# Table 6 Hierarchical model with multivariate Poisson regression for frailty (n = 1030). FIBRA Unicamp, Elderly, 2008–2009

Variables First step Second step Third step
PR* 95 % CI p value PR* 95 % CI p value PR* 95 % CI p value
Age
65–69 1.00 - - 1.00 - - 1.00 - -
70–74 1.42 0.90–2.26 0.136 1.44 0.90–2.28 0.126 1.30 0.81–2.06 0.275
75–79 2.02 1.25–3.28 0.004 2.02 1.25–3.28 0.004 1.88 1.16–3.06 0.011
> = 80 3.71 2.30–5.99 <0.001 3.65 2.27–5.88 <0.001 3.05 1.88–4.95 <0.001
Schooling
0 2.55 1.28–5.09 0.008 2.29 1.15–4.58 0.019 1.56 0.77–3.15 0.218
1–4 1.64 0.84–3.21 0.146 1.52 0.78–2.96 0.224 1.14 0.58–2.25 0.705
5–8 1.39 0.65–2.96 0.399 1.26 0.59–2.70 0.547 1.01 0.47–2.17 0.986
> = 9 1.00 - - 1.00 - - 1.00 - -
No. of diseases
0     1.00 - - 1.00 - -
1–2     2.00 0.91–4.37 0.084 2.01 0.92–4.42 0.081
> = 3     3.28 1.51–7.14 0.003 2.74 1.26–5.97 0.011
GDS13
No        2.09 1.46–3.00 <0.001
Yes        1.00 - -
GDS3
No        1.00 - -
Yes        1.76 1.24–2.51 0.002
GDS15
No        1.00 - -
Yes        1.72 1.21–2.44 0.003
GDS2
No        1.00 - -
Yes        1.71 1.19–2.44 0.004
1. Note.*PR = Prevalence ratio for frailty; (n = 891 non-frail and n = 139 frail)
2. PR 95 % CI = 95 % confidence interval for prevalence ratio. Stepwise criteria for variable selection. Ref. reference level
3. Variables considered in step 1: age, sex, schooling, marital status and family income; step 2: number of diseases and number of disabilities; step 3: items GDS1 to GDS15