Using Multivariate Quantile Regression Analysis to Explore Cardiovascular Risk Differences in Subjects with Chronic Kidney Disease by Race and Ethnicity: Findings from the U.S. Chronic Renal Insufficiency Cohort Study


  • Longjian Liu Drexel University School of Public Health



Background and Aims: Adults with chronic kidney disease (CKD) carry an extraordinarily high risk for cardiovascular disease (CVD).The present study aimed to test two hypotheses that: (1) CVD risk factors disproportionately affect non-Hispanic black (NHB) with CKD than non-Hispanic white (NHW). (2) This difference significantly contributes to an excess risk of CVD in NHB versus NHW.


Methods: A total of 3,939 aged 21-74 years old participating in the Chronic Renal Insufficiency Cohort Study was analyzed. A sum weighted CVDRisk score was constructed from well-established CVD risk factors. Differences in CVDRisk score by race/ethnicity were tested using quantile regression (Qreg) analysis.


Results: The prevalence of CVD was 30.7% in NHW and 38.2% in NHB (p<0.001). The means (SD) of CVDRisk score were 12.6 (5.7) in NHW and 14.6 (6.4) in NHB (p<0.001). Qreg analysis indicated that NHB with estimate glomerular filtration rate (eGFR) 30-59.9 ml/min/1.73m2 had significantly higher (worse) CVDRisk scores across all quantiles (Qs) than NHW. This race differences in CVDRisk were also significantly higher in NHB with eGFR 60-70 ml/min/1.73m2 in Qs 1 and 2 as compared to their NHW counterparts. An estimated 35.8% of the excess prevalent CVD could be attributable to the difference in CVDRisk for NHB versus NHW.


Conclusion: NHB have a significantly higher CVD risk factor score in those with moderate and mild CKD than NHW.

Author Biography

Longjian Liu, Drexel University School of Public Health

Associate Professor, Department of Epidemiology and Biostatistics

Interim Chair, Department of Environmental and Occupational Health

Drexel University School of Public Health, Philadelphia, PA, United States






Original Article