Table 4

Association between prefrailty, frailty and instrumental activities of daily living limitations: unadjusted, age-adjusted and sex-adjusted and multivariable-adjusted zero-inflated Poisson regression results

ModelsLogistic section, OR (95% CI)*Poisson section, RR (95% CI)†
FrailtyPrefrailtyFrailtyPrefrailty
Model 1: unadjusted0.06 (0.02 to 0.24)0.47 (0.18 to 1.21)5.93 (2.99 to 11.75)1.81 (0.87 to 3.74)
Model 2: model 1+sex and age0.10 (0.02 to 0.47)0.44 (0.15 to 1.31)3.92 (2.21 to 6.95)1.29 (0.73 to 2.30)
Model 3: model 2+longest held occupation0.11 (0.02 to 0.56)0.46 (0.16 to 1.32)3.94 (2.11 to 7.36)1.30 (0.70 to 2.43)
Model 4: model 3+social support0.13 (0.02 to 0.79)0.49 (0.15 to 1.63)4.35 (2.31 to 8.22)1.36 (0.73 to 2.53)
Model 5: model 4+multimorbidity0.12 (0.02 to 0.69)0.44 (0.14 to 1.40)4.21 (2.26 to 7.83)1.30 (0.71 to 2.39)
Model 6: model 5+self-perceived vision ability0.11 (0.02 to 0.68)0.34 (0.05 to 2.25)4.13 (2.25 to 7.59)1.22 (0.68 to 2.20)
Model 7: model 6+self-perceived hearing ability0.11 (0.02 to 0.59)0.33 (0.06 to 1.84)4.16 (2.27 to 7.60)1.21 (0.67 to 2.16)
  • Statistically significant estimates (at the 5% level) are displayed in bold.

  • *Logistic section of the regression model estimates the log-odds of belonging to the ‘sure-zero’/‘not-at-risk’ class.

  • †Poisson section of the regression model estimates the count of IADL limitations for those estimated to belong to the ‘non-sure zero’/‘at risk’ latent class.

  • OR, odds ratio; RR, rate ratio.