Testing the proportional hazards assumption in case-cohort analysis

Xue, Xiaonan; Xie, Xianhong; Gunter, Marc J.; Rohan, Thomas E.; Wassertheil‐Smoller, Sylvia; Ho, Gloria Y.F.; Cirillo, Dominic J.; Yu, Herbert; Strickler, Howard D.

doi:10.1186/1471-2288-13-88

Cited by 116 publications

(72 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hazard ratios (HR) and 95% confidence intervals (CI) was calculated. The assumption of proportionality was tested by plotting scaled Schoenfeld residuals41. Several multivariable Cox proportional hazard models were built to adjust for the effects of confounding covariates: (1) Model 1: a univariate analysis; (2) Model 2: adjusted for age, sex, time 1–2 and time 2–3 .…”

Section: Methodsmentioning

confidence: 99%

Cumulative Resting Heart Rate Exposure and Risk of All-Cause Mortality: Results from the Kailuan Cohort Study

Zhao

Wang

et al. 2017

Sci Rep

View full text Add to dashboard Cite

The relationship between cumulative exposure to resting heart rate (cumRHR) and mortality remain unclear in the general population. In the Kailuan cohort study, resting heart rate (RHR) was repeatedly measured at baseline and at years 2 and 4 by electrocardiogram among 47,311 adults aged 48.70 ± 11.68. The cumRHR was defined as the summed average RHR between two consecutive examinations multiplied by the time interval between with two examinations [(beats/min) * year]. A higher RHR was defined as ≥80 beats/min, and the number of visits with a higher RHR was counted. During a median of 4.06 years of follow-up, a total of 1,025 participants died. After adjusting for major traditional cardiovascular risk factors and baseline RHR, the hazard ratio for the highest versus lowest quartile of cumRHR was 1.39 (95% CI: 1.07–1.81) for all-cause mortality. Each 1-SD increment in cumRHR was associated with a 37% (HR: 1.37, 95% CI: 1.23–1.52) increased risk of death and displayed a J-shaped relationship. Compared with no exposure, adults who had a higher RHR at all 3 study visits were associated with a 1.86-fold higher risk (95% CI: 1.33–2.61) of mortality. In summary, cumulative exposure to higher RHR is independently associated with an increased risk of mortality.

show abstract

Section: Methodsmentioning

confidence: 99%

Cumulative Resting Heart Rate Exposure and Risk of All-Cause Mortality: Results from the Kailuan Cohort Study

Zhao

Wang

et al. 2017

Sci Rep

View full text Add to dashboard Cite

show abstract

“…The proportional hazards assumption was validated by a goodness-of-test based on scaled Schoenfeld residuals. (10) We stratified by variables for which the proportional hazards assumption did not hold. For all analyses, a two-sided p value <0.05 was considered statistically significant.…”

Section: Methodsmentioning

confidence: 99%

Aortic Atheroma Increases the Risk of Long-Term Mortality in 20,000 Patients

Butler

Luxford

Huang

et al. 2017

The Annals of Thoracic Surgery

View full text Add to dashboard Cite

Background The association between long-term mortality and aortic atheroma in cardiac surgical patients has not been comprehensively investigated. In this study we determine the relationship between grade of atheroma and the risk of long-term mortality in a retrospective cohort of over 20,000 patients undergoing cardiac surgery over a 20 - year period. Methods We included 22,304 consecutive intraoperative transesophageal and epiaortic ultrasound examinations performed at Brigham and Women’s Hospital between 1995 and 2014, with long-term follow-up. The extent of atheromatous disease recorded in each examination was used for analysis. Mortality data was obtained from our institution’s data registry. Mortality analyses were done using Cox proportional hazard regression models with follow-up as a time scale. We repeated the analysis in a subgroup of 14,728 patients with more detailed demographics, including postoperative stroke, queried from the institutional Society of Thoracic Surgeons database. Results A total of 7,722 mortality events and 872 stroke events occurred. Patients with atheromatous disease demonstrated a significant increase in mortality across all grades of severity, both for the ascending and descending aorta. This relationship remained unchanged after adjusting for additional covariates. Adjustments for postoperative stroke resulted in only minimal attenuation in the risk of postoperative mortality related to aortic atheroma. Conclusions Aortic atheromatous disease of any grade in the ascending and descending aorta is a significant long-term risk of long-term, all-cause mortality in cardiac surgery patients. This association remains independent of other conventional risk factors, and is not related to postoperative cerebrovascular accidents.

show abstract

“…All Cox models were computed with robust variance estimation in order to account for the case-cohort design [36,37]. We tested whether all models met the proportional hazard assumption using a published method [38]. Non-linearity using sRAGE as a continuous variable in predicting CRC risk was tested by restricted cubic spline regression [39].…”

Section: Methodsmentioning

confidence: 99%

A prospective study of soluble receptor for advanced glycation end-products and colorectal cancer risk in postmenopausal women

Chen

Duan

Tinker

et al. 2016

Cancer Epidemiology

View full text Add to dashboard Cite

Objectives Receptor for advanced glycation end products (RAGE) expressed on adipocytes and immune cells can bind to ligand Nε-(carboxymethyl)-lysine (CML) and trigger dysregulation of adipokines and chronic inflammation. Soluble RAGE (sRAGE) mitigates the detrimental effect of RAGE. We examined the associations between circulating levels of CML-AGE and sRAGE and colorectal cancer (CRC). Methods In a case-cohort study of the Women’s Health Initiative Study, blood levels of CML-AGE and sRAGE were measured using ELISA. We used multivariable Cox regression model to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) of incident CRC in relation to quartiles (Q) of biomarker levels. Results Average follow-up was 7.8 years for 444 cases and 805 subcohort members. In the subcohort, CML-AGE and sRAGE were inversely correlated with BMI (P values < 0.0001). Levels of CML-AGE and sRAGE were not associated with CRC. In BMI-specific analysis, the association between sRAGE and CRC was observed. Among women with BMI ≥ 25 kg/m2, those with highest levels of sRAGE had significantly lower risk for CRC as compared to women with lowest levels of sRAGE (HRQ4 versus Q1: 0.39; 95% CI: 0.17–0.91). This inverse association was not observed among women with BMI <25 kg/m2 (P value for interaction = 0.01). Conclusions Among postmenopausal women, the RAGE pathway may be involved in obesity-related CRC.

show abstract

Testing the proportional hazards assumption in case-cohort analysis

Cited by 116 publications

References 28 publications

Cumulative Resting Heart Rate Exposure and Risk of All-Cause Mortality: Results from the Kailuan Cohort Study

Cumulative Resting Heart Rate Exposure and Risk of All-Cause Mortality: Results from the Kailuan Cohort Study

Aortic Atheroma Increases the Risk of Long-Term Mortality in 20,000 Patients

A prospective study of soluble receptor for advanced glycation end-products and colorectal cancer risk in postmenopausal women

Contact Info

Product

Resources

About