Cases were slightly younger than controls, less likely to be African American, and slightly less likely to have started their first job before 1960, but otherwise were comparable with respect to their distributions by sex, SEER site, and education level (). This was also true for the subset of subjects assigned as unexposed or probably exposed (i.e., the subset of participants for whom we calculated exposure metrics). The distributions of NHL histologic subtypes among the cases in the overall study sample and the subsample with TCE exposure metrics were also comparable.
Selected characteristics of participants in the NCI-SEER study, 1998–2001.
We estimated 52 jobs among 45 cases (4% of all cases) and 32 jobs among 27 controls (3% of all controls) to involve probable exposure to TCE. We classified most (n = 46, 64%) of the 72 subjects with probable TCE exposure as such because they performed degreasing. Forty-one of these 46 subjects had been administered a job module asking about degreasing; of these subjects, 16 (39%) reported heated degreasing during a time period when TCE was a common vapor degreaser. The remaining five subjects assessed as having probable TCE exposure from degreasing said that they degreased with TCE in the occupational history portion of the interview.
The most common occupation categories among the 84 probably exposed jobs were mechanics and repairers (SOC 61; n = 21, 25% of exposed jobs); textile, apparel, and furnishings machine operators (SOC 765; n = 7, 8%); assemblers (SOC 772; n = 6, 8%); general managers and other top executives (SOC 121, consisting mainly of owners of small businesses; n = 6, 7%); and precision metal workers (SOC 681; n = 4, 5%). The most common industry categories of the 84 jobs were laundry, cleaning, and garment services (SIC 721; n = 10, 12%); tires and inner tubes (SIC 301; n = 8, 10%); aircraft and parts (SIC 372; n = 6, 7%); motor vehicles and motor vehicle equipment (SIC 371; n = 4, 5%); and national security (SIC 971; n = 4, 5%). The distribution of exposed jobs classified as mechanics and repairers differed significantly between cases and controls, with 18 of the 21 jobs reported by cases (n = 15) and the remaining three jobs reported by two controls (Pearson 1-df χ2 = 6.29, p = 0.01). Otherwise, the distributions of occupational and industrial groupings did not vary meaningfully between the two groups.
summarizes ORs describing the associations between different measures of TCE exposure and NHL. Cases and controls did not differ with respect to the frequency of having an occupational history that involved possible TCE exposure (OR = 1.1 vs. unexposed; 95% CI, 0.9–1.3), but we estimated a statistically nonsignificant higher proportion of cases than controls to have had a work history involving probable exposure to TCE (OR = 1.4; 95% CI, 0.8–2.4). In the analyses of the TCE exposure metrics among probably exposed subjects relative to unexposed subjects, we found that participants who had an average weekly exposure > 150 estimated ppm–hr/week was associated with NHL (OR = 2.5; 95% CI, 1.1–6.1). When we further subdivided this exposure category using the intracategory median defined by the controls’ distribution, the association with NHL for the highest level of average weekly exposure (> 360 estimated ppm–hr/week) became stronger (OR = 7.9; 95% CI, 1.8–34.3). Eight of the 22 subjects with average weekly exposure > 360 estimated ppm–hr/week had a job under the SOC category of mechanics and repairers. When we excluded subjects with such jobs from the analysis, however, the association remained (OR = 4.9; 95% CI, 1.1–22.1). We also observed an association with NHL for cumulative exposure > 112,320 estimated ppm–hours (OR = 2.3; 95% CI, 1.0–5.0), with a stronger association for cumulative exposure above the intracategory median after subdividing the highest exposed category (for > 234,000 estimated ppm–hr: OR = 3.3; 95% CI, 1.1–10.1). The trend test was statistically significant for average weekly exposure (p = 0.02) and approached statistical significance for cumulative exposure (p = 0.08). However, the associations with categories of average weekly exposure were not suggestive of a monotonic relationship; the ORs were 1.6, 0.5, and 2.5, respectively, for tertiles 1, 2, and 3 (and 0.4 and 7.9 for the tertile 3 below- and above-median subcategories). We observed a similar pattern of associations for categories of cumulative exposure (ORs = 1.4, 0.6, and 2.3, respectively, for tertiles 1, 2, and 3; ORs = 1.4 and 3.3 for tertile 3 below- and above-median categories). Overall, neither duration nor intensity of exposure was associated with NHL, although we observed an association with the lowest tertile of exposure duration (1–6 years of exposure; OR = 2.1; 95% CI, 1.0–4.7).
We observed similar associations with high exposure to TCE within strata defined by sex and age group and upon restriction to non-Hispanic Caucasians (data not shown), and in analyses incorporating latency periods of 5 years (e.g., for average weekly exposure > 150 estimated ppm–hr/week: OR = 2.5; 95% CI, 1.1–6.1; p-value for trend = 0.02) and 15 years (OR = 2.3; 95% CI, 1.0–5.8; p-value for trend = 0.03). Similarly, an association with high TCE exposure remained when we included in the analysis (and assumed to be unexposed) the 132 cases and 75 controls previously excluded because they were never employed or had only unknown occupations (e.g., for average weekly exposure > 150 estimated ppm–hr/week: OR = 2.3; 95% CI, 1.0–5.4; p-value for trend = 0.02). Our findings also did not change when we excluded subjects interviewed before the incorporation of the job modules into the interview [see Supplemental Material, Table 1 (doi:10.1289/ehp.1002106)].
summarizes findings from analyses of common histologic subtypes of NHL. Evidence of an association with TCE exposure was strongest for follicular lymphoma (e.g., for average weekly exposure > 150 estimated ppm–hr/week: OR = 3.7; 95% CI, 1.2–11.7; p-value for trend = 0.005), although the numbers of exposed cases for each subtype were small (e.g., cases with average weekly exposure > 150 estimated ppm–hr/week: diffuse large B-cell lymphoma, 9; follicular lymphoma, 6; small lymphocytic lymphoma and chronic lymphocytic leukemia, 4).
Analysis of estimated occupational exposure to TCE and selected NHL histologic subtypes within the NCI-SEER study, 1998–2001.