Cannibis and Cognition-1

8/2/2019 Cannibis and Cognition-1

1/7

American Journal of EpidemiologyCopyright O 1999 by The John s Hopkins University School of Hygiene and Public HealthAll rights reserved

Vol. 149, No. 9Printed In U.SA.

ORIGINAL CONTRIBUTIONSCan nabis Use and C ognitive Decline in Persons under 65 Years of Age

Constantine G. Lyketsos,1"3 Elizabeth G arrett,4 Kung-Yee Liang,4 and James C. Anthony2-3

The purpose of this study was to investigate possible adverse effects of cannabis use on cognitive declineafter 12 years in persons u nder age 65 years . This w as a follow-up study of a probability sam ple of the adulthousehold residents of East Ba ltimore. The analyses included 1,318 participants in the Baltimore, Maryland,portion of the Epidemiologic Catchment Area study who completed the Mini-Mental State Examination (MMSE)during three study waves in 1 98 1, 1982, and 1993 -1996 . Individual MMSE score differences between waves 2and 3 w ere calculated for each study participant. After 12 years, study pa rticipants' scores declined a mean of1.20 points on the M MSE (standard deviation 1.90), with 66 % having scores that declined by at least one point.Significant numbers of scores declined by three points or more (15% of participants in the 18-29 age group).There were no significant differences in cognitive decline between heavy users, light users, and nonusers ofcannabis. There were also no male-female differences in cognitive decline in relation to cannabis use. Theauthors conclude that over long time periods , in persons under age 65 years, cognitive decline occurs in all agegroups. This decline is closely ass ociated with aging and educational level but does not appear to be associatedwith cannabis use. Am J Epidemiol 1999;149:794-800.aging; cannabis; cognition; dementia

Cognitive capacity has multiple determinants,including genetic makeup, nutritional status, healthstatus, formal education, and age-related developmen-tal processes. This capacity generally reaches its peakin early adulthood and then declines later in life (1).Cognitive decline is a significant public health prob-lem, given its association with impaired functioningand increased mortality (1) and its close link to demen-tia (2-4). Dementia is defined as the occurrence ofmeasurable, global cognitive decline sufficient toimpair functioning (5). The prevalence and incidenceof dementia, now on e of the most common and seriousdiseases of the elderly, is rapidly increasing as theworld population ages (6, 7).Epidemiologic studies of dementia and of cognitivedecline have typically investigated individuals over theage of 60 years. The expected prevalence of dem entia inReceived for publication April 14 ,199 8, and accepted for publica-t ion September 2,1998.Abbreviation: MMSE, Mini-Mental State Examination.1 Neuropsychiatry and Memory Group, Department of Psychiatryand Behavioral Sciences, School of Medicine, The Johns HopkinsUniversity, Baltimore, MD.2 Department of Mental Hygiene, School of Hygiene and PublicHealth, The Johns Hopkins U niversity, Baltimore, MD.3 Department of Epidemiology, School of Hygiene and PublicHealth, The Johns Hopkins U niversity, Baltimore, MD .4 Department of Btostatistics, School of Hygiene and PublicHealth, The Johns Hopkins U niversity, Baltimore, MD .Reprint requests to Dr. C. G. Lyketsos, Osier 320, The JohnsHopkins Hosp ital, 600 N orth Wolfe Street, Baltimore, MD 21287-5371.

these age groups is 2 percent or higher (6 ,7) , and preva-lence might be as high as 48 percent in those over age85 (6, 7). In late Me, dementing processes hamper thestudy of cognitive decline as a phenomenon distinctfrom dementia. Additionally, recent research suggests(8) and scientific consensus concurs (9) that dementia isbest understood as the result of cumulative effects on thebrain from diseases (e.g., Alzheimer's disease or cere-brovascular disease) and other exposures (e.g., alcoholor tobacco use), all occurring against background, po s-sibly lifelong, declines in cognition associated withaging itself. However, epidemiologic knowledgeregarding cognitive d ecline in persons younger than age65 is very limited. Indeed, we could find only one pub-lished epidemiologic study of cognitive decline inyounger persons: the Seattle Longitudinal Study (10).The Seattle Longitudinal Study followed a series ofcommunity-based cohorts of individuals enrolled in ahealth maintenance organization. Sample sizes forindividual cohorts were between 500 and 997.Participants were assessed according to a large num berof tests of intelligence and cognitive capacity. Themain findings were that individual cognitive abilitiesdo not change much before age 60, with the exceptionof verbal fluency. Because of attrition, the SeattleLongitudinal Study did not have sufficient samplesizes to detect small cognitive declines in younger agegroups. Furthermore, very few individual participantswere followed for spans of more than 5 years.

794

by

gueston

Jan

uary14

,2

012

http://aje.oxford

journ

als.org

/

Downloa

dedfrom
http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/


2/7

Cannabis Use and Cognitive Decline 795

The major correlate of cognitive decline is increas-ing age (10-14). Higher educational level (14) andhigher functioning (13) are associated with less cogni-tive decline. Being female or encountering stressfullife events is not associated with cognitive decline (11,13). Risk factors for dementia include age, prior cog-nitive impairment, stroke, high blood pressure, heartdisease, diabetes mellitus, alcohol consumption, anddepression (15-28). The use of nicotine via smokinghas also been associated with a lower risk for demen-tia, although this finding is controversial (29). Beingfemale has not been associated with the incidence ofdementia (15, 17). Two recent studies (30, 31) havereported that lesser educational attainm ent is a risk fac-tor for dementia. However, this finding has not beensupported universally (17, 32, 33).The relation between cognitive functioning or cog-nitive decline and use of cannabis (marijuana) hasreceived limited attention in epidemiologic studies.Two cognitive effects of cannabis must be distin-guished: acute effects, those associated with intoxica-tion, and residual effects, which persist after the drughas left the central nervous system (34). The lattereffects might be short term or long term. Cross-sectional studies, either experimentally administeringcannabis or comparing users with nonusers, supportthe existence of short term residual effects of cannabisuse on attention, ability to perform psychomotor tasks,and short term memory (34, 35). These effects aremore severe in women (36) and in heavy users ofcannabis as compared with light users (37).To our knowledge, no study with published resultshas investigated the long term effects of cannabis useon cognition in an epidemiologic sample. According toPope et al. (34), study designs best suited to address-ing this issue are naturalistic com parisons, in large epi-demiologic samples, of heavy users, light users, andnonusers of cannabis. These studies must also accountfor the concurrent use of alcohol and other drugs, bothillicit and legal (e.g., nicotine). In addition, such stud-ies must adjust for other factors known to influencecognition over time, such as age and education, andmust investigate possible interactions between thecognitive effects of cannabis use and gender (beingfemale).We recently reported findings from a 13-year follow-up of 1,488 persons of all ages who had participated inthe Baltimore, Maryland, portion of the EpidemiologicCatchment Area study (38). The Mini-Mental StateExamination (M MSE ) (39), a widely used quantitativemeasure of cognition, was administered to participantsduring wave 1 (1981) and during two follow-up wavesin 1982 and 1993-1996. The design of the studyallowed us to examine cognitive decline between

waves 2 and 3 in a large epidemiologic sample. Wefound that cognitive decline occu rred in all age groups.Age, education, and minority status were all signifi-cantly associated wiui greater cognitive decline.In this follow-up paper, we focus our investigationon persons under age 65 years. To our knowledge, thisis the first population study that has investigated cog-nitive decline in this age group, in which the preva-lence of dementia is very low. This perm its better studyof cognitive decline as a phenomenon distinct fromdemen tia, as well as its associated risk factors. W e hadtwo goals: 1) to further delineate the epidemiology ofage-specific cognitive decline in persons under 65 and2) to investigate any long term association betweencognitive decline and use of cannabis using a designsimilar to the one proposed by Pope et al. (34).MATERIALS AND METHODSBaltimore Epidemiologic Catchment Areafollow-up

The Epidemiologic Catchment Area program hasbeen described in detail elsewhere (40, 41). TheBaltimore arm of this five-site study first entered thefield in 1981, when the first wave of in-person assess-ments was completed. A second wave of assessment(including wave 2 administration of the MMSE) wasconducted 1 year later, in 1982. The Baltim oreEpidemiologic C atchment Area target population con-sisted of the adult household residents of easternBaltimore City, an area with 175,211 inhabitants.During wave 1, 4,238 individuals were designated forinterview by probability sampling methods, and 3,481(82 percent) completed interviews. Of these persons,2,695 completed interviews during wave 2.In 1993, all 3,481 initial participants were targetedfor tracing and interviewing. A total of 848 participantswere found to have died; the remaining 2,633 werepresumed to be alive, but 415 of them could not besuccessfully traced. Of the 2,218 persons located, 298refused to participate, and 1,920 completed interviews.Of these, 1,488 had completed the MMSE during allthree waves, approximately 11.5 years after wave 2.All study participants signed informed consent state-ments approved by the Institutional Review Board ofthe Johns Hopkins University School of Hygiene andPublic Health.

ParticipantsIn these analyses, we included only those partici-pants who w ere under age 65 at wave 1 and who com-pleted the MMSE during all three study waves (n =

1,318).Am J Epidemiol Vol. 149, No. 9, 1999
http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/http://aje.oxfordjournals.org/


3/7

79 6 Lyketsos et al.

Measu rement of cognitive decline. For each par-ticipant, an MM SE score difference was calculated bysubtracting the wave 3 (1993-1996) MMSE scorefrom the wave 2 (1982) MMSE score. The mean timeinterval between the points at which these MMSEswere adm inistered was 11.6 years (standard error 0.01years). The median interval was 11.5 years, the 25thpercentile w as 11.3 years, and the 75th percentile was11.9 years. Change in MMSE score between waves 2and 3 was the primary dependent variable in theanalyses.

Classification of participants according to use ofcannabis. Participants were separated into fivegroups based on their self-reported drug use duringall three waves of the study. Group 1 (nonusers) w erethose who reported in all three waves that they hadnever used cannabis in any form (n = 806 (61 per-cent)). Group 2 (light users) were participants whohad used cannabis but had never used it daily or moreoften for over 2 weeks (n = 235 (18 percent)). Group3 were light users who reported use of any other illicitsubstance in any study wave (n = 131 (10 percent)).Group 4 (heavy users) reported during at least onestudy wave that they had used cannabis daily or moreoften for over 2 weeks (n = 137 (10 percent)). Group5 were heavy users of cannabis who reported use ofother illicit drugs as well (n = 8 (1 percent)).Information on cannabis use was missing for oneparticipant.Classification of participants according to use ofalcohol or tobacco. On the basis of the highest alcoholintake reported for the past month during any of thethree study waves, participants were placed into threegroups: never drinkers (n = 67 (5 percent)), light-to-moderate drinkers (n = 778 (59 percent)), and heavydrinkers, defined as those who had had more than fourdrinks on any one day during the past month (n = 47 3(36 percent)). With respect to smoking, three groupswere defined on the basis of self-report during any ofthe three waves: never smokers (n = 347 (26 percent));occasional smokers (n = 573 (44 percent)); heavysmokers, defined as those who smoked 20-39 cigarettesper day (or the equivalent in cigars or pipefuls oftobacco) (n = 310 (24 percent)); and very heavysmokers, those who smoked two or more packs ofcigarettes per day (or the equivalent) (n = 85 (6 per-cent)). Information on smoking was missing for threeparticipants.Other variables associated with cognitive declineused as covariates. Information on other variablesassociated with cognitive decline was recorded atwave 1. Gender w as indicated as m ale or female. Agewas grouped as follows: 18-30, 31-^ 0, 41-50, 51-6 0,and 61-64 years. Minority status was indicated as

African-American or Hispanic versus other ethnicity(non-Hispanic white). Five educational subgroupswere developed: 0-8 years, 9-11 years, 12 years orGeneral Equivalency Diploma, 13-15 years, and >16years, in conformance with common educational land-marks (grade school, some high school, completedhigh school or the equivalent, some college, and com-pleted college). It is possible that some study partici-pants, especially those in younger age groups at wave1, completed their education after wave 1 and werethus misclassified.Analyses

Mean M MS E score changes between waves 2 and 3 *(with 95 percent confidence intervals) are reported inthe tables for the entire cohort and for subgroups byage. The proportions of individuals who evidenced anyincrease, no change, a one-point decline, a two-pointdecline, a three-point decline, or a four-point or greaterdecline are also reported by age group. Mean changein MMSE score (with its 95 percent confidence inter-val) by level of cannabis use was estimated for menand women separately. The relation between level ofcannabis use and MM SE score change between w aves2 and 3 was examined in a series of linear regressionmodels with MMSE score change as the dependentvariable and cannabis use as the independent variable,with or without inclusion of the other covariates. Forboth univariate and multiple regression models, theassociation of cannabis use with change in MMSEscore is reported in the form of regression coefficients(with 95 percent confidence intervals). Subgroupswere entered into regression models individually as"dummy" variables to allow direct comparisons ofregression coefficients using one of the subgroups asthe reference category.To validate the findings from the linear regressionmodels, we also constructed a series of proportionalodds logit models (42) relating diseases or substanceuse to MMSE score change. These were bivariate ormultivariate "analogs" to the linear models. Thedependent variable was "change in MMSE score,"grouped as follows: any increase, no change, a one-point decline, a two-point decline, a three-pointdecline, or a four-point or greater decline. Findingsfrom these models w ere similar to those obtained fromthe linear models. For simplicity, we report only find-ings from the linear models.RESULTS

Table 1 provides a description of the study cohortat wave 1 with regard to sociodemographic variables.It also shows mean MMSE scores at each study wave.Aw J Epidemiol Vol. 149, No. 9, 1999

by

gueston

Jan

uary14

,2

012

http://aje.oxford

journ

als.org

/

Down

loadedfrom


4/7

Cannabis Use and Cognitive Decline 79 7TABLE 1. Soclodem ographic characteristics of theBaltimore Epldemlologlc Catchment Area study cohort atwave 1 {n = 1,318) and mean MMS E* scores at waves 1-3

VariableAge (years)18-30

3 1 -4 04 1 -5 051-6061-64Gender

MaleFemaleRaceMinority (African-Americanor Hispanic)

Nonminority (other)Education (years)0 -89-1112/GED*

13-151 6Mean MMSE scoreWave 1 (1981)Wave 2 (1982)

Wave 3 (1993-1996)

No.

54 531 917918590

48 883 0

49 082 8

16128 054 121 112528.6528.6527.46

%

412414147

3763

3763

1221411610(1.90)t(1.81)(2.23)

* MMSE, Mini-Mental State Examination; GED, General Equiv-alency Diploma.t Numbers in parentheses, standard deviation.

Cognitive decline between waves 2 and 3Table 2 shows the mean change in MMSE scorebetween waves 2 and 3 for every age group. It alsoshows the proportions of participants in each agegroup with specific changes in MMSE score, asdescribed above. Persons in all age groups had mean

declines greater than zero, with two thirds declining inscore by at least one point. The mean decline and theproportion of persons with declining scores increasedsteadily with age, as expected. It is noteworthy that inevery age group there was a notable proportion of par-ticipants whose score declined three points or morea change of a magnitude that merits clinical attention(43, 44). These estimated declines must be consideredin the context of MMSE measurement error, theMMSE ceiling effect, and normal variation in MMSEscores over time (see Discussion).Association between cannabis use and scoredecline

Table 3 displays estimated mean changes in MMSEscore according to level of cannabis use for men andwomen separately. Women who were nonusers ofcannabis had scores that declined more than those ofmen who were nonusers. However, within male-female groups, there were no evident differences inscore decline by cannabis use for either men orwomen.Table 4 displays results from the linear regressionmodels with MMSE change between waves 2 and 3used as the dependent variable. The numbers shown inthe table are regression coefficients estimating th e rel-ative change in MMSE score for a given group ofcannabis users relative to nonusers. Model 1 includedonly cannabis use as the covariate. Model 2 includedcannabis use and use of alcohol and tobacco. Model 3included cannabis use plus age, gender, education,minority status, alcohol use, and tobacco use. Model 4included cannabis use plus all of the variables frommodels 2 and 3. Both light and heavy u sers of cannabisevidenced less cognitive decline than nonusers,although this finding was not statistically significant at

TABLE 2. Mean change in Mlnl-Mental State Examination (MMSE) score from wave 2 (1982) to wave 3 (1993-1998 ) andproportions of participants evidencing specific MMSE score changes, by age group, Baltimore Epldemlologlc Catchment Areastudy follow-up

Age group(years)

18-30 (no 545)31-4O(n = 319)4 1 - 5 0 {n= 179)5 1 - 6 0 ( n = 1 8 5 )61 -64 (n = 90 )Alleges[n= 1,318)

Meanchange- 0 .98-1 .08-1 .25-1.52-2 .12

-1 .20

ChangeIn

95 %i Cl*-0 .83 to - 1 .13-0.89 to-1.27-0.92 to-1.58-1.20 to -1.84-1.52 to-2.72

-1.10 to-1.30

AnyscoreIncreaseNo.7343352913

193

%1313201614

15

Nochangein scoreNo.12259302910

25 0

%2219171611

19

Score change groupDecline

No.179107394818

39 1

of 1point%3334222620

30

Decfineo f2pointsNo.9360372720

23 7

%1719211522

18

Decline

No.472917217

121

O f3points%9910118

9

I

No.3121213122

126

Declineof 24points%67121724

10* Cl, confidence interval.

Am J Epidemiol Vol. 149, No. 9, 1999

bygueston

Jan

uary14

,2

012

http://aje.oxford

journ

als.org

/

Dow

nloadedfrom


5/7

798 Lyketsos et al.TABLE 3. Mean change In Mini-Mental State Examination(MMSE) score between wave 2 (1982) and wave 3 (1993-1996)In men and women, by level of cannabls use, BaltimoreEpldemlologlc Catchment Area study follow-up

Gender andlevel ofcannabls useMen

Non usersLight usersLight users plus useof drugsHeavy usersHeavy users plus useof drugsWomenNonusersLight usersLight users plus useof drugsHeavy usersHeavy users plus useof drugs

No.

25 11044882

3

5551318355

5

Meanscorechange InMMSE-1 .00-1.03-1.06-0.84-0.33

-1.46-1.04-1.07-1 .15-0.60

95% Cl*

-0 .73 to-1 .27-0.67 to -1.39-0.57 to-1 .55-0.46 to -1.22+5.93 to -6.59

-1.29 to-1 .63-0.71 to-1 .37-0.77 to-1 .37-0.47 to-1 .83+3.09 to -4.29

Cl, confidence interval.

the conventional level of p < 0.05 (model 1). Afteradjustment for the other variables in models 2-4, therewas no association between cannabis use and cognitivedecline.DISCUSSION

Cognitive decline is an age-related phenomenon thataffects persons of all ages, including those und er age 30years. It becomes more pronounced with increasing ageand is most evident in persons over age 59. A signifi-

cant proportion (>15 percent) of persons in all popula-tion age groups evidence declines that approach clini-cal significance. We offer two interpretations of thisfinding. One is that cognitive decline might be aninevitable phenomenon of aging, perhaps modified bygenetic mak eup, education, nutrition, disease, and envi-ronmental ex posure. Another is that the declines are theresult of slowly progressive neurodegenerative diseases(such as Alzheimer's disease) which might be lifelongin evolution but do not lead to clinical symptoms untilmuch later in life (8). While these two lines of reason-ing are not mutually exclusive, the relation between ageand cognitive decline across all age groups reportedhere lends greater support to the former.

To our knowledge, this was the first long termprospective study in the United States that had a com-munity sample large enough to investigate the relationbetween cannabis use and cognitive decline in personsunder age 65 years. Other studies have found shortterm residual effects of cannabis use on memory andcognition (34, 35) that are more severe among women(36) and heavy users (37). However, our data suggestthat over the long term cannabis use is not associatedwith greater declines in cognition among men,women, or heavy users. The study design we usedincluded several of the features proposed by Pope etal . (34) as critical to addressing the long term effectsof cannabis on cognition: naturalistic follow-up, alarge sample size, a population basis, comparison oflight cannabis use with heavy use, and the construc-tion of models accounting for the effects of genderand use of illicit drugs, alcohol, and tobacco.Therefore, these results would seem to provide strongevidence of the absence of a long term residual effectof cannabis use on cognition.

TABLE 4. Regression coe fficients (p) Indicating relative differences In Mini-Mental State Exam ination(MMSE) score change between wave 2 (1982) and wave 3 (1993-1996), by level of cannabls use, in fourregression models, Baltimore Epldemlologlc Catchment Area study follow-upt

Levelofcannablsuse

NonusersLight usersLight users plus use of drugsHeavy usersHeavy users plus use of drugs

Model 1(cannabis use)

P SE+0.28* 0.150.25 0.190.35* 0.180 .81 0 .71

Model 2(cannabis useplus use ofalcohol andtobacco)P SE

0.22 0.150.17 0.190.27 0.190 .66 0 .71

Model 3(cannabis useplus ag e, gender,minority status,and education)

P SE-0.001 0.16-0.07 0.190.08 0.200.79 0.70

Model 4(cannabls useplus age , gender,minority status ,education, anduse of alcoholand tobacco)P SE

-0.02 0.17-0.10 0.190.05 0.200.69 0.70 p < 0 . 1 0 .t Positive numbers indicate MMSE score increases relative to trie reference group; negative numbers Indicaterelative decreases in MMSE score.t SE, standard error. Reference group.


y

g

y

,

pj

j

g


6/7

Cannabis Use and Cognitive Decline 7 9 9

Notable limitations of this study include loss tofollow-up and mortality. Cognitive functioning at base-line was a predictor of both mortality and loss tofollow-up in the Epidemiologic Catchment Area study(40). Additionally, it is possible that some cannabisusers in the study may have used cannabis on the daythe MMSE was administered. Given the acute effectson cannabis on cognition (34), this would have tendedto reduce dieir MMSE score on that day. This mayhave adversely affected accurate measurement ofMMSE score changes over time.Given that a lower level of cognitive functioningwas associated with greater cognitive decline, theseestimates of decline may be underestimates. Theassessment of cannabis use was based on self-reportsand was not confirmed with biologic measures or con-trolled in an experimental setting. This may have led tounderestimation of cannabis use in persons with poormemory.Another important limitation of the study is that theMMSE is not a very sensitive measure of cognitivedecline, even though it specifically tests memory andattention. Thus, sm all or subtle effects of cannabis useon cognition or psychomotor speed may have beenmissed. The MMSE is not intended for the purpose forwhich it was used in this study, and it contains someitems that assess neurologic function as well as cogni-tion. Additionally, MMSE item analysis was not per-formed in this study. Given the MMSE's ease of useand widespread application, it was the most practicalinstrument available for brief assessment of cognitivefunctioning at the time the multisite EpidemiologicCatchment Area study was planned in the late 1970s.Also, given its limited sensitivity, declines noted onthe MMSE are probably underestimates of truedeclines.Other limitations of the MMSE include the fact thatsmall errors, such as forgetting the present day's date,may be due to measurement error and not to truedecline. Measurement error on the MMSE might becaused by a variety of factors, including the ambientenvironment in which the test is taken, the respon-dent's mood or emotional state, the respondent's ade-quacy of sleep the night before, the time of day atwhich the test is given, and other factors. However,such errors ought to be random and not systematic(equally distributed between study waves), so theeffect on mean estimates should "average out" acrossthe population and across waves of assessment.MM SE scores in this study exhibited a ceiling effect,given that most participants scored in the 27-30 rangeduring wave 1. However, the ceiling effect was limitedto a minority of participants, those who scored 30points at baseline, since most declines were small.

Finally, the small but tangible beneficial "practiceeffect" of repeated testing on MM SE score would tendto lead to higher, not lower, MM SE scores at follow-up.We conclude that cognitive decline occurs across allage groups, with a significant proportion of persons ofall ages showing declines near clinically significantlevels after 12 years. Such decline is not associatedwith cannabis use in either men or women. A betterunderstanding of predictors of cognitive dec line in per-sons under age 65 years might lead to interventionsdesigned to slow or arrest such decline. This in turnmight reduce the incidence of dementia at older ages.

ACKNOWLEDGMENTSThis study was supported by grant 1R01-MH47447 fromthe National Institute of Mental Health for Baltimore

Epidemiologic Catchment Area study follow-up.

REFERENCES1. Zee RF. The neuropsycho logy of aging. Exp Gerontol1995;30 :43 l^2 .2. Bickel H, Cooper B. Incidence and relative risk of dementia inan urban elderly population: findings of a prospective fieldstudy. Psychol Med 1994:24:179-92.3. O'Brien JT, Beats B. Benign senescent forgetfulness and ageassociated memory impairment. In: Burns A, Levy R, eds.Dementia. New York, NY: Chapman and Hall Medical, 1994:295-308.4. Huppert FA, Brayne C. What is the relationship betweendementia and normal aging? In: Huppert FA, Brayne C,O'Connor DW, eds. Dementia and normal aging. Cambridge,England: Cambridge University Press, 1994:311.5. American Psychiatric Association. Diagnostic and statisticalmanual of m ental disorders: DSM-IV. 4th ed. W ashington, DC :American Psychiatric Press, 1994.6. Corrada M, Brookmeyer R, Kawas C. Sources of variability inprevalence rates of Alzheimer's disease. Int J Epidemiol1995;24:1000-5.7. Henderson AS. Alzheime r's disease in its epidemiological con-text Acta Neurol Scand Suppl 1993;149:1-3.8. Snowdon DA, Greiner LH, Mortimer JA, et al. Brain infarc-tion and the clinical expression of Alzheimer disease: The NunStudy. JAMA 1997;277:813-17.9. Small GW, Rabins PV, Barry PP, et al. Diagnosis and treatmentof Alzheimer disease and related disorders. Consensus state-ment of the American Association for Geriatric Psychiatry, theAlzheim er's Association, and the American Geriatrics Society.JAMA 1997;278:1363-71.10. Shaie KW. The course of adult intellectual development AmPsychol 1994;49:304-13.11. Grimby A, Berg S. Stressful life events and cognitive func-tioning in late life. Aging (Milano) 1995;7:35-9.12. Hultsch DF, Hertzog C, Small BJ, et al. Short-term longitudi-nal change in cognitive performance in later life. PsycholAging 1992;7:571-4.13. Johansson B, Zarit SH, Berg S. Changes in cognitive function-ing of the oldest old. J Gerontol 1992;47:P75 -80.14. Farmer ME, Kittner SJ, Rae DS, et al. Education and changein cognitive function: The Epidemiologic Catchment AreaStudy. Ann Epidemiol 1995;5:l-7.


by

gueston

Jan

uary14

,2

012

http://aje.oxford

journ

als.org

/

Dow

nloadedfrom


7/7

80 0 Lyketsos et al.15. Minam i Y, Tsuji I, Fu kao A, et al. Physical status and dem en-tia risk: a three year prospective study in urban Japan. Int J SocPsychiatry 1995 ;41:47 -54.16. Yoshitake T, Kiyohara Y, Kato I, et al. Incidence and risk fac-tors of vascular dementia and A lzheimer's disease in a definedelderly Japanese population: The Hisayama Study. Neurology1995;45:1161-8.17. Letenneur L, Commenges D, Dartigues JF, et al. Incidence ofdementia and Alzheimer's disease in elderly community resi-dents of south-western France. Int J Epidemiol 1994;23:1256-61.18. Hebert LE, Scherr PA, Beckett LA, et al. Age-specific inci-dence of Alzheimer's disease in a community population.JAMA 1995;273:1354-9.19. Payami H, Montee K, Kaye J. Evidence for familial factorsthat protect against dementia and outweigh the effect ofincreasing age. Am J Hum Genet 1994;54:650-7.20 . Boothby H, Blizard R, Livingston G, et al. The Gospel OakStudy stage El: the incidence of dementia. Psychol Med1994;24:89-95.21 . Shen YC, Li G, Li YT, et al. Epidemiology of age-relateddementia in China. Chin Med J (Engl) 1994;107:60-4.22 . Paykel ES, Brayne C, Huppert FA, et al. Incidence of demen-tia in a population older than 75 years in the United Kingdom .Arch Gen Psychiatry 1994;51:325-32.23 . Morgan K, Lilley JM, Arie T, et al. Incidence of dementia in arepresentative British sample. Br J Psychiatry 1993;163:467-70.24 . Bachm an DL, Wolf PA, Linn RT, et al. Incidence of dem entiaand probable Alzh eimer's disease in a general population: TheFramingham Study. Neurology 1993;43:515-19.25 . Hagnell O , Franck A, Grasbeck A, et al. Vascular dementia inthe Lundby study. 1. A prospective, epidemiological study ofincidence and risk from 1957 to 1972. Neuropsychobiology1992;26:43-9.26 . Copeland JR, Davidson LA, Dewey ME, et al. Alzheimer's dis-ease, other dementias, depression and pseudodementia: preva-lence, incidence and three-year outcome in Liverpool. Br JPsychiatry 1992;161:230-9.27 . Mann AH, Livingston G, Boothby H, et al. The Gospel OakStudy: the prevalence and incidence of dementia in an innercity area of London. Neuroepidemiology 1992;ll(suppl 1):

7 6 -9 .28 . Li G, Shen YC, Chen CH, et al. A three-year follow-up studyof age-related dementia in an urban area of Beijing. ActaPsychiatr Scand 1991;83:99-104.29 . Lee PN . Smoking and Alzheimer's disease: a review of the epi-demiological evidence. Neuroepidemiology 1994;13:13144.

30 . Stem Y, Gurland B, Tatemichi TK, et al. Influence of educationand occupation on the incidence of Alzheimer's disease.JAMA 1994;271:1004-10.31 . Ott A, Breteler MM, van Harskamp F, et al. Prevalence ofAlzheimer's disease and vascular dementia: association witheducation. The Rotterdam Study. BMJ 1995;310:970-3.32 . Gilleard CJ. Education and Alzheimer's disease: a review ofrecent international epidemiological studies. Aging MentHealth 1997; 1:33-46.33 . Beard CM, Kokmen E, Offord KP, et al. Lack of associationbetween Alzheimer's disease and education, occupation,marital status, or living arrangement. Neurology 1992;42:2063-8.34 . Pope HG Jr, Gruber AJ, Yurgelun-Todd D. The residual neu-ropsychological effects of cannabis: the current status ofresearch. Drug Alcohol Depend 1995;38:25-34.35 . Fletcher JM , Page JB , Francis DJ, et al. Cognitive correlates oflong-term cannabis use in Costa Rican men. Arch GenPsychiatry 1996;53:1051-7.36 . Pope HG Jr, Jacobs A, Mialet JP, et al. Evidence for a sex-specific residual effect of cannabis on visuospatial memory.Psychother Psychosom 1997;66:179-84.37 . Pope HG Jr, Yurgelun-Todd D. The residual cognitive effectsof heavy marijuana use in college students. JAMA 1996;275:521-7.38 . Lyketsos CG, Chen LS, Anthony JC. Cognitive decline inadulthood: an 11.5-year follow-up of the BaltimoreEpidemiologic Catchment Area study. Am J Psychiatry 1999;156:58-65.39 . Folstein MF, Folstein SE, McHugh PR. "Mini-mental state": apractical method for grading the cognitive state of patients forthe clinician. J Psychiatr Res 1975;12 :189-98.40 . Eaton WW, Anthony JC, Gallo J, et al. Natural history ofDiagnostic Interview Schedule/DSM-rV major depression: theBaltimore Epidemiologic Catchment Area follow-up. ArchGen Psychiatry 1997;54:993-9.41 . Eaton WW, Kessler LG, eds. Epidemiologic field methods inpsychiatry: the NIMH Epidemiologic Catchment Areaprogram. Orlando, FL: Academic Press, Inc, 1985.42 . McCullagh P. Regression models for ordinal data (with discus-sion). J R Stat Soc [B] 1980;42:109-42.43 . Rebok G, Brandt J, Folstein M. Longitudinal cognitive declinein patients with Alzheimer's disease. J Geriatr PsychiatryNeural 1990;3:91-7.44 . Schmand B, Lindeboom J, Launer L, et al. What is a signifi-cant score change on the Mini-Mental State Examination? IntJ Geriatr Psychiatry 1995;10:411-14.


y

g

y

,

pj

j

g

Documents

Cannibis and Cognition-1