Table 3. Characteristics of COPD decedents and the remaining cohort members.

		COPD Decedents	Non-COPD Decedents	Non-Decedents
Characteristics		n = 4	n = 11	n = 496	p-valuea
Belong to Subcohort 1, n (%)		3 (75.0)	9 (81.8)	279 (56.2)	0.19
Belong to Former worker group, n (%)		2 (50.0)	6 (54.5)	147 (29.6)	0.26
Median age, years		49.5	45.0	30.0	0.0015
Median age at death, years		56.5	48.0		0.08
Ever worked in a high risk job, n (%)		1 (25.0)	0	54 (10.9)	0.32
Median tenure, months		17.2	23.9	9.1	0.54
	Min–Max	0.3–30.0	0.5–121.8	0.03–243.9
Median cumulative diacetyl exposure, ppm-yrs		3.29	5.34	0.57	0.07
	Min–Max	0–24.33	0.02–25.11	0–65.26
Median average diacetyl exposure, ppm		1.99	2.20	0.37	0.09
	Min–Max	0.004–9.74	0.37–2.70	0.002–9.89
Ever smoking status, n (%)		2 (50.0)	9 (81.8)	307 (61.9)	0.42
Never smoking status, n (%)		2 (50.0)	2 (18.2)	189 (38.1)
Usual coughb, n (%)		3 (75.0)	5 (45.5)	159 (32.1)	0.10
Shortness of breathc, n (%)		2 (50.0)	3 (27.3)	133 (27.3)	0.63
PFT resultsd (n = 511), n (%)
	Obstruction	1 (25.0)	1 (9.1)	24 (4.8)	0.0009
	Restriction	0	1 (9.1)	49 (9.9)
	Mixed obstructive-restrictive pattern	2 (50.0)	2 (18.2)	27 (5.4)
	Non interpretable test or no test	1 (25.0)	0	8 (1.6)

^ap-values for tests of equality of medians or prevalence comparing COPD decedents, Non-COPD decedents and Non-decedents.

^bPositive response to "Do you usually have a cough?"

^cPositive response to "Do you get short of breath walking with people of your own age on level ground?"

^dPulmonary function test (PFT) interpretation of each worker's initial test.