Tobacco Documents Online

LOiUa na 114U 1b:4U l'r1.1 u1/US(/`J/1J SiIY: AId1VA(:Fi11%Nf 28/09 '99 09:54 FAX 01703798939 512 Etucrotn and Hcach Epidemidogy Sepcembe. 1999. Val. tO Na 5 A5_ Agt Adjusted Death Rate (DR) and Relati.re Risk of Deat6 tDR Ratio and Cox Propon;o..al Ha.zards (PH) Ratio] by Follow-Up Period for 1959 Cutrmt Cipteue Smokers Compared with Nevu Smokers ~ng California CPS I Subjcctz Fnllav-np Pctiod Papulaticu 1960-1964 1965-1969 1970-1974 1975-1979 1980-I984 1985-1989 1990-1994 1995-1997 1960-1997 Ma1es AU csuus Cucrent smoken DR (C) 19-57 21.47 18.95 ll38 16.8) 14.18 I0-44 8.87 18.68 Hcvc sma4m DR (N) 10-41 10:78 1132 8.8i 8.05 6.72 5-80 4-05 9-46 DRtauu (CIb1) PH wtlo (ON) 139 1.99 1.95 (1.87-2.07) 1.67 1.97 1.74 (L65-1.84) 2.09 2.11 1.80(1-71-190) 1-80 48 (139-1Z57) 1. 1.74 (1.6979-i.79) Lung caers G.umc smArc.a DR (C) 1.369 1.715 105 1$85 2.257 1673 1.678 ABTO 1-726 Never smdmn DR (N) 0.124 0.128 0.151 0.085 0.121 0.090 0.103 D.015 0.133 DR ratio (C(N) 11.04 D-40 10.63 22-18 18.65 18.59 1629 58.00 12.99 Ri iuro (C!N) 12.0 (7.91-20.16) 12.09 (8.12-1BA0) 1248 (B.45-18A2) 10.21 (6-12-173`3) It960.63-14A5) p.nwlre All o.w Cimcnc zmokeu; DR (C) 940 9-54 11.40 I031 8.94 489 7-48 5$0 i0.14 ~m DA (N) ~ swd 6.76 7.14 7.16 6.50 5J8 530 4.43 2.74 6.44 JNi C l L4Z 134 1 53 L 1 1.66 1-87 1b9 Z E2 157 Pli tacb C1N) 1.47(138-1371 . (1.46- 6 j L61 (1.54-1.68) L5011A4-138) 1.53 fll0-1S7) Lung GSrrmc ewuten DR CC1 0.132 0.286 0.656 1.006 1.056 l.(87 0.726 0644 0.806 Neveramokns DR (Nl 0A77 0.123 0.118 0.117 0.091 0.111 0.065 0A47 0.116 DR txro (GN) 1-71 233 536 8.60 11.60 10-69 1120 13.70 6-95 Plt ntio (CJN) 2.49 (1b5-3-77) 6.25 (4.70431) 696 (539-9.00) 7.76 (5.e1-1036) 6.13 (530-7AB) Co< l~po.6om1 t~Ns rs6uW hude.

28/09 '99 09:54 FAX 01703798939 Epidemioloyry Scpcember 1999, VoL 10 Nu. > A2. 1960 CPS ( Mal¢ 1[ev¢T $vtokes Follo.ccd m 1997 All oMmcd de 30-34 adu by attaine 1 d aEc 0 0 15-39 1 1 0 4o-q4 3 3 2 50~-54 47 35 8 55-59 62 79 60 60-64 71 93 96 65-69 111 125 151 70-74 114 136 176 75-79 110 - 139 177 80-84 90 151 175 "9 41 103 138 90-94 17 36 60 30-99 709 015 1061 0 0 0 0 5 11 65 152 159 207 209 f50 76 25 1039 T~pesau.- ~ d dxcrvation0 v¢amd :0 0 35-1439 1311 424 0 0 45-49 8078 2364 1137 348 50-54 11475 7892 2241 1030 55-59 9201 11056 1369 2031 60-64 6734 8791 10106 6588 65-69 5046 6160 7787 -9008 70-74 3530 4396 5152 6672 75-79 2362 2843 3436 3985 80-84 1106 1705 1965 2346 85-89 358 630 972 1062 90-94 67 131 235 374 95-98 9 19 25 73 30-99 52078 47663 40824 33517 Age-specific death nm (1000'tluctved desths(persoa-ycarsl u 1 \. JL1.I.iILJ1LlV 1 5`- 'AlN.'. CESSATION ANU MORTALITY T9ENDS 509 198t-:c-) 1995-1989 199'J-1994 1995-1997 1960-1997 0 0 0 1 J 0 0 2 0 0 0 B 0 0 0 40 0 0 0 96 2 0 0 226 lc 7 1 0 349 9- 28 14 2 680 18? 121 43 7 9 93 223 269 183 36 1354 2a_ 293 337 176 1633 183 237 346 234 1432 132 148 121 690 34 38 65 39 224 1051 1127 1137 615 7614 0 0 0 443 0 O 0 1735 0 0 0 1186 343 0 0 0 22980 993 336 0 0 30987 1 953 325 0 35439 6 71 1833 913 263 37181 8184 5663 1643 574 35814 5655 7106 4855 944 31187 2969 4343 5559 2871 22865 140 1810 2731 2120 11140 45i 8; 586 ~ 857 172 699 164 3402 656 28201 _-80 17046 7634 249764 Age 1960-1964 1965-1969 1970-19iJ 1975-I979 BO-84 8139 8854 89-0i 89.08 85-89 114-58 163.50 142.02 14 L.71 90-94 25352 274.0B 255.13 203.39 95-99 534.93 368-83 438.64 342.28 30-34 2.26 0.00 (100 0.N 35-39 0.76 2.36 0.00 0A0 40-44 1.2 S 2.40 5.01 0_00 45-49 324 2-96 6.16 0.00 50-54 4-10 4.43 3 57 4.86 55'-59 6.74 7.15 8.14 5.42 60b4 10.54 (0.58 9.50 9.87 65-69 22.00 20.29 19.39 16.87 70-74 32.29 30.94 34.16 23.83 75-79 5080 48E9 51.52 5i.94 Drath raze aditts*ed to 1960 1J.5- pnpdation lot anained aqes 35~4 35-Sp 10-41 10.78 11.37 8.81 0A0 0.0] 0.00 21b 0.00 0A0 0.00 1.15 0.00 0A0 0.00 1.97 0A0 0.00 '• 0.6 337 0.00 ODO 0.00 4.18 5.95 0.00 0A0 7.29 7.35 3.07 0.00 9-85 )527 1533 7.61 18.29 21.37 26.17 1;.21 26.22 37.85 37.69 38.12 43.42 65.:? 67A6 6041 11,34) 71-42 13Q: 5 126.74 127.15 110.40 12855 22i?l 225.19 172.75 173 06 202.82 403.c$ 34666 377.71 237.80 341.27 6.71 5.80 4.05 9.46

40, .4 Jy lUG lu:4V PAA utrVJrl`Jr1J 28/09 '99 09:54 FAX 01703798939 bfG A1.LivAl,LiliF.iv1 J lpJutz Epidcmioingy Scptcmhcr 1999, VoL 10 N0. 5 SMOKING CF55Al7ON AND MOftTALi7Y TRFNDS 511 A4- 1960 CPS I Pemale Nn.er Smokera Folin.red to 1997 Age 1960-1964 1965-1969 1970-1974 1975-1979 1980-1984 1965-1989 1990-1994.. . 1995-1997 1960J997 Al[ ob~ed dnd. by attalned age 30-34 3 D 0 0 0 0 0 0 3 35-39 6 1 0 0 0 0 0 0 7 4044 14 13 0 0 0 0 0 0 27 45-49 54 20 8 5 0 0 0 0 87 50-54 88 90 59 15 3 0 0 0 255 55-59 141 141 123 50 72 484 60.64 175 195 246 i30 60 32 10 0 8 48 65-69 260 279 276 229 (75 88 40 5 1352 70-74 269 409 387 316 318 297 121 28 2165 75-79 315 444 630 5ZZ 521 502 386 115 3435 80-84 276 ~ 500 632 661 659 757 731 371 4587 85-89 167 349 496 652 811 912 818 559 4764 90-94 91 148 238 370 51B 719 756 445 3265 95-99 29 42 79 148 Z13 360 408 215 1494 30-99 1910 Z631 3174 3098 3300 3674 3270 1738 22795 Totaip~-grazs of a6te:variun by aveincd agc 30-34 1439 0 0 0 0. 0 0 0 1439 35-39 5294 1430 0 0 0 0 0 0 6724 40-44 12661 5099 1356 0 0 0 45-49 27076 12436 4700 1124 0 0 0 45335 50-54 32592 26430 11411 4063 1107 0 0 0 75602 `,es 55-59 29744 31690 24021 9815 3990 1072 0 0 100332 60-64 25846 28535 27796 20648 9557 3863 1016 0 117260 65-69 22554 24260 24365 23311 19876 9213 3706 833 128118 70.74 16290 20478 19954 19713 21981 18752 8669 2476 128313 75-79 9043 L4159 16301 15502 17619 19824 17104 5572 I151L4 80-84 4515 7042 10308 11622 12465 14508 16680 9743 86883 85-89 1640 2836 4191 6208 7876 8556 10192 7377 48876 90-94 460 679 1336 1934 3142 3933 4323 3331 19137 95A9 80 l54 209 393 566 1018 [033 879 4391 30-99 189237 175227 145946 114334 98178 80737 62784 30210 896654 V psific dead6 na (1C00'ohscrvcd dcadu/prrwn-ycara) 34 Z-08 0.90 0DO 000 0A0 0.00 0 00 0.00 L08 35-39 Li3 0-70 000 0.00 0.00 0.00 0.00 0.00 1.04 40-44 1-Il 2.55 0.00 0.00 0.00 0A0 0.00 . 0.00 1.41 45-49 1.99 1.61 1.70 4 45 ODO 0.00 0.00 0A0 1-92 50-54 2.70 3.41 5.17 3.69 I.71 0.00 0.00 •.0.00 3.37 55-59 4.74 4.45 5.12 5.09 5.51 653 0.00 0.00 4.82 60-(M 6_77 6.83 8.85 6.30 6.26 81$ 9.84 0.00 7.23 65-69 11.53 11.50 1133 9.82 8.80 955 10.79 6-00 10.55 70-74 17.74 I9.97 1939 ] 663 14-47 15-84 1396 11.31 1687 75-79 3483 31.36 38.65 33.67 2957 25.32 22.57 20.64 2934 80-84 61.13 71.00 6131 56.88 5287 52.18 43.82 38.08 52.80 85-,R9 101.80 123.04 11836 105k3 102.97 10639 80.26 75.77 97.47 90-94 197.91 217-98 178-ZO [ 9l 27 i64.84 18293 174.87 133.61 171.65 95-99 363.72 273.44 378A6 376.38 37661 353.67 373.29 244.69 340.27 Dea,F nre adjucred m 1960 US.populscion (« atlained ages 35-84 35-84 6.76 7-14 7.16 650 538 530 4.43 2-74 6.44

- ~•I a yy i,i 1o:en rAn ot,~orearla .... U11. Mr4~rll.hDiC~~1 011 28/09 '99 09:54 FAX 01703798939 -1Q ~ 510 Eeutram end Heorh A3. 1960 CPS I Female Current Ggevctie SmokcrsFo{lowed to 1997 EpWeeialogy September 1999, Vol. 10 \u. 5 Epi A4 Age 1960-1964 1965-1969 197t11974 1975-1979 1980-1984 1985-1989 L990-1994 1995-199i 1960-1997 All obnerved dotha h attzicw3 a¢e 30-34 0 0 0 0 0 0 0 0 .' A 35-39 5 1 0 0 0 0 0 0 6 4054 29 6 3 0 0 0 0 0 3> 45-49 78 51 28 3 0 0 0 0 16C 50-34 109 118 82 27 2 0 0 0 338 55-59 112 L73 191 73 34 12 0 0 593 60-64 114 144 236 252 96 56 1L 0 909 65-69 83 ' 136 232 265 300 194 62 12 1784 70-74 52 106 197 28l 394 461 189 51 1731 75-79 33 87 158 Z26 402 563 553 184 7206 880-64 24 56 101 173 ' 249 423 675 385 2086 85-39 6 iB 48 95 143 275 389 332 1296 90-94 2 9 11 43 73 109 l46 126 519 95-99 2 1 4 4 78' 30 40 35 134 36A9 - 649 906 1291 1432 1711 2173 2065 1125 11302 Tool ensenryeas uf o}nc[vauon DY sttained age 30-34 L398 0 0 0 0 0 0 0 1396 T< 35-39 5273 1352 0 0 0 0 0 0 6625 40-44 12779 5046 1278 0 0 0 0 0 191m 44-49 25463 123L9 4573 1039 0 0 0 0 43395 50-54 '75287 74464 li1[6 3859 1029 0 0 0 65755 55-59 15753 23932 21725 9281 3714 999 0 0 75404 60-64 8597 14589 Z0440 18098 8828 3482 9I1 0 74953 65b4 4683 7759 12055 16877 16814 8093 3164 733 70176 70-74 2389 4060 6008 9422 15204 14924 7t54 1991 61133 75-79 983 1950 3073 4390 7768 12743 IZ327 4175 47409 80-84 252 737 1272 1994 3096 5705 9531 6210 28797 85-89 70 140 456 773 1167 1825 3526 3396 115.] 90-94 13 2B 51 215 308 480 760 921 2778 95-99 3 . 4 1 10 62 84 111 110 386 30-99 102944 96380 82047 65908 57989 48336 37495 17737 508837 Age-specific dcarh nre (10t70"cbsenxd deathslp ,soo-yeanJ 30-34 0-tC' 0.00 0D0 0.00 0.00 0.00 O.W 0.0C 0ZI AI 35-39 0.95 0.74 O.CO 0.00 0.00 0.00 000 . O.CO ~.91 40-44 117 139 2.35 0.00 0.40 0.00 0.00 0.00 1.99 4549 JR6 414- 6.12 2.89 0.00 0.00 0.00 00.~ 169 50-59 4.31 4.82 7.38 7.00 1.94 0.00 000 0.00 5.14 55 59 7.1! 7.Z3 8.79 7.87 9.16 1Z01 0.00 OOC F39 60-64 1316 9.87 11.55 13.92 10.88 16A8 11.95 0.0C 17.13 65-69 17.72 1753 19-25 15.70 1784 23.97 1959 1637 I:3Q 70-Y4 21.76 26.11 32.79 29.82 25.91 30.89 Z6.42 25.62 :&31 75-79 33.58 44.6251.41 51.48 51.75 44.18 4486 4}.07 46.53 80$4 95.13 7598 79.40 86.78 80.44 74.14 70.82 61.99 72.44 85-89 86.32 128.86 10536 11759 12256 150b7 11031 92.33 1 O3i8 90-94 l53&? 316.10 217.54 I99.St 23681 227.01 192A0 136.75 1Ffl26 95-99 650.04 170.42 2708.Z9 38Z01 Z89_46 355.85 360.00 318.4'- 3q,.0i D.d, a[c adj..sad m 1960 U.S. pop,Wtion (or aaainei xgc. 35-84 35-84 9.60 9.54 11.40 1037 8.94 9-B9 7.48 5.80 NMi Om

2ri/U9 'J9 LLh 1tl[3tl PAd UliU31l971J 28109 "99 09:54 FAX 01703798939 508 FnsROm and Heaazh i3'1'l: (4.11v.iluEDlt•,1v1 Appendix Al. t960 CF'S 11.Oale Cuerenx Cig.vecte Smoken Followed to 1997 Epideminio¢y Sep.embcc 1999, Vol. 10 No. 5 EF A Age 1960-1964 1965-1969 1970-1974 1975-1979 1980-1984 1985-1989 1990-1994 l995-SD97 1960-1997 ~~d- ~26yx~~0 0 0 D 0 0 0 7 35-39 11 3 0 0 a 0 0 0 14 41Y44 25 - 16 1 0 0 0 0 0 42 45-49 105 28 18 4 a 0 0 0 155 2 50-54 238 179 47 - 21 7 0 a 0 4Y 55-59 274 384 236 47 28 11 0 0 980 6(1-64 285 428 486 315 95 35 9 a 1653 65-16 270 367 443 557 370 88 37 30 2142 70-74 193 335 409 556 651 . 438 89 39 2710 75-79 107 186 289 388 654 753 509 90 2975 B0-84 49 121 174 235 359 537 706 337 Z518 85-89 16 35 64 106 159 261 372 268 1281 90-94 4 3 14 27 42 70 111 85 356 95 99 0 1 2 5 11 1) 23 15 70 30-99 1579 2086 2182 2261 2376 1206 1856 844 15390 Toral~ yeaa ol oFs.mricc by amined~e 30-34 821 0 0 0 0 0 0 0 821 35-39 2738 755 0 0 0 0 0 0 3523 40-44 5042 2565 747 0 0 0 0 0 8354 45-49 (8064 4775 2357 646 a 0 0 0 25843 50-54 23877 17100 4334' 2102 613. 0 0 0 50025 55-59 19134 23944 15532 3772 2006 577 0 0 64965 60-64 t1288 16867 20650 - 13227 3417 1830 519 0 67798 65-69 6407 9407 13846 - 17090 - 11490 3002 1658 411 63312 70-74 3336 4813 7052 1072D 14077 9435 2354 976 52962 75-79 1407 2294 3228 4785 7731 10514 7034 1371 38364 8054 438 930 1278 1821 2959 4723 6762 3452 22270 85-89 57 175 149 538 842 1386 7780 2085 7811 90-94 11 21 57 103 172 Z63 437 :546 16t0 95-99 0 1 t0 10 25 43 60 '56 214 30-99 94620 83586 69440 54824 43332 31772 21404 8897 407874 A(~~ci8c ;k eh noc (10Woixe;vW dndss/pc~somt'ealsl N 2.44 0.0.1 D.00 0-00 0-00 35-39 4.02 3_82 0.00 0110 0.00 A"1-44 4.96 6.24 134 0.00 0.00 45-49 5Bt 5.86 7.64 6.19 0.00 30•54 9.20 1047 1034 999 11.42 53-59 14}2 16.04 15.19 12.46 13.96 60-64 25.25 25.37 23.54 Z3.81 77.80 65-69 42.14 39.01 32.W 3259 32.20 70-74 57.86 69.61 SB.W -51_87 46.25 75-79 76.03 81.09 89.21 81.08 84.60 80-84 112.00 144.43 136.13 129.0) 12134 85-89 28156 200.09 183.37 197.14 186.80 90-94 364.62 141.49 243.76 Z6189 244.15 95-99 000 717.56 200.47 264.48 439.69 C.d. n.e adjmccd m 1960 US.pap~ fauon far ac~a'sed agu 33-89 35-84 19.51 21.4"! I8.95 1738 ¢6.83 O.W O.W 0.W 2.44 6.00 0.00 0.C0 3.97 a00 O.M 0.00 5.03 O.aO 0.00 0.00 6-W 0.00 O-W 0.~0 9.83 19.Q7 0. W 0.00 15.08 19.13 1733 0.00 2436 29.32 223Z - 2431 31.83 46.42 3485 39.94 51.17 71.62 7236 65.63 77.55 113.71 104-4I 97-63 113.07 l88.27 136.32 12a-57 163.99 266.20 251.91 155.81 221.06 301.86 383.50 Z68.62 326.84 14.18 10.44 &87 IB.68 V/y6s kr 060 USa p.pr,lumnfx e¢es 35-84 in 3-yea sge @e.ps 35-39.0.t74; A0-44.0.t15; 45-4? 0.14V, 50~4:0.155; 55-59-0.11t; f0-bk 0.096; 65-69: 0.080; 7u-74:0061: TS-79:00)5; 90-84:;7.0i1 I wo

_,~...., sb 1VL: 1n:Jl rt1A u1rUJrIyr1J 28/09 '99 09:54 FAX 01703798939 Epidemiology September 1999, Vol. 10 No. 5 SMOKING CESSATION AND MORTALI7Y TRENDS 501 TABLE I. Cigarette Smoking Trends annng Califorma CPS I Subjetta and U.S. Whitea by Age As of December 31, 1959 Agc u Yetemtaee of Cunent Cigarette Smoions of Males Femzlez 1959 (Yms) Population 1959 1965 1972 1994 1999 1959 1965 1972 1994 1999 30-34 CPS 57 45 36 0 6 44 39 32 0 5 US. 14 67 56 16 49 47 43 16 35-39 CPS 52 45 32 8 3 45 40 35 0 5 Us 7Z 64 52 13 46 44 39 12 40-44 54 31 0 2 44 40 32 2 3 US ~ 7 0 61 49 10 45 43 36 9 45-49 CPS 51 40 26 2 1 42 37 29 6 2 U.S. 66 59 43 7 40 38 31 6 50-54 CPS 49 37 24 1 29 22 0 1 53 4 . 32 29 73 2 55-59 CPS 44 20 25 2 27 21 IS 0 0 U.S. 57 46 30 2 23 10 15 1 60-M CF5 36 27 14 l9 14 lo 48 37 22 17 14 10 65-69 CP8 20 10 14 10 6 U5. 40 31 19 11 9 6 70-74 CP3 26 - 17 6 Il a 4 U.S. 33 24 8 8 6 5 75-79 CPS 21 12 3 7 5 3 80-a4 CPS 13 10 10 4 1 0 85-99 CPS 7 6 0 3 3 0 Torel C15 46% 34% 23% 4% 2% 32% 28% 23% 2% 3% (All resyadrns) (51,343) (44,990) (26,255) (114) (2.252) (66.751) (61,236) (36.083) (167) (4,796) Pe,cenng a(ew,ent cira,ane smakets amonr CPS 1 wblreu art 6azACn rtgimaes xo ehe wyhu( 1959 puesri«wee aed rn d,e fallev.up qunriorm.:e iu 19E5. 1sT1. 1994. a,d 1999- Tcra{ numkcr d(8S I rtspondmn uc shown (n pannd. Pmeneage am".,e 1I5. ".hins - 6a-d w. mc".el s.ren.,- heaith at the rime of cessation do noc have lower death rates than cutrent smokers. These limitations are gener- ally not mentioned when stating the benefits of cessa- tiOn. The most rigorous way to evaluate cessation is to assign smokers randomly to either a cessation interven- tion ot no intervention tn a randomized controlled trial (RCF). Given the evidence ftorst nonexperimental stud- ies, it would be expected that in an RCT with a smoking cessation intervention, smoking-relaced mortality would be fowec in the intervention group than in the control group. Therc has been just one completed RCT designed to evaluate smoking cessationn aione: the Whitehall Civil Servants Study of 1,445 middle aged whice men in London.' There has been one RCT where smoking ces- sation was the major risk factur change during 6 years of intervention: the Multiple Risk Factor Intervention Trial (MRFIT) of 12,866 middle aged white men in America 61-here is ~ncly one ongoing RCT designed to evaluace smoking cessationt the Lung Healrh Study (LHS) oF 3,702 men and 2,185 women in the United States and Canadaj lnitiatly these RCTs showed substantially moa smok ing cessation in the intervention group relative to the control group (about 45% tn 20% in MRFIT averaged over 6 years, about 60% vs 25% in Whitehall averaged -ovet 5 years), but the cessation di(ferences diminished substantially over time- For the intervettian groups rel- ative to their respecrivc control gruups, the total moe- rality during 16-20 years of follow.up was 6% less in MRFIT and 7% less in Whitehall; lung cancer mortality was 17% more in MRFIT and I1 % less in Whitehall. No differences in total and lung cancer mortality were ob- ult, einhebED1b1~1 served m the first 5 years of fal(bw-up in LHS. A sy:- tematic review of nine RCTs shows rhat mulufactotial interventions, including smoking cessation, resulted in a reduction of only 3% in total morrality e So the RCts have not definitively conftrmed the value of smoking cessation, but they are limited by tlhe fact that they were not able to sustain a large difference in smoking cessa- tion between incervencion and conaol groups. We provide additional insiglu into this issue in cwo ways. First, we analy:e the conventional observational data on the benefits of cessation among a large cohtx-t of Califomiaas, making tse of repeated measurements of smoking status and extended follow-up noe previously analyzed. Second, we evaluate the "natum) experiment" of long-term smoking cessation that has occurced among this cohortc Here, the amokers as a whole can be con- sidered as an intervention group thar has undergone extensive, sustained cessation and the never smokets as a whole can be considered as a control group diat has never smoked. The smokiag-related death rates of both i groups would be expected ra converge as a eonseque.nce ~ of smoking cessation. i Although a"natural experiment" does noc invol..e randomization of subiecrs and is not nearly as rigorous as an RCT, infetences about etiologic relations derived ftom such situations are stmnger than inferenccs derived solely from an absctvational aady.°'[he study of British physicians is a prime exampleg~l+ n; appmadt.r0 as is out somewhat similar scudy of CaG'dwtna physicians.tl 7he British study involved 34,990 physicians who substaa- tially reduced their cigacacte smoking from 1951 to 1971 and whose lung cancer dearh rare declined relative to the general popellationtate and convecged towatd that p 002

28/09 '99 09:54 FAX 01703798939 Lly ~~r~syis`i'°i UuU4 Epidemiology Sepceznber 1999. VoL 10 No- 5 9MOKING CESSATION AND MORTetUTY 1REND5 503 TABLE 3. Age-Ad)usted Deat6 Rate (Deaths per 1.000 Person-Years for Ages 35-84 $mdardire3 to 1960 U.S. Popu- lation) during 1960-1969 for Feemer Cigarette Smokers by Length of T'mx Quit Compared with Current and Never Smokers among California CPS I Subjects - Males Fenrales 1959 Smoking Srsrut Num4r of Sub)ecn All Ca,aa L.vng Cazxer Number of Sobjeca All (Surzs Q.nmr (as of 1959 and followd 1960-1969) 19,899 1 21.137 9-54 E3a6041 13071 [1•5171 Farmer (quir <1 yeu aa of 1959) Fu9o-ed 1960-1961 1 137 28 239 6 693 IS.ZB , ~81 ) E I121 Survi•ed rlunugh 1961-ard fdlowed 1962-1964 106 1 21 47 1 7 681 00 12 . . ~] . [ 21 Sunived duwgh 3964.ad follow.d 1965-1969 1 04D 16.47 1 Z 663 7 1 , 19) 1Z91 pncmer (+. of 1939 x,d fullwed 1960-1969) Qaic el rear aa of 1959 1.137 18.68 (190) 1.651 [18) 693 9(A~17 Quir lly ycya w. af 1959 2.291 1631 1.135 t,27B 9-10 Quir 5-9 yeazs as of 1959 2 ¢67 1 0 1 032 k0 , ~~) {~) , ( Quk 10-19 Yeza s of 1959 2 535 1 .5 8 D 2 1111 BSB , ~ ~ ) `I 01 l Qu1t 20* rearr ss of 1959 2 149 10.69 0. 675 5JS . ~81 13z1 N~ fas of 1959 a,d fol7nwmd L960-1969) 10470 i051 0127 39,218 6.95 11.414) 1181 [3,712) N.,nbe, d wbir<a af.4 ~ .nd m,mb. o(dodu W 1,h .ae~d zen 35-85 I I..c aMvn in ihu and u,Le, ubin. Follow-up was done without knowledge of the smoking stauu of rhe subjects. To confirm that cohort members are still alive, dlose not known to he dead were matched by rhe Califomia Departrnenl of Motor Vehicles (DMV) widl California driver's License information. Matrhes were established based on name, sex, date of b"uth, height, weight, and spouse's address. Other sources of folE.ow-up have been used extensive[v: the Social Security death index, voter registration records, and national name and address files. Suhjects are considered to be alive if they have a valid driver's license match and are rsot known ro be dead from any source. The drivers' licenses for most of these macches were lasr renewed in 1995-1998. We believe most deaths chrough 1997 have now been identified through the above sources. Thu.• we selected a cutoff date of December 31. 1997, representing 38 years of follow•up. As of this date, among 51,343 males, there were 36,274 (71%) deaths with underlying cause known, 2,451 (5%) deaths wirh cause unLno.w-n (uroatly outside of California since 1972), 8,023 (16%) known alive dtraugh DMV records, 1,395 (3%) not folbwed beyond 1972, and 3200 (6%) lost as of 1997. Among 66,751 females, [here were 35,183 (53%) deaths widh underlying caeue knawn, 2.030 (3%) deaths with cause unknown, 16,552 (25%) known alive. 5.449 (8%) nor followed beyond 1972, and 7537 (11%) lost as of 1997. To determine recent smoking patterns in the cohorr, a mailing was done in 1994 co random samples of 200 houceholds where both husband and wife were alive and 100 households with a single [emnle using driver't li- cense addresses obtafne I from the DMV. A rotai nf 114 men (57%) al.d 167 w.men (56%) completed a iimple one-page smoking questionnaite_ A second question- naire was mailed in 1999 to 6,354 men and 12,918 women with a dsivcr's license addless a9 of 1995 or later. Preliminarily, a total of 2,252 (43%) men and 4,796 (45%) women completed a two•pagt smoking and life• style questionnaire, out of 5,275 men and 10,73 8 women who apparently received the quesrionnaire: at ]easr 3,259 questionnaires were undeliverable owing to an incorrect address or recent death. Al.so, comparison is made with recent US- smoking surveys by birth co- hort'a To determine the mortality impact of smoking cessa- tion, we have used a survival ptogmm developed by Monson to derermine pecson-years of follow•up, deaths and death rates by attained age, and age-ad'Nsted death rates for attained ages 35-84, standardiud to the 1960 U.S. population in 5-yeaz age grasps.19 Also, we used the . Cox proportional hazards model to evaluate the age- adjusted relative risk of death as a function of smoking srattu and time period of follow-up_1D Current, former, and never smoking status has been detetminerl initially- as of the 1959 questionnaire and then again as of the 1965 and 1972 questionnaires. Current smokets are de- fined to be those persons who smoked cigaterces only. Former smokers in 1965 are defined as those 1959 smok- ers who did nor smoke in 1965; former smo)cets in 1972 ate defined as dtose 1965 smokers who did not sm, ke in 1972- Never smokers are defined to be those "ars who never smoked any fonn of mbacco. Age at entry is defined to be 30 99 yeara as ~d De- cember 31, 1959. The fi•Ilow-up period is from rime od entry into the study (Jnni.uy 1 to March 31, 1960) ur.cil death, withdrawaI (dare last kn.,wn alive),o end a folkrw •,p (Dec ,mber 31 19'97). Yetsonryeaa e f ebser J 1

28/09 '99 02:54 FAX 01703798939 502 Faxrmm awf Heath tl11. ill~i,V~ar,LdLi~~ Epu3emiology Sepcember 1999, Vol. 10 No. 5 TABLE 2. Numher of Subieces Foflowed, Percentage Followed Only w September 1972, Pv,-tnlage Lost As of December 1997. tdumber of 19fi0-f969 aad 1.960-1997 Person-Ycwrs for Cnrxnt Cigaactte Sntokexa and Ne.ror Smokers atnong Cafifornia Cp615obkcn _ l,da16 Feordo f ~ N Fvllov-up (%) Peam.Yean d Ot~uivatiw. h f N Folln..-up (%) P r5ln YtaS of Qhseevnelon umber o A l9 9 Su6isra {Yearsl Fd1Rw<d Onlq ro 197Z lar u d 1997 1950-1969 1960-1997 er u wc. Subjecrs Folln.td t7nly tv 1972 Lost as d 1997 1960.1%9 1960-I997 C.urenr eiguene smdten 30-34 410 6.1 123 821 1121 649 171 13.4 1,398 1,398 35-39 70i 6.l 109 3$23 3,523 1.661 9.9 13.4 6b25 6,625 40-49 1,593 5.1 9-4 7,608 8354 51 9 45-49 5,560 3.0 7S 22940 25,843 ,997 . 9-5 12.5 37,782 433 5 50-54 4,821 2.9 65 42,976 50,075 4,194 - 95 12.6 49,751 65,755 55-39 3,153 39 6.4 43,078 64,965 2,406 8.4 14.8 39.685 75.404 " ~ A i5 i 1,276 B.6 7.4 23,186 74,953 6 5~ i,0 56 22 2 a 5 . 63,312 715 7-0 7.3 12,443 70,178 y0-74 527 1-9 1.1 8,148 52,962 346 510 4.6 6,450 61,153 75-79 215 05 0.0 3,701 38,364 116 0.9 1.7 2,933 47,409 80-89 44 0.0 0.0 1,275 22,270 30 0.0 0.0 969 2a,797 85-99 7 QO 0.0 765 9,635 10 0A 01) 257 14,666 Ton1 19,899 33% 6.5% 178,206 407,874 21,137 9-1% 12.2% 199,324 508,837 Nnxrsmoken 30-34 198 8.5 15.5 443 443 681 10.6 13.6 2,439 1.439 35-39 353 3.7 73 1,735 1.735 1,615 8.8 11.4 4,724 6,724 4tt-M 693 2-4 112 3,659 4,058 _3,820 7.7 1L8 17,759 19,i15 45-49 7,45? 2,6 8-4 10,392 11,576 Q610 85 il-9 3R5t1 45,335 50-54 2,11' 1.9 24 19,367 22,980 6346 8. 7 13.2 59,021 75,602 55-i9 1-662 23 73 20J57 30y87 5.665 7_7 16S 61,434 100,332 60-64 1,199 1.8 44 15,525 35-439 5,006 8.1 9.6 54381 717,260 65-64 911 2.4. 4.0 11.206 37,381 4,180 7b 6.6 46,814 128,114 7h74 668 0.6 13 7,926 35.814 2,630 5.3 5-7 36,768 I28,513 75-79 393 03 0.2 5Z05 31,187 1,473 0.8 0.6 23,202 115,124 80-84 16' 0.0 0.0 2,811 22,865 680 0.5 0.1 11,556 86983 8i-99 59 1.7 0.0 1,214 15.198 310 03 0.0 '. 5$19 72,404 Teal. 10,870 21% 69% 99,741 249,764 39,218 7.5% 30.8% 446t 96,654 All whjeon Twt 51,343 2.7% 6.2% 462397 1.093,429 66,751 B1% 11.3% 623,552 IS63,058 T9 x+a ak e ut pRe..le, 31. 1959. x,d aaeEdt, sanu u ddiont im ¢xe of non:mokets during those 20 years. The major rttotral- iry• 6enefits of cessation were among smoking-related causes in physicians under age 65 at death, with no overall benefit for ages 65 and aUove:'r Methads The original ACS Cancer Prevention ^.Study (CPS I) of 1 million Amenicans from 25 states is a prusp4ctive epidemiologic eohort study begun in October 1959. It is described in detail elsewhcre.12-t4 CPS I used 68,00Q ACS vohtnteec researchers to distribute and collect quescinnnaires fiom ftiends, relativee,and othets with whom they would.expect: to be In contact fot 6yeats. A rocal of 1,078,000 men and womenn were enrolled. in-cluding 118,094 Californians (51,343 alenand 66;751 women). The fau:-page confideatlai questionnaire con- tained data on many facmxs svspected as being related to cancer, such as, family histoty of disease; medical his- rory, smoking, drinking, and diet; use of dn>gs and med- icines; occupational history; and environmenmleltpo- sures- In 1961, 1963, 1965, and 1972, survivutg subjects were asloed to complete a shorc questt«maire regarding changes in smoking hablts and detailed somroaries of the s.noking changes after 6 aad 13 years have been pub- lished.'x'6 The subjects were followed axulually for 6 years, +aith 99% t:aced chrough September 1965, Follow-up was begun again in 1971, with 93% traced through Septem- ber 1972. ACS secured death crrttificatrs from state health departme-nts for ehose known dead. Th.ere werc diffieulties in distributing the 197Z questionnaire and only 75% of sutvivors returned it. il(pst follow.t3ps were teqqf{sated in 1972 becatlse ttacFng became increasingly di4(`jet3lc owing to movement atgl/or death of the te- seirchers and subjects. Follow-up was conunoed through 1987 for the 5% of subjeets bnm befote 1887!7 Afcer a 1990 feasibility study showed thhat it was possible to successfully follow most of the petsotss in a 03% sample beyond 1972, follow-up was urrdettaken on al1118,094 Califotnia C:PS I su161ecrs, e><cluding 22,210 deaths previoualy found by ACS and 6,844 pusotu with an incomplece or missing name who could not be fol- iowed beyond 1972. To identify new deaths, the CPS I subjects were macehed several Ciraes with the I960- 1997 California death file. We evaluated death tnatchcs baaed on the following variables available on most sub- jects: lasr name, fm:t name, middle initial, marixa] statLs, spoust's initials, sex, tace, birth year, birth month, cou¢- rry of birth. county of cesidence, and initial health status. E; Fc sc te an ca 15 I, ]o 3! 2. kr 7!

-o.oa os ...n 1o-ov r,L~ oleuSttaela b1C dlA,SAGL•ML•A1 28/09 '99 09:54 FAX 01703798939 to 005 i04 Ersscrom and Heath Epidemiology Seprember 1999, Vol- 10 No. 5 TABLE 4. Age-Adjusred Death Bnre by LYear Fallow-llp Pertods for Current snd Former CqPretm Smokers (.ompsetd witl, Never Smokers vmong Califoraia CPS I Subjects for All G.vses, with Smoking SmNS Deteirttined As of 1959, 1965, and 1972 Questionnaires Males Females f N b Deedi Aeee by Fdlw.-l>p Pe:,od N b f Death Race by FoBnw-Up Perial Smking5ntus er o um Subjects 1960-1961 1962-1963 nm ec o 1964-19658ubqecos 1960-1961 1962-1963 1964-1965 1959 (:Lncnt 19,899 17.61 15101 19.17 1656) 21,63 ] 21.137 [85~5 R~ I~I3521 Fovner fqar P6 vwsl 773 4 1655 2 526 143 2 96 8 7-14 . 17471 11611 , ~ 7 ~ 63 . 6 6 731 10.870 9.02 11.16 87 39,218 10. 9 f206] [Z81) (Z991 [5861 1656) 17711 1966-1967 1968-1969 1970-1971 1966-1967 1968-1969 1970-1971 Cunxnr 12,366 20.05 20.61 1958 15,244 937 8.19 11.30 ~ ~ 2 ] 1601~ 12381 Cd641 P781 Fwmet (vit 0-S reaa 4Z24 P 9 D 0.44 17.i 3,381 11.02 7.72 9.91 aierc 1959) (Z181 (2201 [1941 [801 l70] t84) Nevec 9,330 10.7E 1034 12-32 34,052 6.90 7.02 1.53 [2781 1292] 13271 R43) 18171 [9521 1972 1973-1974 1975-1976 1977 1978 1973-1974 1975-1976 1977-1978 G>ne.u 4,892 18.74 20.14 19.23 6.122 9.96 10.00 12.71 69 6 5 13251 [152] (1871 (2491 Fermer (quit 0-6 yean 2.fl98 0 I 15 56 13.95 Z,I28 10D2 I2S4 9.18 sm e 1965 ~ 11891 4167] [65) 1801 176] Ne.<r . 5.435 0 8-6 9.63 17.553 5.71 6.15 6.62 11921 12361 12401 142131 1*eN f534) aarion and deatlu in 1959 were excluded. Petaons living past age 100 areassumed dead at age 99-Causea ofdcarh. as defined by the Ninth ILevision of the International Classificarion of Diseases (ICp9), are limited hete to lung cancer (1CD9 = 162) and all causes. Causes of deaths before 1979 under the Seventh andEighdt Re- i-isions have been reclassified to the Ninth Revision- Similar methodology has been used recently for 19-year follow-up of a U.S. sample of males and fetnales and 26-year follow-up of the U.S. male veoerans cohorr." Results The cigarctte smoking habits of theCalifornia CPS I cohott in 1959 have been described and compared with follow-up smoking dan.fr~om-1965, 1972, 1994, and 1999 qtr Table 1. Based on surviving subjecr5 with complete smoking historfes, dte percentage of cigaretce smokkea decliaed from 46% in 1959 to 3% in 1994/1999 among maks and from 32% in 1959 to 2% in 1994/ 1999 among femafes. The derlines were . more rapid among the older subjeca- These data demonstrates the high degree of smoking cessation among t6is cohort and excend previtws.findings.'sde These ftodings agree qua[- itanvely with the cessation dara for US. whiresie as shown in Table 1. Evea if die 1994 and 1999 CPS 1 surveys were inacenrace. the aaCtonal saveys.indicate chac the 1994 smoking preevalence among those~.bora during 1900-1929 (aged 30 59 in 1959) would aveta<ge less than 10%.'s Although the 1994 and 1999 suru•eys are based on selected samples, the baseline characteristics of the re- spondents and notuespoadenrs are similar. Thus, these two su:veys should be reasonably reprtesenra[ive of re- cent survivots. The main finding ln both is tbat there has been.ajmou tntal cessatitm among the respondents. Of the men who smoked cigaretces-In 1959, 5I%smoked in 1972 and-7% smoked in 199411999; of the. women who smoked in 1959, 66% smoked in 1972 atrd 7% smoked -m 199411999. Among other respondents, chere Is good agreement between smoking starus reported in 1959 and 1994/1999. Of die former smokers in 1959, IOfN% of males and 85% of females were former smokers in 199411999, w(th the odler females reported to be never smokers. Of the never smoke[s in 1959, 94% of ma]esand 93% of females were never smokers in 1994/ 1999. Table 2 shows the number of subjects by age as of December 31, 1959 and -1959 smoting sratus, thee per- ceniage followed only to 1972, the petcenrage lost as of 1997, and 1960-1969 and 1960-1997 person-yeats of observation. The percentage wirhd[awn from follow-up at each age is independent of smoking status and de- cllineswirhh age. The person-yeats of observstiotr during 1960-1997 have been 1,093,429 among men and 1,563,058 among women, making this the largest cpide- mialogic cohort followed for 38 years. Table 3 shows the beneAs ofsmokmg cessarion based on the midiflonal method of examining self-selected former smokerswho had quit for a number of years as of the beginning of the study? The death rates farformer cigarette smokers compared with never smokers have heen calcuLared duringrhe fol]ow ap period 1960-1969. The death rates for rhose who had quit for <I and 1-4 yearswere close tothe death rates for cucrent stnokets, whereas the death rates for faamcr srnoken who had not smoked for 20+ years were close to the death tates fot n. h. D[ .~a T r. tn I Nw

28/09 '99 08'54 FAX 01703798939 Epidemidogy Sepcember 1999. Vol. IG No. 5 ull. n1:L\:LbL•aLLG,Yl SMOKING CESSATION AND MORTAUTy TREsVDS 505 TASLE 5. Agc-Adiustcd Deadt Rate aod Aate Difk::reoce and Re6ti.e Rixk o4 Deada (Rado of AyE-Adjmud Dnth Rates and Cax Proportional Hanrdc Rafio) by Follow-Up Period for 1959 Cutrent.Cigarctrt Smokers Compared with Ne.er Smokers among California CPS I Suhjects . Csuse of 73ead+ and Fo11as-Up Pcriod N M f All Camu Lung Cancv 1959 Smoking Sv. x o w¢ S.,6jem 1960-1969 1960-1997 c960-1969 1960-199? Male. (ag~-edfustd run) AB melecu (T) 51.743 15.30 14.13 0.793 0.888 (8,519) I30,856) 14891 f1.8931 Cuumt vnoken (C) 19,999 20.67 1&68 lsse 1.778 ` ( 65 1 [13,6811 (3071 11,3051 Nwer smo4us (N) 10.870 10 i 9.46 0.127 0.133 (1.4141 1 71 ~ 1 9 ~ tlifl'. (C-N) 16 10 9 22 1 43 15 5 Reur. . Age9d/mrrd sus ( ) 1.97 (1.8~-L09) 1.37(i.9i-2.03) 1217(7b8-19S9) IZ.99(10.46-16-13) ptopomo,al haaMs C/N) . 1.95 (1%3-2.07) 1.74 (1.69-7.79) 12.63 (7.91-2R16) I1-96(9.63-14.65) Femzlo (spo-sd"wste.l stes) All swb)ac¢ ({~ 66,751 7-7z 1 7 7.70 0,138 0339 Q.neut sebkc~ (C) 21,137 9 ~ (10747 Q~ 0~ 11,5171 19,353} [52) (7431 Ne+c amokcrn (N1 39,218 6.95 6.44 0.094 0.116 13.7123 (13.2491 1461 (1971 Dtf(emKe (C-N) 2.59 3.70 0.114 0590 (la io Agc-adiumcd rue+(C/N) 137 (1-29-1.45) 157 (154-1.61) 221 (1.46-3.35) 695 (6D1$04) PropmtWaal hamrtls (GN) 1.47 (138-1.57) s 5.53 (t.56-1.57) 2.49 (L65-3.77) 613 (5.30-7.09) never smokers, in agreement with previous results for this cohort." The top rows of Table 3 show the death rate for former smokers who had nor smOked for <1 year as of 1959 and those who survived for 2 years through 1961 had higher death rates than current smokers. The death rate for the former smokers who survived dfrough 1964 was somewhat less than the death rare for current smokers- These dara show that in the first few years afrer ces:ation, the death rues for former smokcrs are at least as high as the death rates for current smokcrs_ To examine the morrality risk among former smoken more closely, Table 4 shows the death rates by 2-year follow-up periods for formet cigarette smokers who had not smoked for 0-6 years as of 1959, 1965, and 1972. The 1959 data are for former smokers who had quic 0-6 years according to the.1959 quesuonnairn, the 1965 data are for the 1959 smokers who did not smoke in 1965; the 1972 data are for the 1965 smokets who did not smoke in 1972. The death rates by 2-year follow-up periods for cunetat cigacette smokers as of 1959. 1965, and 1972 are also shown. Note that for 2-year follow-up of ma)es and 2-ycar and 4-year follow-up of females, the death rates for former smoke*s are slightly larger than rhose for current smokers. These fitulings are cotssistent with a previously noted fioding of the 1959 quest(onnaire data, nantcly, that around the time they answer a questionnaire a substan- rial number of recent quitters are in relatively poor healrh27 The poor health effect is laegesr in the 1 st year of cessariore and diminishes with each additional year of cessarion before the quescfotulaire and each additional year of follow-up after the questionnaire. The results in Tables 3 and 4(ndicate that the benefrts of cess2eion are very Gmfted during the firsc 5 years of cessation, but are substantial for rhose who survive many years after quit- ting. The special value of this study is that it is possible to exsmine the overall impact of cesgation on a large co- hotr of persons who smoked cigarettes in 1959 and then largely quit smoking during the next 38 years. This assessment is notjust of long-term healthy formersmok- ers, but of a cohorr of smokets as a whole as ]t tmdenvent almost total cessation, which includes both short-cerm and long-term quictets, a portion of whom were nor healthy. Table 5 compares the death tates foc an initial period, 1960-1969, when most of the 1959 smokets were still smoking, with the death races for the entire follow-up period, 1960-1997, during which there was a substantial degree of cessation. All subjecs and 211 1959 current cigarette smokers are compared with 1959 ttever smokers for death from lung cancer and all causes. Both absolute and relative differences in dearh tates are presented. When comparing the 1960-1969 and 1960-1997 pe- riods, the death rates generally decrease slightly far all causes and generally increase subsrantially for lung can- cer. The absolute difference in death rates between curretu smokers and never smokers remains relatively constant for males but increases substantially for females. The ratio of death taces for current smokers to never smokers has remained at 1.97 for males and has in• creased for females frOm 137 to 157 for all causes. The proportional hazards ratio, based on all deaths and twt jmt those with attained age 35-84, decreased from 1.95 to 1.74 for males and increased from 1.47 to 1.53 for females. The corresponding ratios foc lung cancer re• mained rqitghly constant for males and increased three- foid for feinales. Detai(ed trends in deadt rates for all u