Multiple logistic analysis of the onset of microalbuminuria according to quartile of HbA1c and smoking status

Univariate odds ratiosAdjusted odds ratios
HbA1c (%)
 4.3–7.21 (reference)1 (reference)
 7.3–8.10.98 (0.4–2.4)0.85 (0.3–2.1)
 8.2–9.02.3 (1.1–5.0)1.9 (0.9–4.2)
 9.1–16.56.3 (3.1–12.7)4.7 (2.3–9.7)
Smoking status
 Never1 (reference)1 (reference)
 Past0.84 (0.4–1.8)0.81 (0.4–1.8)
 Current3.8 (2.4–6.0)3.1 (1.9–5.1)
  • Data are n (95% CI) and include only patients with HbA1c and smoking information. (This includes 80 out of 109 of those with onset of microalbuminuria and 1,183 2-year follow-up intervals for analysis.) Univariate model was adjusted for period of analysis only. Adjusted model included both the quartiles of HbA1c and smoking status, and is adjusted for age at diabetes diagnosis, duration of diabetes, and period of analysis. When the analysis was repeated using only individuals with three or more ACR measurements (or two concordant measurements) in both the baseline and follow-up interval (784 2-year follow-up intervals and 44 onsets of microalbuminuria), the results were unchanged. The odds ratios were 0.79, 1.9, and 4.4 for the second, third, and fourth quartiles of HbA1c relative to the risk of microalbuminuria in the first quartile. The odds ratio for current smokers was 3.5 and for past smokers was 0.31 relative to the risk of microalbuminuria for those who never smoked.