Characteristics of Airway Responsiveness (Symptoms): Byssinosis, chest tightness, chronic bronchitis, cough, and dyspnea 2+ were the symptoms of primary interest. Consistency of symptom reporting was defined as follows: Subjects were “always” responders if he or she reported the symptom in both surveys; “inconsistent” responders if they reported the symptom on one of the two surveys; and “never” responders if they did not report the symptom on either survey. Inconsistent responders included two potentially distinct groups, those who became symptomatic during the 5-year study and those who became asymptomatic.
Characterization of Across-shift Change in FEVj (A FEVi Percent): Airway responsiveness to cotton dust was characterized according to the subject’s across-shift change in FEVi (Д FEVi percent) in either survey. In adjusting for differences in lung size, the change in FEV i over the work shift was standardized by the average FEVi level during the day, ie, adjusted day change=(AM FEVi—pm FEVi)-r0.5 (am FEVi+pm FEVi) Link my canadian pharmacy.com. The average level, rather than the initial level, was used to standardize the acute change in order to avoid biases that might arise from measurement error in FEVi. This potential bias is due to regression to the mean whereby spuriously high preshift measurements would be followed by apparently large negative acute changes, resulting in a spurious negative correlation between initial level and acute change.
Consistency of response was defined as follows: Subjects were “always” responders if the across-shift drop in FEVi was 5 percent or greater on both surveys; “inconsistent” responders if the across-shift drop in FEV i was 5 percent or greater on one but not both surveys; and “never” responders if the across-shift drop in FEV i was less than 5 percent in both surveys. Again, the inconsistent responders were a heterogeneous group of individuals who either gained or lost acute responsiveness.
The relationship between measures of both symptom reporting and across-shift change in FEV] was examined in relation to the 5-year decline in lung function. In addition, mean preshift FEVi was examined both in relation to consistency of acute change and chronic decline in lung function.
Longitudinal change in FEVi was measured by the difference of the second survey preshift FEV i and the first survey preshift FEVi: (FEVJ 1981—(FEVi) 1986.
Linear regression models were used to examine the relationship between symptoms and the 5-year decline in FEVi while adjusting for appropriate confounders. Cigarette smoking was adjusted for by including both smoking status (yes, no) and cumulative exposure (pack-years) in the same model. The six ex-smokers were excluded from the model.