Seniori Costantini et al TABLEL.Characteristics of Interviewed Subiects.Frequencies and Percentages Italian Case-Control Study emia(CD:04-08) acD送：203 Cases Control Cases 586 1278 263 1.100 686(85,4 1796(712 3168324%) 1,446(76.1% SAS199 96 44 34727.2 82(312 30627 29(5.0 62(49 228.4 5752 621(48.5 756 52948 3356 6752 5954 0 0 2371404 12464 102 202 104 2102 6205 109/85 28/10 109/9j 253 90(7.0 18(68 90(8.2 65247 29526.8 2699 924 95% hown). In contrast. an OR of 2.3(95%Cl:1.0-5.4.based on 11 out of 14cases were somewhat elevated,albeit with wide data not shown). The e of inc =09-4 7)as well CTable D.Associations were stronger for duration more Benzene.xylene and toluene exposures were correlated. than 15 years for several solvents and both malignancies.The of subjects ne ne an ce in nd with of subjects exposed to benzene and also to toluene andor increasing duration of medium/high exp sure for CLL and xylene was about 33%.Risk of CLL appeared to be elevated e to toluene (P-value for 1=0.03),benzene and xylene (P=0.09).For CLL the exposure【between medium/high benzene exposure and all leukemia (OR ¼ 1.3, 95% CI ¼ 0.7–2.3), CLL (OR ¼ 1.8, 95% CI ¼ 0.9–3.9), and MM (OR ¼ 1.9, 95% CI ¼ 0.9–3.9). Risks of CLL were somewhat elevated, albeit with wide confidence intervals, from medium/high exposure to xylene (OR ¼ 1.9, 95% CI ¼ 0.8–4.5) and toluene (OR ¼ 2.1, 95% CI ¼ 0.9–4.7) as well. Benzene, xylene and toluene exposures were correlated. Considering all levels of exposure, about 67% of subjects exposed to benzene were exposed also to toluene and/or xylene. For medium/high intensity exposure, the percentage of subjects exposed to benzene and also to toluene and/or xylene was about 33%. Risk of CLL appeared to be elevated from exposure to all three aromatic solvents: the OR was 3.2 (95% CI: 1.1–9.3, based on 7 out of 12 cases; data not shown). In contrast, the risk of MM seemed to be concentrated among subjects exposed only to benzene with an OR of 2.3 (95% CI: 1.0–5.4, based on 11 out of 14 cases; data not shown). The effects of increasing duration of medium/high solvent exposures were investigated for MM and CLL (Table III). Associations were stronger for duration more than 15 years for several solvents and both malignancies. The numbers of cases and controls were generally quite small in these analyses, so that confidence intervals were often wide. Nevertheless, there was evidence of a positive trend with increasing duration of medium/high exposure for CLL and exposure to toluene (P-value for trend ¼ 0.03), benzene (P ¼ 0.05), and xylene (P ¼ 0.09). For CLL, the risk from exposure to benzene was concentrated among subjects with TABLE I. Characteristics of Interviewed Subjects, Frequencies and Percentages, Italian Case^Control Study Leukemia (ICDIX: 204-208) Multiple myeloma (ICDIX: 203) Cases Controls Cases Controls No. interviewed 586 1,278 263 1,100 % ofrespondents 686 (85.4%) 1,796 (71.2%) 316 (83.2%) 1,446 (76.1%) Gender Men 345 (58.9) 658 (51.5) 129 (49.1) 567 (51.6) Women 241 (41.1) 620 (48.5) 134 (50.9) 533 (48.4) Age (years) 0-34 50 (8.5) 163 (12.8) 1 (0.4) 141 (12.8) 35^44 53 (9.1) 146 (11.4) 10 (3.8) 124 (11.3) 45^54 92 (15.7) 237 (18.5) SAS1999 (14.4) 196 (17.8) 55^64 181 (30.9) 347 (27.2) 82 (31.2) 306 (27.8) 65^75 210 (35.8) SAS19995(30.1) 132 (50.2) 333 (30.3) Missing 0 0 0 0 Education Illiterate 29 (5.0) 62 (4.9) 22 (8.4) 57 (5.2) Primary school 286 (48.8) 621 (48.6) 175 (66.5) 529 (48.1) Middle school 144 (24.6) 295 (23.1) 33 (12.5) 259 (23.5) High school 94 (16.0) 233 (18.2) 23 (8.8) 196 (17.8) University 33 (5.6) 67 (5.2) 10 (3.8) 59 (5.4) missing 0 0 0 0 Smoking Yes 348 (59.4) 717 (56.1) 140 (53.2) 628 (57.1) No 237 (40.4) 559 (43.7) 122 (46.4) 470 (42.7) Missing 1 (0.2) 2 (0.2) 1 (0.4) 2 (0.2) Area Florence 165 (28.2) 379 (29.7) 97 (36.9) 379 (34.4) Forl 62 (10.6) 135 (10.6) 29 (11.0) 135 (12.3) Imperia 62 (10.6) 109 (8.5) 28 (10.7) 109 (9.9) Ragusa 25 (4.3) 90 (7.0) 18 (6.8) 90 (8.2) Turin 177 (30.1) 295 (23.1) 65 (24.7) 295 (26.8) Verona 47 (8.0) 178 (13.9) Latina 48 (8.2) 92 (7.2) 26 (9.9) 92 (8.4) 4 Seniori Costantini et al.