Estimate | SE | p Value | |
---|---|---|---|

Arthritis (N=3653) | |||

Time 1 total effects (path A) | −0.701 | 0.660 | 0.288 |

Time 2 total effects (path C+(path A×path B)) | 0.196 | 0.580 | 0.736 |

Time 2 total effect estimate | |||

Direct effect (path C) | 0.479 | 0.549 | 0.383 |

Effect mediated through time 1 work status (path A×path B) | −0.283 | 0.267 | 0.289 |

Model fit: CFI=0.989, TLI=0.978; RMSEA 0.015 (0.009–0.021) | |||

Back problems (N=4439) | |||

Time 1 total effects (path A) | 0.447 | 0.568 | 0.432 |

Time 2 total effects (path C+(path A×path B)) | −0.057 | 0.566 | 0.920 |

Time 2 total effect estimate | |||

Direct effect (path C) | −0.282 | 0.507 | 0.577 |

Effect mediated through time 1 work status (path A×path B) | 0.226 | 0.287 | 0.431 |

Model fit: CFI=0.942, TLI=0.889; RMSEA 0.027 (0.023–0.030) | |||

Diabetes (N=1029) | |||

Time 1 total effects (path A) | 0.084 | 1.088 | 0.939 |

Time 2 total effects (path C+(path A×path B)) | 0.382 | 1.085 | 0.725 |

Time 2 total effect estimate | |||

Direct effect (path C) | 0.335 | 1.046 | 0.749 |

Effect mediated through time 1 work status (path A×path B) | 0.047 | 0.609 | 0.939 |

Model fit: CFI=0.966, TLI=0.939; RMSEA 0.023 (0.013–0.032) | |||

Hypertension (N=2946) | |||

Time 1 total effects (path A) | −0.777 | 0.637 | 0.223 |

Time 2 total effects (path C+(path A×path B)) | 0.173 | 0.689 | 0.802 |

Time 2 total effect estimate | |||

Direct effect (path C) | 0.478 | 0.637 | 0.453 |

Effect mediated through time 1 work status (path A×path B) | −0.305 | 0.254 | 0.229 |

Model fit: CFI=0.989, TLI=0.978; RMSEA 0.013 (0.006–0.018) | |||

Heart disease (N=803) | |||

Time 1 total effects (path A) | −2.452 | 1.223 | 0.045 |

Time 2 total effects (path C+(path A×path B)) | −0.546 | 1.31 | 0.677 |

Time 2 total effect estimate | |||

Direct effect (path C) | 0.402 | 1.194 | 0.736 |

Effect mediated through time 1 work status (path A×path B) | −0.949 | 0.471 | 0.044 |

Model fit: CFI=0.970, TLI=0.946; RMSEA 0.022 (0.007–0.033). |

Path model corresponds to figure 1 and is stratified by a chronic health condition of interest. Negative estimates indicate lower than expected work hours.

A positive and larger estimate indicates an increase in the likelihood of an outcome.

*Estimate adjusted for baseline measures of sex, age (grouped) comorbidities, body mass index (grouped), occupational strength requirements, self-reported depression, marital status, geographic location, survey year and work hours. Based on the process of selecting respondents, each model may consist of a different subsample of the overall cohort and may to some extent overlap.

CFI, comparative fit index; RMSEA, root mean square error of approximation; TLI, Tucker-Lewis index.