NIOSH3 | Ranch Hand2 | |
Subjects (n) | 281 Exposed, 260 non-exposed1-150 | 989 Exposed, 1276 non-exposed1-151 |
Age (y, mean) | 56 | 53 |
Confounder adjustment | Age, race, sex, BMI, family history, medication | Stratification on dichotomized year of birth, BMI |
Analysis method | Logistic regression for diabetes, linear regression for fasting serum glucose | Frequency matched; RRs for diabetes, abnormal glucose by exposure groups |
Median serum dioxin concentration of exposed subjects (lipid adjusted) | 68 ppt in 1987–1988 (range 2–3388) | 12 ppt in 1987 (some 1992) (range 0 to 618) |
Median serum dioxin of non-exposed subjects | 6.1 (range 2.0–19.7) | 4.0 (over 10 ppt excluded) |
Diabetes definition | Self report of physician diagnosis or fasting glucose ⩾140 mg/dl | Physician diagnosis confirmed via med records or oral glucose tolerence test above 200 mg/dl |
Diabetes prevalence | 9.3% Exposed, 7.0% non-exposed (10.8%, 9.3% using glucose ⩾126 criterium mg/dl) | 14.7% Exposed, 13.2% non-exposed |
OR or RR of diabetes, exposed v non-exposed | OR 1.49 (0.77–2.91) | Not reported; RR=1.11 (0.91–1.37) calculated from data, unadjusted |
OR or RR of diabetes, categorical analysis by serum dioxin level, v non-exposed | OR 2.1 (0.8–5.7), 1.5 (0.5–4.3), 0.7 (0.2–2.6), 2.0 (0.8–4.9) for ascending quartiles of 1987 serum level | RR=0.7 (0.5–1.0) for background exposed (<10 ppt), 1.3 (1.0–1.7) for low (below median 94 ppt, back-extrapolated dioxin), 1.5 (1.2–2.0) for high (above median), |
Trend in diabetes risk with serum dioxin concentration | No dose-response trend found | Risk of diabetes increased with dioxin |
Time to diabetes onset | Not reported | Decreased in high exposure group |
Prevalence of abnormal fasting glucose, >115 mg/dl (known diabetics excluded) | Not presented; mean serum glucose did not differ greatly between exposed and non-exposed | 12.9%, 9.5%, 16.3%, 15.8% in non-exposed, background exposed, low exposed, and high exposed |
Trend in abnormal serum glucose by concentration of serum dioxin | No positive dose-response; however highest quartile of back extrapolated serum dioxin had higher glucose than the non-exposed | Not presented; RRs for abnormal glucose v non-exposed were 0.7 (0.5–1.0), 1.3 (0.9–1.7), and 1.2 (0.9–1.7) for background, low, and high exposed groups |
↵1-150 Includes women and some non-exposed subjects with TCDD concentrations >10 ppt; these were excluded in the present analysis, which includes 267 exposed and 227 non-exposed subjects.
↵1-151 It was later discovered that one of the non-exposed group was in fact exposed, leading to 990 exposed and 1275 non-exposed subjects.