Skip to main content

Table 2 Stepwise approach to examine energy-adjusted zinc as an independent predictor of a new diagnosis of diabetes

From: Is dietary zinc protective for type 2 diabetes? Results from the Australian longitudinal study on women’s health

  Quintile of energy-adjusted zinc
  Q1 Q2 Q3 Q4 Q5 P
Number of women 1785 1784 1784 1784 1784  
Energy-adjusted zinc [median (min, max)] −1.25 (-4.8, -0.79) −0.48 (-0.79, -0.23) 0.01 (-0.23, 0.26) 0.50 (0.26, 0.79) 1.24 (0.79, 4.45)  
Number of diabetics 80 60 59 74 60  
Odds ratio       
• Age adjusted 1.00 0.74 (0.53 to 1.05) 0.73 (0.52 to 1.03) 0.99 (0.67 to 1.28) 0.75 (0.53 to 1.05) 0.319
• Age & non-dietary factors adjusted 1.00 1.00 0.82 (0.56 to 1.19) 0.65 (0.44 to 0.96) 0.83 (0.58 to 1.19) 0.56 (0.38 to 0.83) 0.010
• Age, non-dietary and dietary factors adjusted 1.00 0.78 (0.53 to 1.15) 0.60 (0.40 to 0.91) 0.77 (0.52 to 1.13) 0.48 (0.31 to 0.75) 0.004
• Age, non-dietary and dietary factors adjusted plus alcohol intake and use of supplements   0.80 (0.54 to 1.17) 0.60 (0.40 to 0.90) 0.78 (0.53 to 1.15) 0.50 (0.32 to 0.77) 0.006
  1. Non-dietary factors were BMI; smoking status; HRT; exercise group; and history of arthritis, CHD, hypertension, asthma and depression.
  2. Dietary factors were energy-adjusted fiber, iron and fat.
  3. Adjustment for family income in the models resulted in a loss of 1300 observations but the p-values for the test for trend were very similar (p = 0.010 for all 3 adjusted models).