Table 6

Risk of prevalent disease, by supplement user group.
Outcome
 Cases: n/(%) **
 OR*
 95% CI*
Diabetes


     Multiple Supp users
 8/(2.9)
 0.27
 0.12–0.59
     Single Supp users
 23/(11.1)
 1.11
 0.74–1.67
     No Supp users
 70/(8.5)
 1.00
Coronary Heart Disease


     Multiple Supp users
 14/(5.0)
 0.48
 0.22–1.09
     Single Supp users
 17/(7.1)
 1.08
 0.52–2.25
     No Supp users
 47/(5.6)
 1.00
Heart Attack


     Multiple Supp users
 7/(2.5)
 0.51
 0.18–1.41
     Single Supp users
 16/(7.6)
 1.76
 0.90–3.42
     No Supp users
 44/(4.4)
 1.00
Angina


     Multiple Supp users
 6/(2.2)
 0.41
 0.14–1.18
     Single Supp users
 12/(6.8)
 1.71
 0.82–3.59
     No Supp users
 35/(4.7)
 1.00
Congestive Heart Failure


     Multiple Supp users
 4/(1.4)
 0.73
 0.18–2.95
     Single Supp users
 9/(2.8)
 0.88
 0.26–2.97
     No Supp users
 31/(3.5)
 1.00
Stroke


     Multiple Supp users
 5/(1.8)
 1.42
 0.39–5.17
     Single Supp users
 6/(2.6)
 1.26
 0.34–4.67
     No Supp users
 29/(2.5)
 1.00
Arthritis


     Multiple Supp users
 88/(31.7)
 1.00
 0.69–1.46
     Single Supp users
 71/(31.8)
 1.23
 0.77–1.98
     No Supp users
 189/(26.7)
 1.00
Emphysema


     Multiple Supp users
 3/(1.1)
 0.55
 0.09–3.41
     Single Supp users
 4/(1.8)
 0.73
 0.32–1.64
     No Supp users
 23/(2.3)
 1.00
Health Status†


     Multiple Supp users
 44/(15.8)
 0.26
 0.17–0.41
     Single Supp users
 86/(44.4)
 0.75
 0.45–1.23
     No Supp users
 302/(50.9)
 1.00

* Logistic regression methods were used to estimate odds ratios (OR) with 95% confidence intervals (CI), adjusted for sex, age and age2, education, income and income2, and body mass index. The referent group for risk estimation was the No Supp group.

** Number of events, weighted percent

†Self-assessed as 'good, fair, or poor' versus 'excellent or very good.'

Block et al. Nutrition Journal 2007 6:30   doi:10.1186/1475-2891-6-30