Cigarette smoking and cardio-renal events in patients with atherosclerotic renal artery stenosis

Abstract

Cigarette smoking causes cardiovascular disease and is associated with poor kidney function in individuals with diabetes mellitus and primary kidney diseases. However, the association of smoking on patients with atherosclerotic renal artery stenosis has not been studied. The current study utilized data from the Cardiovascular Outcomes in Renal Atherosclerotic Lesions (CORAL, NCT00081731) clinical trial to evaluate the effects of smoking on the risk of cardio-renal events and kidney function in this population. Baseline data showed that smokers (n = 277 out of 931) were significantly younger at enrollment than non-smokers (63.3±9.1 years vs 72.4±7.8 years; p<0.001). In addition, patients who smoke were also more likely to have bilateral renal artery stenoses and peripheral vascular disease (PVD). Longitudinal analysis showed that smokers experienced composite endpoint events (defined as first occurrence of: stroke; cardiovascular or renal death; myocardial infarction; hospitalization for congestive heart failure; permanent renal replacement; and progressive renal insufficiency defined as 30% reduction of GFR from baseline sustained for ≥ 60 days) at a substantially younger age compared to non-smokers (67.1±9.0 versus 76.1±7.9, p<0.001). Using linear regression and generalized linear modeling analysis controlled by age, sex, and ethnicity, smokers had significantly higher cystatin C levels (1.3±0.7 vs 1.2±0.9, p<0.01) whereas creatinine and estimated glomerular filtration rate (eGFR) were not different from non-smokers. From these data we conclude that smoking has a significant association with deleterious cardio-renal outcomes in patients with renovascular hypertension.

Fig 1. CONSORT flow diagram of CORAL study.

Fig 2. Age in years for smokers and non-smokers at study enrollment.

The Red curve represents the distribution of age for smokers (N = 277) and the blue curve represents age at study enrollment for non-smokers (N = 644). Numbers at the peak of the distribution curves are the mean years of age at study enrollment ± SD for the patient populations as defined above. The insert boxplot shows age with interquartile ranges and 95% confidence intervals differentiated by smoking status. The asterisk (*) indicates that the means are significantly different (p<0.001) as determined by two-sample t-test.

Fig 3. Kaplan-Meier curves of event-free survival for age-at-composite endpoint delineated by smoking status.

The hazard ratio, assessed by log-rank test for age-at-composite endpoint delineated by smoking status, was 2.32 [1.79, 2.98], p<0.001.

Fig 4. Longitudinal analysis of the effect of smoking on kidney function.

Least square means measured over time-in-study are delineated by smoking status, and the panels display the following: a): log values of the means for creatinine (mg/dL); b): log values of the means for MDRD-eGFR (mL/min per 1.73m²); c): log values of the means for Urine Albumin to Creatinine Ratio (mg/g); and d): log values of the means for Cystatin C (mg/L). An asterisk (*) indicates the mean for smokers is significantly different than non-smokers value at same time point (p<0.05).

Fig 5. Fitted longitudinal slope of the natural log of MDRD-GFR over time-in-study by age at enrollment.

For each individual, the longitudinal slope of MDRD-GFR was obtained by linear regression between the log of MDRD-GFR and time-in-study for that patient. These slopes were used as the response variable and fitted into a multiple variable regression model with age, sex, ethnicity, smoking, diabetes and BMI as covariates. The plot of the predicted slope of MDRD-GFR over time with 95% confidence intervals was generated for age at enrollment grouped by smoking status (red for smokers and blue for non-smokers).

Fig 6. Longitudinal effect on blood pressure for smokers versus non-smokers.

Graphs represent the following: a): systolic blood pressure (mmHg); b): diastolic blood pressure (mmHg); and c): pulse pressure (mmHg). Mean ± SD for the patient data from baseline through follow-up are given. No significant differences were observed.