McGinnis (1993)^*reported the following findings pertaining to the DFAP-attributable risk factors: "… studies have associated dietary factors of sedentary lifestyles with 22% to 30% of cardiovascular deaths, 20% to 60% of fatal cancers, and 50% to 80% of diabetes mellitus cases, including 30% of diabetes deaths." These estimates (denoted as DAF_MIN_k and DAF_MAX_k) along with corresponding mortality figures obtained from VISTA database were used for estimating deaths and YPLL attributable to DFAP in Pierce County. Average attributable risk factors were defined as DAF_AV_k = (DAF_MIN_k+DAF_MAX_k )/2.

Here is an example. The numbers of deaths and YPLL due to heart disease were found to be equal to 1,299 and 2,806, respectively. The calculation suggests that on average (22% + 30%)/2 = 26% of deaths due to heart diseases were associated with DFAP. Therefore, the estimated average numbers of deaths and YPLL due to heart diseases would be equal to 1,299*0.26 = 338 deaths and 2,806*0.26 = 730 YPLL, respectively.

To estimate the number of hospital bed-days A_k pertaining to disease k and related to DFAP, it was assumed that A_k is proportional to BDAYS_k(the total number of bed-days pertaining to disease k calculated from CHARS data base), with proportionality coefficient DAF_AV_k:

A_k = DAF_AV_k*BDAYS_k.

The number of visits B_k to the outpatient hospital that pertained to disease k and related to DFAP was estimated in a similar way as it was in the Tobacco section. The difference was in choosing the proportionality coefficient: DAF_AV_k was used instead of SAF_k.

*McGinnis GM, Foege WH (1993) Actual causes of death in the United States. JAMA Vol. 270, No 18, November 10.