Table 4 summarizes the results for several PM10 lags
and mortality end points when estimated concentrations
of PM10, based on airport visibility, were used. For the full
four years of data, the results were similar to those obtained
using measured PM10 (Table 2). Positive and statistically
significant associations were found between estimated PM10
concentrations and natural mortality, respiratory mortality,
and mortality for those above age 50 using same-day,
three-day lag, or five-day moving average of estimated PM10,
with the five-day moving average generating the largest
PM10 effect. The three-day lag of estimated PM10 was statistically
significant for cardiovascular mortality, similar to
the result obtained with measured PM10.