Male rats: NTP (humerus), Whitford (femur diaphysis), Dunipace (femur).
Zipkin data: Regression results from crude and adjusted model, the latter assuming males and 70 years residency.
Dunipace: y = 625 + 147x (r2 = 0.97)
NTP: y = 443 + 63.1x (r2 = 0.99)
Human (crude): y = 517 + 1,549x
Human (adjusted to male, 70 years residence): y = 1,300 + 1,527x
tee also estimated two slopes for the human data, crude and adjusted for length of residency and sex. The crude and adjusted estimates are similar, barely changing the ratios in Table D-2.) These results suggest that rats require water concentrations 10 to 20 times higher than humans to achieve comparable bone fluoride concentrations.
Why are the Dunipace bone concentrations larger than the NTP results? As shown in Table D-1, the NTP study was longer and had higher fluoride concentrations in feed, but both of those factors should increase bone concentrations. The use of different rat strains could contribute to the difference. Type of bone is unlikely to explain the difference. Even if water concentrations are the same, doses might be different. The NTP study provided estimates of average absorbed fluoride doses (assuming 100% from water, 60% from feed) of 0.2, 0.8, 2.5, and 4.1 mg/kg/day for the four experimental groups. Using data provided by Dunipace et al. (1995), the committee estimates average fluoride doses of 0.042, 0.34, 0.96, and 2.83