兰德-Healthy-Vitality-Age_43页_3mb

Summary of Healthy Vitality Age (HVA) Model

Core Content

The Healthy Vitality Age (HVA) model is an extension of the Vitality Age (VA.3) model, which assesses an individual's health status based on their lifestyle and clinical risk factors. HVA incorporates both mortality and morbidity to provide a more comprehensive health-adjusted age estimate. It compares an individual's health-adjusted life expectancy (HALE) with that of the general population to estimate how many healthy years of life they have lost.

Main Points

• VA.3 is a risk-adjusted health assessment tool that estimates an individual's life expectancy by comparing their risk-factor profile to the population average.
• HVA builds upon VA.3 by incorporating Disability-Adjusted Life Years (DALYs) and Years Lived with Disability (YLDs), which account for the years of life lost due to disability.
• The HVA formula is:
  $$
  \text{HVA} = \text{chronological age} + \text{population HALE} - \text{individual HALE}
  $$
• HALE is calculated using Sullivan's method, which integrates mortality rates and disability weights to estimate the number of years lived in good health.

Key Information

2.1 Summary of Vitality Age Model

• The model uses risk factor-cause of death pairs and relative risks (RR) to compute individual mortality rates.
• It calculates individualised probability of dying by adjusting the baseline mortality rate (BD) using Population Attributable Fractions (PAFs) and relative risks (RRs).
• The process is repeated for all ages up to 120 years to generate a comprehensive overview of mortality risks.
• Vitality Age (VA) is derived as the difference between the chronological age and the life expectancy of the individual compared to the population.

2.2 Estimation of Healthy Vitality Age

• Population HALE is calculated using life tables that include YLDs and disability weights.
• Individual HALE is estimated by adjusting population mortality rates and YLDs for the individual's risk profile.
• YLDs are calculated as:
  $$
  YLD_o = DW_o \times p_o
  $$
  where $DW_o$ is the disability weight and $p_o$ is the prevalence of the cause.
• Adjusted YLDs are estimated using counterfactual prevalence rates and relative risks (RRs), where:
  $$
  YLD_o' = DW_o \times p_o'
  $$
  and $p_o'$ is the adjusted prevalence rate based on the individual's risk profile.
• HALE' for an individual is then calculated as:
  $$
  HALE'(x) = (1 - YLD'(x)) \times L(x)
  $$
  where $L(x)$ is the number of person-years lived between ages $x$ and $x + 1$.

3. Data

• The HVA model uses data from the Global Burden of Disease (GBD) database.
• The GBD database includes mortality rates (MRs), Population Attributable Fractions (PAFs), relative risks (RRs), and disability weights (DWs).
• Data are available for 5-year age groups, and include prevalence rates, YLDs, and disability weights for each cause.
• The data file contains the following fields:
  • Cause ID
  • Risk ID
  • Proximal Risk
  • Risk level category
  • Gender
  • Age group
  • Mortality rate
  • PAF for the cause-risk pair
  • RR for the cause-risk-proximal risk combination
  • Mediating risk factor
  • YLD per 100,000 population
  • Prevalence rate per 100,000 population
  • Disability weight

4. Validation and Testing

• The HVA model is validated by comparing its outputs with GBD interpolated life expectancy and Office for National Statistics (ONS) estimates.
• Figure 5 compares GBD interpolated population life expectancy with ONS estimates.
• Figure 6 compares GBD interpolated population HALE with HVA estimates.
• Figure 7 shows the comparison of GBD interpolated population HALE with HVA estimates.
• The model also assesses the sensitivity of HVA to different risk factors, including:
  • Alcohol
  • Smoking
  • Fruits and vegetables
  • Nuts and whole grains
  • Processed meat
  • Glucose
  • Cholesterol and blood pressure
  • Physical activity and BMI
• These sensitivity analyses are presented in Figures 13–19 and are based on risk factor exposure levels and their associated relative risks (RRs).

5. Limitations

• The model does not account for morbidity in the same way as mortality, which may affect the accuracy of HALE estimates.
• The use of mortality-based RRs for morbidity may introduce some bias.
• The GBD database provides disability weights and YLDs but does not always distinguish between morbidity and mortality RRs.
• The model is based on predefined risk factors and may not account for all possible risk factors or individual variations.
• The sensitivity testing is limited to the risk factors included in the GBD database.

Conclusion

The HVA model provides a more comprehensive health assessment than VA.3 by integrating both mortality and morbidity into the calculation of health-adjusted life expectancy (HALE). It uses data from the Global Burden of Disease (GBD) database and applies Bayesian meta-regression to estimate the impact of individual risk profiles on health outcomes. While the model offers valuable insights into health-adjusted aging, it has limitations related to morbidity estimation, risk factor coverage, and data assumptions.