Skip to main content

Table 1 Statistical power expected for logistic analysis under a sample size of 2000 with various assumptions.

From: Design and baseline characteristics of a prospective cohort study for determinants of osteoporotic fracture in community-dwelling elderly Japanese men: the Fujiwara-kyo Osteoporosis Risk in Men (FORMEN) Study

Risk factor studied Annual incidence rate of fracture (%) in population without a risk factor studied  
Prevalence rate (%) RR 0.8 1 1.5 2 PARP (%)
20 1.25 14 18 24 41 4.8
  1.5 41 50 68 81 9.1
  1.75 72 81 94 99 13.0
  2 91 96 100 100 16.7
30 1.25 17 21 30 38 7.0
  1.5 43 59 78 89 13.0
  1.75 80 88 98 100 18.4
  2 95 98 100 100 23.1
40 1.25 19 23 32 42 9.1
  1.5 53 63 81 92 16.7
  1.75 83 91 98 100 23.1
  2 96 99 100 100 28.6
  1. Statistical power was estimated for a logistic regression analysis to yield a significant estimate for a risk factor with a two-tailed level of significance at 5% under a sample size of 2000 with various combinations of assumptions.
  2. RR: Relative risk of fracture posed by a risk factor studied.
  3. PARP: Population attributable risk proportion.
  4. Boldface: Power > 80%
  5. 81: Assumptions used when the present sample size of 2000 was determined.