|
O
<br />..� O >larpuana UWarwl Vorti lily
<br />TABLE 2 -Relative Risk of Death lot Ever Users and Current Users of Marijuana, by Sex and Cause of Death: Kaiser
<br />Permanente Medical Care Program Members (n = 65 171), Oakland and San Francisco, June 1979 through
<br />December 1985
<br />Ever Users
<br />Current Users
<br />No Deaths in
<br />Nonsmokers/
<br />Nonsmokers/
<br />Reference
<br />Full Model- Occasional Dnnkersb
<br />Full Model- Occasional Drinkers
<br />Group
<br />Cause of No.
<br />No
<br />(Nonusers)
<br />Death Deaths
<br />RR (95% CI) RR (95% GI)
<br />Deaths°
<br />RR (950,16 CI) RR (95% CI)
<br />Experimenters)
<br />Men
<br />AIDS 152
<br />1 80 (1.29. 2.52) 2 34 (1 32, 4 14)
<br />104
<br />1.90 (1 33, 2 73) 1.91 (1.02, 3.55)
<br />55
<br />Non -AIDS 266
<br />1.11 (092, 1.34) 1.25 (0 82, 1.89)
<br />153
<br />1 12(089. 1.39) 1.15 (0 67, 1.97)
<br />315
<br />Neoplasms 45
<br />0.78(052. 1.18) 0.46(0 15, 1.41)
<br />30
<br />0 97 (0.61, 1.55) 0.75 (0.21, 2 69)
<br />90
<br />Circulatory disease 60
<br />1.08 (0.75. 1.55) 1.36 (0 58. 3 19)
<br />37
<br />1.22 (0.80. 1.87) 1.80 (0 68, 4 73)
<br />102
<br />Injury/poisoning 92
<br />1.24 (0.87, 1.75) 1.65 (0.84, 3 23)
<br />47
<br />0 99 (0.65, 1.50) 1.23 (0 51, 2.95)
<br />72
<br />Other causes 69
<br />1.47 (0.98, 2 22) 1.71 (0 62. 4 68)
<br />39
<br />1.39 (0.86, 2.24) 0 68 (0.14, 3.45)
<br />51
<br />Unknown 12
<br />.. ...
<br />8
<br />...
<br />7
<br />Total 430
<br />1.28 (1.09, 1.50) 1.58 (1.14, 2.16)
<br />265
<br />1.33 (1 11, 1.59) 1.50 (1.01, 2 22)
<br />377
<br />Women
<br />Neoplasms 36
<br />0 82 (0.54, 1.22) 0.70 (0.31, 1.58)
<br />19
<br />0 86 (0 51, 1.45) 0 56 (0 17, 1.90)
<br />155
<br />Circulatory disease 13
<br />0.68 (0.35, 1.33) 0.34 (0.(14, 2.90)
<br />10
<br />0.96 (0.46, 2.02) 0 70 (0 08. 5.95)
<br />64
<br />Injury/poisoning 28
<br />1.39 (0.79, 2 44) 1 40(065, 3 04)
<br />19
<br />1.86 (0 99, 3.51) 2.04 (0 85, 4.91)
<br />38
<br />Other causes^ 16
<br />0.76 (0.39, 1.46) 0.40 (0.05, 3.35)
<br />10
<br />0.95 (0.45, 2.04) 0 84 (0 10, 7 23)
<br />50
<br />Unknown 4
<br />... ...
<br />3
<br />...
<br />4
<br />Total 97
<br />0.90 (0.69, 1.16) 0.81 (0.49, 1 34)
<br />61
<br />1.09 (0.80, 1.48) 1.03 (0 55, 1 90)
<br />311
<br />Note. RR = relative risk, Cl = confidence Interval. we
<br />•Adjusted for age, race. education, marital status, obesity, cigarette smoking,
<br />and alcohol use.
<br />bAdlusted for age, race, education, mantal status. and obesity
<br />9nctudes 3AIDS deaths.
<br />dNo. deaths In current users is included in no. deaths in ever users
<br />Service Cornnussmn, and the Railroad
<br />Retirement Board. While Social Security
<br />numbers were avadable for about two
<br />thirds of the study cohort, only 27 of 1215
<br />deaths (2.2170 were ascertained by out-of-
<br />state search. Since ca s of death were
<br />unavailable for oui-of-state deaths, these
<br />deaths were included in analyses of total
<br />mortality but excluded in subcategory
<br />r nronaluy analyses.
<br />The overall age-specific mortality
<br />rates of this group were about three
<br />quarters as large as the corresponding
<br />1987 United States rates.)' a discrepancy
<br />we attribute to the probable better health
<br />and predominantly employed status of our
<br />Insured population and to our inability to
<br />ascemmn mortality lin subjects without
<br />Social Security numbers who had left
<br />California.
<br />ArUAsis
<br />SAS programs were used for slatistl-
<br />cal analyses." Cox propottional hazards
<br />models were used to examine the joint
<br />effect of sociodemographhc characteristics
<br />and use of marijuana, tobacco, and
<br />alcohol on mortality risk; esumutei of
<br />relauve risks and associated 95`'e conh-
<br />dence intervals were obtained from these
<br />models.16 Age -squared temis were en-
<br />tered into Cox proportional hazards mod-
<br />els to detennme whether there was a
<br />nonlmrar relationship between age and
<br />mortality and were Included when signifi-
<br />cant. Interactions between marijuana and
<br />tobacco use and between marijuana and
<br />alcohol use were tested In the selected
<br />models (total mortality. AIDS mortality
<br />[men only], non -AIDS mortality [men
<br />onlyl, and morality from mjuries/poison-
<br />mgs) by including cross -product terms In
<br />our profxinional hazards modek. None of
<br />the Interactions were statistically signlfi-
<br />cant(P < .05).
<br />Results
<br />Soctalenhugraphic characteristics of
<br />the sample are shown in Table I. The
<br />cohort comrsted of 389c nonusers. 20`rr
<br />experimenters, 201b former users, and
<br />22% current users.11he percentage of ever
<br />users was highest in the 20- through 29 -
<br />yew -old agc group. Ever use of marijuana
<br />was inure common among men than
<br />among women and was highest among
<br />Whites. Never -manned men and women
<br />were about twice as likely to be ever users
<br />as their marred counterparts. Soclolenm-
<br />graphic patterns were generally smular
<br />for current marijuana use.
<br />Current marijuana users were twice
<br />as likely as never users to be current
<br />tobacco cigarette snickers and nearly 2.5
<br />tines as likely to be alcohol drinkers. The
<br />percentage of current snmukern was 'It2
<br />for never marijuana users. 31"2 fur
<br />experimenters. 32% liar former users. and
<br />42'%e for current users. The corresponding
<br />percentages of those amsummg one ser
<br />more drink per Jay were I o<2. '752. 3 1'; ,
<br />and 3917. While few marijuana users were
<br />nonusers of alcohol. a substantial propur-
<br />lion of ever marijuana users 125rr of men.
<br />30% of wonmem) and current manjuana
<br />users (22% of nhen. 28% of women) were
<br />nonsmokers of tubacco crgaremes and
<br />occasional (less than once per month)
<br />drinkers.
<br />We compared risks of mortality as-
<br />sociated w uh ever and current use Manse
<br />to never or experimental use of marijuana
<br />There were 807 deaths arnong men and
<br />408 deaths among women lin this cohort.
<br />Ne performed analyses for total mortal-
<br />ity. AIDS (men only). neoplasms, clrcula
<br />Anil 11)97 Vni V pan a ' i Amrnrar Intimal nl Pnhhr Ilrallh i97
<br />
|