-likelihood method for jointly estimating the covariate effects on the marginal probabilities of the outcomes and the parameters of the missing data mechanism. The pseudo-likelihood requires specification of the marginal distributions of the missingness indicator, outcome, and possibly missing covariates at each occasions, but avoids making assumptions about the joint distribution of the data at two or more occasions. Thus, the proposed method can be considered semi-parametric. The proposed method is an extension of the pseudo-likelihood approach in Troxel et al. to handle binary responses and possibly missing time-varying covariates. The method is illustrated using data from the Six Cities study, a longitudinal study of the health effects of air pollution.
Geographical patterns of end-stage renal disease incidence and risk factors in rural and urban areas of South Carolina.
To assess the geographical patterns of end-stage renal disease (ESRD) incidence and to identify the risk factors on the regional differences, the authors conducted an ecological study on incidence of ESRD and related risk factors in the 46 counties of South Carolina (SC). Age and gender adjusted, race specific incidence rates for each county in SC were calculated for the 11,346 ESRD patients of all ages who registered in the United States Renal Data Systems Network 6 from 1990 to 1999. County level exposure measures on population physician density, hospitalization rates of diabetes and hypertension, per capita income, percent college degree, and percent below poverty were evaluated. There was a significant increase in mean incidence rates of ESRD from 1990 to 1999 in SC (p<0.0001). The incidence rates were consistently higher in rural than in urban counties. Population physician density (relative risk (RR) 0.49, 95% confidence interval (95%Cl, 0.41-0.58) and rural residence (adjusted RR 1.66, 95%Cl 1.59-1.74) were significantly associated with ESRD incidence. The strong relationship between ESRD and physician density suggests that access to adequate treatment of diabetes and hypertension is of paramount importance for ESRD prevention, and has important public policy implications.
Estimation in regression models for longitudinal binary data with outcome-dependent follow-up.
In many observational studies, individuals are measured repeatedly over time, although not necessarily at a set of pre-specified occasions. Instead, individuals may be measured at irregular intervals, with those having a history of poorer health outcomes being measured with somewhat greater frequency and regularity. In this paper, we consider likelihood-based estimation of the regression parameters in marginal models for longitudinal binary data when the follow-up times are not fixed by design, but can depend on previous outcomes. In particular, we consider assumptions regarding the follow-up time process that result in the likelihood function separating into two components: one for the follow-up time process, the other for the outcome measurement process. The practical implication of this separation is that the follow-up time process can be ignored when making likelihood-based inferences about the marginal regression model parameters. That is, maximum likelihood (ML) estimation of the regression parameters relating the probability of success at a given time to covariates does not require that a model for the distribution of follow-up times be specified. However, to obtain consistent parameter estimates, the multinomial distribution for the vector of repeated binary outcomes must be correctly specified. In general, ML estimation requires specification of all higher-order moments and the likelihood for a marginal model can be intractable except in cases where the number of repeated measurements is relatively small. To circumvent these difficulties, we propose a pseudolikelihood for estimation of the marginal model parameters. The pseudolikelihood uses a linear approximation for the conditional distribution of the response at any occasion, given the history of previous responses. The appeal of this approximation is that the conditional distributions are functions of the first two moments of the binary responses only. When the follow-up times depend only on the previous outcome, the pseudolikelihood requires correct specification of the conditional distribution of the current outcome given the outcome at the previous occasion only. Results from a simulation study and a study of asymptotic bias are presented. Finally, we illustrate the main results using data from a longitudinal observational study that explored the cardiotoxic effects of doxorubicin chemotherapy for the treatment of acute lymphoblastic leukemia in children.
Frequency of clinically unsuspected myocardial injury at a children's hospital.
BACKGROUND: Ill children are at risk but rarely screened for myocardial injury. The frequency of such injury in ill children is unknown. Elevated levels of plasma cardiac troponin I (cTnI) can detect subclinical myocardial injury. METHODS: We measured cTnI levels from 283 Children's Hospital, Boston patients (median age 2.10 years, range 0.13-22.4 years) seen in an outpatient or emergency clinic without clinically apparent cardiac disease. We took > or = 0.5 ng/mL as an indication of myocardial injury. We also measured plasma creatine kinase-MB, total creatine kinase, and myoglobin, and performed a chart review. RESULTS: Fifteen (7.8%) of the 193 acutely ill children and 4 (4.4%) of the 90 well children had an elevated cTnI level (P = .44). Within the acutely ill group, the children with elevated cTnI were younger and had lower mean hemoglobin and hematocrit levels. Cardiac troponin I levels correlated with creatine kinase-MB (r = 0.22; P < .001) but not with creatine kinase or myoglobin. The 4 children with cTnI > 0.89 ng/mL, who also had plasma cardiac troponin T measured, showed cardiac troponin T elevations that were consistent with unstable angina levels in adults. Four children had high-level cTnI elevations (> 2 ng/mL) consistent with acute myocardial infarction levels in adults. CONCLUSIONS: Elevated cTnI levels occur in children without clinically apparent cardiac disease and can be at adult unstable angina or acute myocardial infarction levels. Prospective studies to determine the clinical significance of these findings and their relationship to the development of cardiomyopathy are warranted.
Resource utilization in the treatment of critical limb ischemia: The effect of tissue loss, comorbidities, and graft-related events.
OBJECTIVE: Resource utilization (RU) in the care of patients with critical limb ischemia (CLI) is not well quantified. We present a cohort study to quantify in-hospital RU and analyze the role of tissue loss (TL), comorbidities, and vascular graft-related events (GREs) in patients undergoing peripheral bypass for CLI. METHODS: A retrospective analysis of 1404 patients enrolled in a multicenter clinical trial (PREVENT III) of vein bypass grafting for CLI was performed with analysis of RU during the 1-year follow-up period. Univariate and multivariable linear regressions were performed to determine RU predictors and outcomes. RESULTS: Compared with patients with rest pain, patients presenting with TL as the indication for bypass surgery had a longer index length of stay (mean, 9.8 vs 6.2 days), more rehospitalizations (mean, 1.6 vs 1.2), and a longer cumulative length of stay (mean, 27.7 vs 17.3 days; P < .0001 for all comparisons). Rehospitalizations over the ensuing year were for additional procedures (37.5%), wound infection (14.6%), graft failure (10.7%), and other cardiovascular (10%) and noncardiovascular (26%) reasons. Early GRE (stenosis > or =70%, thrombosis, revision, or major amputation within 30 days) occurred in 162 (11.5%) patients, resulting in a longer index length of stay (mean, 11.8 vs 8.6 days; P = .0002) and cumulative length of stay (mean, 25.9 vs 24.6 days; P = .0043), but no difference in the number of rehospitalizations (mean, 1.6 vs 1.5 days; P = .3272). During the 1-year follow-up, 554 (39.5%) patients had GREs, and this resulted in more rehospitalizations (mean, 2.1 vs 1.1; P < .0001) and a longer cumulative length of stay (mean, 28.2 vs 21.9 days; P < .0001) compared with patients without GRE. Multivariable analysis demonstrated the highly positive association of TL (hazard ratio [HR], 1.75) and early GRE (HR, 1.77) with the index length of stay, whereas comorbidities-namely, dialysis dependency (HR, 1.31), nonsmoking status (HR, 1.29), hypertension (HR, 1.26), and increasing age (HR, 1.01)-also had strong effects. The effect of TL and GRE on later RU (number of rehospitalizations and cumulative length of stay) was present but less pronounced than patient comorbidities (namely, dialysis). CONCLUSIONS: The stage of disease at presentation (TL vs rest pain) and the patency of the bypass graft (freedom from GRE) are critical determinants of RU over the first year after limb-salvage surgery. These effects predominate early (index length of stay) and persist through 1 year. Patient-specific factors, particularly dialysis-dependent renal failure, are also critical comorbidities affecting RU in these patients.
Chronic progressive cardiac dysfunction years after doxorubicin therapy for childhood acute lymphoblastic leukemia.
PURPOSE: Cross-sectional studies show that cardiac abnormalities are common in long-term survivors of doxorubicin-treated childhood malignancies. Longitudinal data, however, are rare. METHODS: Serial echocardiograms (N = 499) were obtained from 115 doxorubicin-treated long-term survivors of childhood acute lymphoblastic leukemia (median age at diagnosis, 4.8 years; median follow-up after completion of doxorubicin, 11.8 years). Results were expressed as z scores to indicate the number of standard deviations (SDs) above (+) or below (-) the normal predicted value. Median individual and cumulative doxorubicin doses were 30 mg/m2 per dose and 352 mg/m2, respectively. RESULTS: Left ventricular fractional shortening was significantly reduced after doxorubicin therapy, and the reduction was related to cumulative dose. z scores for fractional shortening transiently improved before falling to -2.76 more than 12 years after diagnosis. Reduced fractional shortening was related to impaired contractility and increasing afterload, consequences of a progressive reduction of ventricular mass, and wall thickness relative to body-surface area. Left ventricular contractility fell significantly over time and was depressed at last follow-up in patients receiving more than 300 mg/m2 of doxorubicin. Systolic and diastolic blood pressures were below normal more than 9 years after diagnosis. Even patients receiving lower cumulative doxorubicin doses experienced reduced mass and dimension. Fractional shortening and dimension at the end of therapy predicted these parameters 11.8 years later. CONCLUSION: Cardiac abnormalities were persistent and progressive after doxorubicin therapy. Inadequate ventricular mass with chronic afterload excess was associated with progressive contractile deficit and possibly reduced cardiac output and restrictive cardiomyopathy. The deficits were worst after highest cumulative doses of doxorubicin, but appeared even after low doses.
Sex differences in the effect of diabetes duration on coronary heart disease mortality.
BACKGROUND: It is not known whether the coronary heart disease (CHD) mortality risk associated with recent (RDM; <10 years) or long-standing diabetes mellitus (LDM; > or =10 years) varies by sex. METHODS: The relationship between diabetes duration and CHD mortality was evaluated among 10 871 adults (aged 35-74 years at baseline) using the 1971-1992 National Health and Nutrition Examination Survey Epidemiologic Follow-up Study. RESULTS: The CHD mortality rates per 1000 person-years in men with no myocardial infarction (MI) or diabetes, MI only, RDM only, LDM only, MI and RDM, and MI and LDM were 5.5 (95% confidence interval, 4.8-6.2), 15.2 (11.6-20.0), 13.2 (7.9-22.1), 11.4 (6.4-20.3), 36.0 (16.7-77.7), and 35.4 (14.0-89.7), respectively. The corresponding rates in women were 2.9 (2.5-3.3), 7.3 (5.0-10.8), 5.2 (3.5-7.7), 10.7 (7.5-15.5), 9.3 (4.3-19.9), and 21.6 (6.1-76.0), respectively. Compared with MI, the multivariate hazard ratios and their 95% confidence intervals (adjusted for age, race, smoking, hypertension, total cholesterol level, and body mass index) for fatal CHD in men with RDM, LDM, MI and RDM, and MI and LDM were 0.7 (0.3-1.3), 0.8 (0.4-1.4), 3.2 (1.4-7.4), and 2.4 (0.8-6.7), respectively. The corresponding ratios in women were 0.9 (0.6-1.3), 1.8 (1.1-3.2), 1.3 (0.5-3.5), and 1.6 (0.2-10.9), respectively. CONCLUSIONS: In men, RDM and LDM were associated with as high a risk for CHD death as MI. In women, although RDM had a CHD mortality risk similar to MI, LDM had an even greater risk. Because women with LDM are at very high risk for CHD mortality, current guidelines may need to be further refined to match intensity of treatment to risk in these women.
Coronary artery calcifications in the long-term follow-up of Kawasaki disease.
BACKGROUND: To determine if detection of coronary artery calcifications in patients with Kawasaki disease may serve as a noninvasive predictor of future coronary artery events. METHODS: A prospective, cohort pilot study that included 18 patients with Kawasaki disease > 1 year from the acute disease was performed including 9 patients with coronary abnormalities during the acute illness (Group 1) and 9 without coronary abnormalities (Group 2). Patients were classified by echocardiography as having none, resolved, or residual coronary artery abnormalities. Electron beam computed tomography (EBCT) scans were completed using the Agatson coronary calcium scoring system. Intermediate follow-up was performed 2.5 years after EBCT to determine if clinically significant coronary artery events (myocardial infarction or sudden death) had occurred. RESULTS: Late echocardiographic abnormalities corresponded with the early echocardiographic abnormalities in 5 of 9 patients (P = .029) in Group 1. The late echocardiographic abnormalities significantly correlated with detection of calcifications by EBCT in 4 of 5 patients (95% CI 28%-99%). One patient with residual coronary abnormalities and coronary artery calcifications with the highest calcium score subsequently had a sudden death. Detection of coronary artery calcifications may be predictive of sudden death (P = .056). No residual echocardiographic abnormalities, coronary artery calcifications, or coronary artery events occurred in Group 2 patients. CONCLUSIONS: Patients with Kawasaki disease with residual coronary abnormalities show EBCT evidence of coronary artery calcifications. Detection of coronary artery calcifications may be useful for risk stratification in the long-term management of patients with Kawasaki disease.
Quasi-likelihood estimation for relative risk regression models.
For a prospective randomized clinical trial with two groups, the relative risk can be used as a measure of treatment effect and is directly interpretable as the ratio of success probabilities in the new treatment group versus the placebo group. For a prospective study with many covariates and a binary outcome (success or failure), relative risk regression may be of interest. If we model the log of the success probability as a linear function of covariates, the regression coefficients are log-relative risks. However, using such a log-linear model with a Bernoulli likelihood can lead to convergence problems in the Newton-Raphson algorithm. This is likely to occur when the success probabilities are close to one. A constrained likelihood method proposed by Wacholder (1986, American Journal of Epidemiology 123, 174-184), also has convergence problems. We propose a quasi-likelihood method of moments technique in which we naively assume the Bernoulli outcome is Poisson, with the mean (success probability) following a log-linear model. We use the Poisson maximum likelihood equations to estimate the regression coefficients without constraints. Using method of moment ideas, one can show that the estimates using the Poisson likelihood will be consistent and asymptotically normal. We apply these methods to a double-blinded randomized trial in primary biliary cirrhosis of the liver (Markus et al., 1989, New England Journal of Medicine 320, 1709-1713).
Cardiac changes associated with growth hormone therapy among children treated with anthracyclines.
OBJECTIVE: The objective was to assess the cardiac effects of growth hormone (GH) therapy. Anthracycline-treated childhood cancer survivors frequently have reduced left ventricular (LV) wall thickness and contractility, and GH therapy may affect these factors. METHODS: We examined serial cardiac findings for 34 anthracycline-treated childhood cancer survivors with several years of GH therapy and baseline cardiac z scores similar to those of a comparison group (86 similar cancer survivors without GH therapy). RESULTS: LV contractility was decreased among GH-treated patients before, during, and after GH therapy (-1.08 SD below the age-adjusted population mean before therapy and -1.88 SD 4 years after therapy ceased, with each value depressed below normal). Contractility was higher in the control group than in the GH-treated group, with this difference being nearly significant. The GH-treated children had thinner LV walls before GH therapy (-1.38 SD). Wall thickness increased during GH therapy (from -1.38 SD to -1.09 SD after 3 years of GH therapy), but the effect was lost shortly after GH therapy ended and thickness diminished over time (-1.50 SD at 1 year after therapy and -1.96 SD at 4 years). During GH therapy, the wall thickness for the GH-treated group was greater than that for the control group; however, by 4 years after therapy, there was no difference between the GH-treated group and the control group. CONCLUSIONS: GH therapy among anthracycline-treated survivors of childhood cancer increased LV wall thickness, but the effect was lost after therapy was discontinued. The therapy did not affect the progressive LV dysfunction.
Cardiovascular status in long-term survivors of Hodgkin's disease treated with chest radiotherapy.
PURPOSE: Cardiovascular status was assessed in 48 Hodgkin's disease (HD) survivors at a median of 14.3 years (range, 5.9 to 27.5 years) after diagnosis because they may be at increased risk for cardiovascular abnormalities. PATIENTS AND METHODS: Patients completed the Short-Form 36 quality-of-life instrument and were screened by echocardiography, exercise stress testing, and resting and 24-hour ECG. RESULTS: All patients received mediastinal irradiation (median, 40.0 Gy; range, 27.0 to 51.7 Gy) at a median age of 16.5 years (range, 6.4 to 25.0 years). Four patients received an anthracycline. Although every patient described their health as good or better, and none had symptomatic heart disease at screening, all but one had cardiac abnormalities on screening. Restrictive cardiomyopathy was suggested by reduced average left ventricular (LV) dimension (P < .001) and mass (P < .001), without increased LV wall thickness. Significant valvular defects were present in 42%; 75% had conduction defects. One survivor developed complete heart block shortly after the study visit. Autonomic dysfunction was suggested by a monotonous heart rate in 57%, persistent tachycardia in 31%, and blunted hemodynamic responses to exercise in 27%. Peak oxygen uptake (VO2max) during exercise, a predictor of mortality in heart failure, was significantly reduced (< 20 mL/kg/m2) in 30% of survivors. VO2max was correlated with increasing fatigue, increasing shortness of breath (both, r = -0.35; P =. 02), and decreasing physical component score on the SF-36 (r = 0.554; P = .00017). CONCLUSION: A variety of unsuspected, clinically significant cardiovascular abnormalities are common in long-term survivors of HD who are treated at a young age with mediastinal irradiation. We recommend serial, comprehensive cardiac screening of HD survivors who fit this profile.
The effect of dexrazoxane on myocardial injury in doxorubicin-treated children with acute lymphoblastic leukemia.
BACKGROUND: Doxorubicin chemotherapy is very effective in children with acute lymphoblastic leukemia (ALL) but also injures myocardial cells. Dexrazoxane, a free-radical scavenger, may protect the heart from doxorubicin-associated damage. METHODS: To determine whether dexrazoxane decreases doxorubicin-associated injury of cardiomyocytes, we randomly assigned 101 children with ALL to receive doxorubicin alone (30 mg per square meter of body-surface area every three weeks for 10 doses) and 105 to receive dexrazoxane (300 mg per square meter) followed immediately by doxorubicin. Serial measurements of serum cardiac troponin T were obtained in 76 of 101 patients in the doxorubicin group and 82 of 105 patients in the group given dexrazoxane and doxorubicin. A total of 2377 serum samples (mean, 15.1 samples per patient) were obtained before, during, and after treatment with doxorubicin. Troponin T levels were evaluated in a blinded fashion to determine whether they were elevated (>0.01 ng per milliliter)--the primary end point--or extremely elevated (>0.025 ng per milliliter). RESULTS: Elevations of troponin T occurred in 35 percent of the patients (55 of 158). Patients treated with doxorubicin alone were more likely than those who received dexrazoxane and doxorubicin to have elevated troponin T levels (50 percent vs. 21 percent, P<0.001) and extremely elevated troponin T levels (32 percent vs. 10 percent, P<0.001). The median follow-up was 2.7 years. The rate of event-free survival at 2.5 years was 83 percent in both groups (P=0.87 by the log-rank test). CONCLUSIONS: Dexrazoxane prevents or reduces cardiac injury, as reflected by elevations in troponin T, that is associated with the use of doxorubicin for childhood ALL without compromising the antileukemic efficacy of doxorubicin. Longer follow-up will be necessary to determine the influence of dexrazoxane on echocardiographic findings at four years and on event-free survival.
Association between congenital cardiovascular malformations and neuroblastoma.
OBJECTIVE: We explored the association between neuroblastoma and congenital cardiovascular malformations (CCM), previously described in case reports. STUDY DESIGN: Echocardiogram and chart reviews of a series of 158 patients with neuroblastoma and a control group of 192 children with leukemia were performed. The proportion of patients with CCM in each group was compared. RESULTS: Fourteen of the 70 (20%) patients with neuroblastoma and echocardiography had CCM, compared with 7 of the 192 (3.6%) patients with leukemia with echocardiograms (P=.0001). If all of the patients with neuroblastoma without echocardiograms (n=88) are considered to have normal cardiac anatomy, this difference remains significant (14 of 158 patients with neuroblastoma have CCM detected [8.9%] versus 7 of 192 patients with leukemia [3.6%]; P=.045). Neural crest-derived CCM were more common in patients with neuroblastoma, detected in 5 of 70 patients with neuroblastoma versus 2 of 192 patients with leukemia (P=.016). Congenital cardiovascular malformations in patients with neuroblastoma were associated with a cancer diagnosis at age less than 1 year and a lower neuroblastoma stage, but there was no association with tumor MYCN amplification. CONCLUSIONS: Neuroblastoma and CCM may be associated. We recommend echocardiography for CCM screening in patients with newly diagnosed neuroblastoma.
A clinic-based, comprehensive care model for studying late effects in long-term survivors of pediatric illnesses.
OBJECTIVE: Long-term survivors of several childhood illnesses are at risk for multiple late effects of their illness or therapy, and children with documented toxic exposures may also experience long-term health consequences. Clinical studies of these effects are difficult to conduct. The Cardiovascular Status in Childhood Cancer Survivors Study is an established study that highlights the ability to perform comprehensive clinical investigations when patients are cared for in a formal, long-term follow-up clinic. This clinic model facilitates long-term retention and recruitment of patients, allowing comprehensive clinical studies of late effects of illness or exposures, in this case, of cardiovascular complications of cancer treatment during childhood. METHODS: The study is funded through the National Institute of Health Office of Cancer Survivorship. Participants are recruited from the Long-Term Survivors Clinic at the University of Rochester. The clinic provides care for all survivors of childhood cancer in the region. The Long-Term Survivors Clinic provides medical care and psychosocial and educational support to patients and facilitates coordination of care. Patients remain in close contact with clinic staff for extended periods. RESULTS: We recruited a representative sample of this long-term survivor population, with a wide range of ages, diagnoses, and time since diagnosis. Longitudinal collection of detailed clinical data will enable us to conduct cohort studies of late effects as well as case-control studies of toxic exposures. CONCLUSIONS: The success of this study shows the advantages of formal programs for continued care of patients with chronic illnesses or treatment or toxic exposures. The Long-Term Survivors Clinic provides an excellent model for clinical care and research that is applicable to multiple pediatric and young adult populations.
Analysis of clustered matched-pair data.
Evaluation of the performance of a new diagnostic procedure with respect to a standard procedure arises frequently in practice. The response of interest, often in a dichotomous form, is measured twice, once with each procedure. The two procedures are administered to either two matched individuals, or when practical, to the same individual. A large sample test for matched-pair data is the McNemar test. The main assumption of this test is independent paired responses; however, when more than one outcome from an individual is measured by each procedure, the data are clustered. Examples of such cases can be seen in dental and ophthalmology studies. Variance adjustment methods for the analysis of clustered matched-pair data have been proposed; however, because of unequal cluster sizes, variability of correlation structures within a cluster (within paired responses in a cluster as well as between paired responses in a cluster), and unequal success probabilities among the clusters, the performances of some available methods are not consistent. This research proposes a simple adjustment to the McNemar test for the analysis of clustered matched-pair data. Method of moments is used to calculate a consistent variance estimator. Using Monte Carlo simulation, the size and power of the proposed test are compared to those of two currently available methods. To illustrate practical application, clustered matched-pair data from two clinical studies are analysed.
Sex differences in risk for coronary heart disease mortality associated with diabetes and established coronary heart disease.
BACKGROUND: The sex-specific independent effect of diabetes mellitus and established coronary heart disease (CHD) on subsequent CHD mortality is not known. METHODS: This is an analysis of pooled data (n = 5243) from the Framingham Heart Study and the Framingham Offspring Study with follow-up of 20 years. At baseline (1971-1975), 134 men and 95 women had diabetes, while 222 men and 129 women had CHD. Risk for CHD death was analyzed by proportional hazards models, adjusting for age, hypertension, serum cholesterol levels, smoking, and body mass index. The comparative effect of established CHD vs diabetes on the risk of CHD mortality was tested by testing the difference in log hazards. RESULTS: The adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) for death from CHD were 2.1 (95% CI, 1.3-3.3) in men with diabetes only, and 4.2 (95% CI, 3.2-5.6) in men with CHD only compared with men without diabetes or CHD. The HR for CHD death was 3.8 (95% CI, 2.2-6.6) in women with diabetes, and 1.9 (95% CI, 1.1-3.4) in women with CHD. The difference between the CHD and the diabetes log hazards was +0.73 (95% CI, 0.72-0.75) in men and -0.65 (95% CI, -0.68 to -0.63) in women. CONCLUSIONS: In men, established CHD signifies a higher risk for CHD mortality than diabetes. This is reversed in women, with diabetes being associated with greater risk for CHD mortality. Current treatment recommendations for women with diabetes may need to be more aggressive to match CHD mortality risk.