Wpic.pitt.edu

Sedative-Hypnotic Use of
Diphenhydramine in a Rural, Older Adult,
Community-Based Cohort
Effects on Cognition
Ranita Basu, M.D., Hiroko Dodge, Ph.D.
Gary P. Stoehr, Pharm.D., Mary Ganguli, M.D., M.P.H.
Objective: The authors sought to identify patterns and associations of prescription
and over-the-counter sedative-hypnotic use in an older, rural, blue-collar, community-
based cohort in southwestern Pennsylvania over 10 years. Methods: A group of 1,627
individuals age 65 and over were recruited and assessed during 1987-1989 and re-
assessed during approximately biennial waves. Data included sleep medications, de- mographics, depressive symptoms, sleep complaints, and cognitive functioning (Mini- Mental State Exam [MMSE]). Results: At Waves 1 through 5, the mean age of the cohort
increased from 73.4 to 80.5 years. Use of prescription sedative-hypnotics (primarily
benzodiazepines) increased from 1.8% to 3.1%, and over-the-counter sedative-hypnotic use (primarily diphenhydramine) increased from 0.4% to 7.6%. At Wave 5 (1996- 1998), 8.17% of the sample reported using diphenhydramine as a sleep aid. After adjusting for age and sex, diphenhydramine use was associated with higher education and more depressive symptoms, the latter becoming nonsignificant after controlling for initial insomnia. MMSE became significantly associated with diphenhydramine use when 143 subjects with dementia were excluded from the analysis. Conclusion:
As the cohort aged, prescription sedative-hypnotic use remained relatively stable,
whereas over-the-counter sedative use, principally diphenhydramine, increased sub- stantially. The association of this drug with cognitive impairment in persons without dementia highlights its potential for causing adverse reactions in older adults. (Am JGeriatr Psychiatry 2003; 11:205-213) Olderadultsinthecommunitycontinuetoconsume as inadequate, or inadequately refreshing. Some, but a disproportionately large share of sedative- probably not all of these individuals suffer from clini- hypnotic drugs.1 Presumably, they perceive their sleep cally significant insomnia, perhaps complicated by ad- Received August 10, 2001; revised November 11, December 4, 2001; accepted December 5, 2001. From the Division of Geriatrics and Neuropsy-chiatry, Department of Psychiatry, University of Pittsburgh School of Medicine, Pittsburgh, PA (RB,MG), the Department of Epidemiology, Universityof Pittsburgh Graduate School of Public Health, Pittsburgh, PA (HD,MG), and the Department of Pharmaceutical Sciences, University of PittsburghSchool of Pharmacy, Pittsburgh, PA (GPS). Address correspondence to Dr. Ganguli, Western Psychiatric Institute and Clinic, 3811 O'Hara Street,Pittsburgh, PA 15213-2593. e-mail: gangulim@msx.upmc.edu.
Copyright ᭧ 2003 American Association for Geriatric Psychiatry Am J Geriatr Psychiatry 11:2, March-April 2003 ditional conditions, such as pain. Some seek prescrip- Survey (MoVIES) study is being conducted within the tion sedative-hypnotics from their physicians, whereas mid-Monongahela Valley area of southwestern Pennsyl- others self-medicate with alcohol or over-the-counter vania. Study procedures receive annual approval from sleep aids, with or without the knowledge of their phy- the University of Pittsburgh Institutional Review Board; sicians. Reasons for self-medication may include the sampling and recruitment of the study cohort have pre- higher cost of prescription medication, limited access viously been described in detail.4-6 Eligibility criteria for to healthcare, beliefs in the superiority of "natural" entry into the study cohort, between 1987 and 1989, products, suggestions from friends and family, and in- included community residence (i.e., not already being creased direct marketing of sleep aids to the public.
in long-term care), age of 65 years or older, fluency in The use of prescription drugs for sleep has not in- English, and at least a sixth-grade education; the latter creased over time in North America, despite heavy ad- two criteria were intended to facilitate interpretation of vertising and the introduction of new benzodiazepines the neuropsychological (cognitive) tests used to screen and new non-benzodiazepine sedative-hypnotic drugs for dementia. The total number of randomly selected such as zolpidem and zaleplon. However, the use of participants who met eligibility criteria and consented nonprescription sleep aids, including antihistamines, to participate was 1,422. An additional 259 volunteer valerian, melatonin, herbal remedies, and kava powder participants met the same entry criteria and brought the has increased over the last decade and now exceeds total cohort size to 1,681 at study entry (1987-1989).5 prescription hypnotic use in the United States.2 Elderly All participants underwent a screening interview at Canadian subjects responding to a survey of nonpre- study entry, and a subset of them (see below) under- scription drug sleep aids reported similar patterns of went a clinical assessment for dementia. All surviving nonprescription drug use for sleep-alcohol, antihista- participants were subsequently contacted for follow-up mines such as dimenhydrinate and diphenhydramine, screening interviews at approximately 2-year intervals, analgesics such as acetaminophen, aspirin, and codeine- in a series of data collection "waves." Data reported in containing products, and herbal products such as this article were collected from participants who sur- herbal teas, St. John's wort, and melatonin.3 The more vived and consented to participate in each data collec- recent introduction of a variety of nonprescription an- tion period, from Wave 1 (1987-1989) through Wave 5 algesic/hypnotic combinations appears to have contrib- uted to the increased use of nonprescription hypnoticdrugs.2 Screening
In an ongoing, prospective, epidemiological study of a cohort drawn from a rural American community, After providing written informed consent, each we examined the use of prescription and over-the- subject underwent an in-home screening and risk-factor counter sedative-hypnotic drugs over 10 years of follow- assessment interview. The interview include approxi- up. We hypothesized that the use of these drugs might mately 25 minutes of cognitive screening,7 including be associated with depressive symptoms as well as cog- the Mini-Mental State Exam (MMSE)8 and other, more nitive impairment in these older adults. Observing that specific, tests of cognitive domains known to be af- by far the most common sedative-hypnotic agent used fected in dementia. The test battery included (but was was diphenhydramine; we focused our analyses on fac- not limited to) the neuropsychological battery consti- tors cross-sectionally associated with the use of this tuted by the NIA Consortium to Establish a Registry for Alzheimer's Disease (CERAD).9 Subjects also respondedto a previously described modified version (mCES-D)10of the Center for Epidemiological Studies-Depressionscale(CES-D).11 Briefly, in the mCES-D, the questions are MATERIALS AND METHODS
asked in the second-person by a trained interviewer. All20 original CES-D items are included, but rated (0: no, Study Site and Population
1: yes) according to whether or not the subject expe-rienced them "most of the time" (operationalized as "3 Originally designed as a population-based dementia or more days") during the previous week. Thus, the registry, the Monongahela Valley Independent Elders mCES-D score reflects the number of symptoms expe- Am J Geriatr Psychiatry 11:2, March-April 2003 rienced on 3 or more days of the preceding week, the score according to the CERAD protocol.9 On the CDR maximum possible score being 20. A history of pre- scale, scores (stages) of 0, 0.5, 1, 2, and 3 indicate no scription and nonprescription (over-the-counter) med- dementia, possible/incipient dementia, mild, moderate, ication use was taken as part of the screening interview, and severe dementia, respectively. For the current anal- with drug information being obtained both from self- yses, we categorized as "demented" all those with CDR report and from medication bottle labels.12,13 Subjects (and reliable informants, when applicable and available)were asked to report their use of medications as of the Categories of Drugs
day of the assessment, using a 2-week frame of refer-ence if necessary. Questions regarding sleep complaints Prescription drug data were initially classified included difficulty falling asleep at night (initial insom- within American Hospital Formulary System (AHFS) nia), difficulty staying asleep or sleep continuity distur- categories.12 Over-the-counter (nonprescription) drugs bance (intermittent insomnia), or early-morning awak- were classified according to therapeutic category.13 Benzodiazepines classified as sedatives were fluraze- On the basis of their screening cognitive test scores pam, temazepam, triazolam, and estazolam. For the cur- at study entry, subjects were classified as cognitively rent analyses, we did not include other benzodiazepines intact or impaired on the basis of the following opera- (e.g., alprazolam and lorazepam), or antidepressant or tional criteria: scores at or below the 10th percentile of antipsychotic drugs, that subjects may or may not have the sample on either the general mental status test been using for sleep. Our rationale was that we did not (MMSE)8 and/or at least one test of memory and one have access to information regarding the indications for test of another cognitive domain. These criteria have which the subjects' physicians had prescribed these previously been shown to be sensitive and specific for dementia.4 During subsequent biennial follow-up wavesof cognitive screening with identical measures, subjects Statistical Methods
whose decline in scores (from previous waves) was atthe 95th percentile of the cohort and those whose First, we examined the frequencies and propor- scores had newly fallen to below "impaired" levels as tions of those taking any sedative/hypnotic drugs at defined above, were classified as "cognitively declined." each wave. Observing that the active component of the At each wave, subjects classified as either cognitively vast majority of these drugs was diphenhydramine, we impaired or declined were asked to undergo a clinical focused our attention on this agent. Preliminary analy- (diagnostic) evaluation for dementia, described below.
ses showed that diphenhydramine use increased dra- Also, a sample of cognitively intact subjects, matched matically at the most recent wave (Wave 5, 1996-1999), for age, sex, and education with subjects diagnosed as when it was present in high enough frequency to per- "demented" (see below) was also selected at Wave 1 as mit multivariate analysis. We therefore focused our sub- a control group for clinical evaluation.15 sequent analyses on diphenhydramine use at this wave.
At Wave 5 (Nס845), we examined the difference Diagnosis of Dementia
between diphenhydramine users and non-users in theirdistributions of age, sex, education, mental status The MoVIES clinical evaluation protocol followed (MMSE) scores, and depressive symptom (mCES-D) the diagnostic protocols established by CERAD9 and the scores, using Mann-Whitney tests (for continuous vari- University of Pittsburgh Alzheimer's Disease Research ables) and chi-square or Fisher's exact test (for categor- Center (ADRC). It included a standardized history, brief ical variables). The distributions of MMSE scores and general physical and detailed neurological exams, men- mCES-D scores were highly skewed, as would be ex- tal status exam, an informant interview, and a standard pected in a largely healthy community sample. There- laboratory panel (hematology, chemistry, serology).
fore, in addition to treating them as continuous vari- Clinical evaluations were carried out blind to subjects' ables for analyses using nonparametric tests, we also screening cognitive scores. Dementia was diagnosed ac- categorized them using cutpoints at the 10th percentile cording to DSM-III-R criteria.16 Diagnostically evaluated of our sample (MMSEՅ23) to identify the most cogni- subjects received a Clinical Dementia Rating (CDR)17 tively impaired tenth of our sample; and the 90th per- Am J Geriatr Psychiatry 11:2, March-April 2003 centile of the sample (Ն5 on the mCES-D), to identify 1, which shows combined prescription and OTC di- the most depressed tenth of our sample.
phenhydramine). The proportion of those taking seda- For multivariate analyses, logistic-regression models tive-hypnotic drugs increased at each consecutive wave, were applied to data from Wave 5 to examine the as- with a sharp increase in the most recent wave (2.2% at sociation between diphenhydramine use (dependent Wave 1 to 10.6% at Wave 5). This increase was mainly variable) and age, sex, education, MMSE, and mCES-D due to the increase in diphenhydramine use. At Wave scores, and subjective sleep complaints (independent 1, equal numbers (nס6 each) of subjects were taking variables). In the first model, the independent variables prescription and OTC diphenydramine; by Wave 5, 12 were age, sex, education, original recruitment in ran- and 57 subjects were taking prescription and OTC di- dom sample versus volunteer sample, MMSE score, and mCES-D score. In the second model, we added three At Wave 5, the cohort size was 845, representing insomnia complaints: initial, intermittent, and terminal attrition since Wave 4 of 13.6% from mortality, 2.7% due insomnia. We excluded from these analyses 50 subjects to permanent relocation and dropout, and 3.2% who who had incomplete data on one or more of the above skipped Wave 5 only. Of the 845, 657 were originally variables, reducing the sample size to 795 for these anal- recruited from the random sample and 188 from the volunteer sample. The mean (SD) ages of diphenhydra- We then fit the above logistic regression-models mine users and non-users were almost identical, at 80.1 again, excluding 143 subjects with dementia, defined as (4.6) and 80.5 (4.6) years, respectively, not significantly those with CDRՆ0.5, in case these subjects had "bot- different by Mann-Whitney test (pס0.436; Table 2).
tomed out" on the MMSE and were not susceptible to Women comprised 73.9% of users and 64.2% of non- further variation related to use of the drug. The sample users, but these proportions were not significantly dif- size for this analysis was therefore 652.
ferent by chi-square test (1 df; pס0.104). Users were Goodness of fit for the logistic-regression models significantly better educated: those with greater than was examined by use of the Hosmer-Lemeshow high school education comprised 62.5% of non-users goodness-of-fit test.17 All the models indicated ade- and 76.8% of users, a significant difference by chi-square Of the 845, 25 did not complete the MMSE, and 26 had incomplete mCES-D data. The mean (SD) MMSE scores of all 820 subjects, of the diphenhydramineusers, and the diphenhydramine non-users, were 26.6 A total of 1,681 subjects were enrolled at Wave 1 (base- (3.7), 26.5 (3.8), and 26.6 (3.7), respectively. No sub- line, study entry). Participants at subsequent waves jects with MMSE Յ18 were taking diphenhydramine.
numbered 1,342 (Wave 2), 1,165 (Wave 3), 1,016 (Wave Table 2 also shows the proportions of subjects with 4), and 845 (Wave 5). Subjects with incomplete medi- low MMSE scores (Յ23), high mCES-D scores (Ն5), cation information were deleted from the sample for and insomnia complaints: initial, intermittent, and ter- the present analyses; the total sample sizes used for this minal insomnia, among users and non-users of diphen- study at Waves 1 through 5 are, therefore 1,627, 1,338, hydramine. The table also summarizes the results of 1,164, 1,015 and 845. Mean (standard deviation [SD]) unadjusted and adjusted (multiple logistic regression) ages at each wave were 73.4 (5.9), 74.9 (5.5), 76.9 (5.3), analyses of associations of these variables with diphen- 78.8 (5.1), and 80.5 (4.6) years, respectively.
hydramine use. A diagnosis of dementia, with a CDR Table 1 shows the number of subjects taking score Ն0.5, was received by 143 individuals. Their sedative-hypnotic drugs, within prescription and non- mean (SD) MMSE score was 23.7 (3.5), whereas that of prescription (over-the-counter, OTC) categories at each the non-demented subjects was 27.7 (1.8).
wave. At these consecutive waves, use of prescription The results of the four regression models are sum- sedative-hypnotics (primarily benzodiazepines) was re- marized in Table 2 as odds ratios (OR) with 95% confi- ported in 1.8%, 2.5%, 2.0%, 1.3%, and 3.1% of the sam- dence intervals (CI) and p values derived from Wald chi- ples, respectively, and OTC sedative-hypnotics (primar- square tests with 1 df. In Model I, including all subjects, ily diphenhydramine) were reported as used by 0.4%, and not including sleep complaints, higher education 0.6%, 1.6%, 3.0%, and 7.6%, respectively (not in Table (OR: 2.2; pס0.021) and higher depression scores (OR: Am J Geriatr Psychiatry 11:2, March-April 2003 2.5; pס0.031) were independently associated with di- ble between 1987 and 1999, while the use of over-the- phenhydramine use. By contrast, in Model II, including counter sleep aids increased dramatically. Diphenydra- all subjects, with sleep complaints included as indepen- mine was the active ingredient in the majority of the dent variables, depressive symptoms were no longer sta- sedative-hypnotics used. This finding corresponds to na- tistically significant, but initial insomnia became statis- tional trends showing that use of nonprescription hyp- notics containing diphenhydramine or doxylamine has association with diphenhydramine use. In Model III, increased yearly since 1987, with an even greater in- limited to non-demented subjects, (i.e., excluding sub- crease in sleep aids containing antihistamines combined jects with CDRՆ0.5) and not including insomnia vari- ables, higher education (OR: 2.2; pס0.047), and lower We found no association of diphenhydramine use MMSE scores (OR: 6.7; pס0.005) were statistically sig- with age in our sample, with its relatively narrow age nificantly associated with diphenhydramine use. These range of 73.3 to 99.0 (mean age 80) years. Blazer et al.,1 two associations remained significant even after sleep similarly, found no relationship between age and complaints variables were introduced into the model sedative-hypnotic use in their aging community sample.
We also found no association with sex but did find one with higher educational level, even after adjusting for the effects of age and sex. Potential explanations in- DISCUSSION
clude a confounding association between higher edu- cation and higher income (reflecting greater purchasing Diphenhydramine has been available in the United power), or between higher education and propensity States since 1945, and its potential adverse effects were to self-medicate with nonprescription drugs, or perhaps reported as early as 1947.19 It has both antihistaminic even a greater receptivity to advertising and direct- and anticholinergic actions and is prescribed for the treatment of allergies, motion sickness, and Parkinson As did Blazer et al. with regard to sedative-hypnotics disease. It is also frequently used as a sleep aid, both by in general,1 we found that a greater number of depres- prescription and over the counter, because of its effi- sive symptoms were independently associated with the use of diphenhydramine in our cohort. However, the In our community-based sample, the frequency of addition of sleep complaints to the model revealed a prescription sedative-hypnotic drug use remained sta- significant association of diphenhydramine use with Prescription and nonprescription sedative-hypnotic use over 10 years, n (%)
1987-1989
1989-1991
1991-1993
1993-1996
1996-1998
Sedative-Hypnotic Drug
(N؄1,627)
(N؄1,338)
(N؄1,164)
(N؄1,015)
(N؄845)
aDiphenhydramine-containing products include: generic diphenhydramine, Benadryl, Benylin, Nervine, Aid-To-Sleep, Nytol, Robitussin PM, Tylenol PM, Motrin PM, acetaminophen PM, Excedrin PM, Legatrin PM.
bTotal number of drugs exceeds total number of subjects when one or more subjects reports taking more than one drug.
Am J Geriatr Psychiatry 11:2, March-April 2003 including
insomnia
complaints
Subjects,
excluding
Non-demented
Analyses
insomnia
complaints
(multivariate)
including
Adjusted
insomnia
complaints
(N
Subjects
excluding
I,
insomnia
complaints
(1996-1998)
5
Analyses
n
Non-Users
(n
Unadjusted
diphenhydramine
n
(n
associated
Complaints:
Insomnia
Am J Geriatr Psychiatry 11:2, March-April 2003 difficulty in falling asleep, whereas the association with caused delirium in mildly demented patients.32 A few depressive symptoms became non-significant. Thus, the studies have also suggested that anticholinergic drugs initial insomnia related to depression was the likely ba- cause cognitive deficits in patients with Parkinson dis- sis of the relationship between depression and diphen- ease, a group for whom such drugs are frequently pre- Our primary hypothesis was that the potential benefits of diphenhydramine as a sleep aid might be healthy older adults have yielded mixed results with re- overridden by its negative consequences for cognitive spect to cognitive impairment.36-39 In one study, de- functioning in these older adults. The cognitive effects tectable anticholinergic levels of commonly prescribed of diphenhydramine in older persons are usually attrib- medications were associated with impairments in mem- uted to its anticholinergic properties.23,24 A role may ory and attention in normal elderly subjects.40 In an- also be played by interactions mediated by cytochrome other study, older women appeared to suffer lesser im- oxidase enzyme-2D6 interactions, in which diphenhy- pairments than younger adults.41 In studies of younger dramine may raise the anticholinergic effects of other adults, diphenhydramine produced less consistent memory impairment than did scopolamine,42 but more In our overall sample, our initial hypothesis was not sedation and cognitive deficits than the newer, less confirmed. Although there was a twofold increased centrally-acting antihistamines.43-45 One study found probability of a low general mental status (MMSE) score that drowsiness and mental impairment had parallel in diphenhydramine users, even after adjusting for the slopes relative to diphenhydramine concentrations, al- possible confounding effects of depression, the associ- ation was not statistically significant. However, when Our study had some limitations. Medication use individuals with dementia were excluded from analysis, data were obtained by self-report from subjects/infor- the association between the MMSE and diphenhydra- mants and bottle labels. We had no access to medical mine became much stronger and statistically significant.
records or other sources of information regarding what One possible interpretation is a floor effect, that is, that the participants' physicians may have prescribed and individuals with dementia had sufficient brain dysfunc- for which indications. However, the bulk of the re- tion that their background cognitive impairment was ported diphenhydramine use in our study sample was not substantially affected by diphenhydramine use. The purchased over the counter. Our data on drug use is explanation may also partly be that the 143 demented presumed accurate as regards "regular" use as of the subjects included the 13 subjects with MMSE Ͻ18, time of assessment at Wave 5; we have no objective data none of whom were taking diphenhydramine.
on actual frequency or duration of use, but only 8 of Evidence from previous studies is mixed. We are the 69 users at Wave 5 had reported being users at Wave not aware of any previous population-based studies on 4. The number of users provided sufficient power for the relationship between cognitive functioning and hypothesis-testing only at Wave 5. Thus, our analyses chronic use of diphenhydramine or other anticholiner- are cross-sectional in nature and do not permit the de- gic drugs. Results are difficult to compare across studies termination of the direction of the associations we have because of variation in study population (most often reported here. For example, we cannot state whether patients or volunteers), subjects' age, study design (ob- diphenhydramine use led to cognitive impairment or servational or experimental), and the cognitive mea- vice versa. Power may also have been insufficient to detect small effects because, in this population-based Previous studies of clinical samples have revealed sample, the proportion of individuals with significant clear associations of anticholinergic drug levels or an- depressive symptoms and cognitive impairment was ticholinergic activity with delirium or diminished cog- low. However, the population-based nature of the sam- nitive/functional ability in surgical,26 medical,27,28 nurs- ple enhances its generalizability to the community at- ing home,29 and psychiatric (depressed) patients.30 large, since it is less affected by selection bias than sam- Although diphenhydramine was found beneficial in one ples of patients or volunteers on whom many previous treatment study of non-cognitive behavioral distur- bances in dementia,31 another report found that as low Diphenhydramine use as a sleep aid is on the rise a single dose as 25 mg-50 mg of diphenhydramine among older adults. Many users of nonprescription Am J Geriatr Psychiatry 11:2, March-April 2003 sleep aids, and indeed many physicians, may be un- collection and management, and the MoVIES study aware that diphenhydramine is an ingredient of these participants for their cooperation. products. Its potential adverse effects on cognition This work was previously reported in part at the should be considered. Practitioners should be alert to 14th Annual Meeting of the AAGP, February 23-26, their patients' use of over-the-counter sleep aids, par- ticularly when cognitive impairment is present.
The work reported here was supported in part by Research Grant AG07562 from the National Institute on Aging, U.S. Department of Health and Human Ser- The authors thank Linda Wake for her assistance vices. It was performed at the University of Pittsburgh with the preparation of an earlier version of this (School of Medicine and Graduate School of Public manuscript, the staff of the MoVIES project for data References
1. Blazer D, Hybels C, Sinonsick E, et al: Sedative, hypnotic, and 16. American Psychiatric Association: Diagnostic and Statistical Man- antianxiety medication use in an aging cohort over 10 years: a ual of Mental Disorders, 3rd Edition, Revised. Washington, DC, racial comparison. J Am Geriatr Soc 2000; 48:1073-1079 2. Lasagna L:Over-the-counter hypnotics and chronic insomnia in 17. Hughes CP, Berg L, Danziger WL, et al: A new clinical scale for the elderly. J Clin Psychopharmacol 1995; 15:383-386 the staging of dementia. Br J Psychiatry 1982; 140:566-572 3. Sproule BA, Busto UE, Buckle C, et al: The use of nonprescription 18. Hosmer DW, Lemeshow S: Applied Logistic Regression. New sleep products in the elderly. Int J Geriatr Psychiatry 1999; 19. Barman MC: Danger with Benadryl: of self-medication and large 4. Ganguli M, Belle S, Ratcliff G, et al: Sensitivity and specificity for dementia of population-based criteria for cognitive impairment: 20. Rickels K, Morris RJ, Newman H, et al: Diphenhydramine in in- The MoVIES Project. J Gerontol Med Sci 1993; 48:M152-M161 somniac family practice patients: a double-blind study. J Clin 5. Ganguli M, Lytle M, Reynolds MD, et al: Random vs. volunteer selection for a community-based study. J Gerontol Med Sci 1998; 21. Kudo Y, Kurihara M: Clinical evaluation of diphenhydramine hy- drochloride for the treatment of insomnia in psychiatric patients: 6. Ganguli M, Seaberg EC, Belle S, et al: Cognitive impairment and a double-blind study. J Clin Pharmacol 1990; 30:1041-1048 use of health services in an elderly rural population: The MoVIES 22. Roehrs T, Zwyghuizen-Doorenbos A, Roth T: Sedative effects and Project. J Am Geriatr Soc 1993; 41:1065-1070 plasma concentrations following single doses of triazolam, di- 7. Ganguli M, Ratcliff G, Belle S, et al: Effects of age, gender, and phenhydramine, ethanol, and placebo. Sleep 1993; 16:301-305 education on cognitive tests in an elderly rural community sam- 23. Bartus RT, Dean RL, Beer B, et al: The cholinergic hypothesis of ple: norms from the Monongahela Valley Independent Elders Sur- geriatric memory dysfunction. Science 1982; 217(4558):408-414 vey (MoVIES). Neuroepidemiology 1991; 10:42-52 24. Molchan SE, Martinez RA, Hill JL, et al: Increased cognitive sen- 8. Folstein MF, Folstein SE, McHugh PR: "Mini-Mental State:" a prac- sitivity to scopolamine with age and a perspective on the sco- tical method of grading the cognitive state of patients for the polamine model. Brain Research Brain Research Reviews 1992; clinician. J Psychiatr Res 1975; 12:189-198 9. Morris JC, Heyman A, Mohs RC, et al: The Consortium to Estab- 25. Lessard E, Yessine MA, Hamelin BA, et al: Diphenhydramine alters lish a Registry for Alzheimer's Disease (CERAD), Part I: clinical the disposition of venlafaxine through inhibition of CYP2D6 ac- and neuropsychological assessment of Alzheimer's disease. Neu- tivity in humans. J Clin Psychopharmacol 2001; 21:175-184 26. Tune L, Holland A, Folstein M: Association of postoperative de- 10. Ganguli M, Gilby J, Seaberg E, et al: Depressive symptoms and lirium with raised anticholinergic drug levels. Lancet 1981; associated factors in a rural elderly population: The MoVIES Proj- ect. Am J Geriatr Psychiatry 1995; 3:144-160 Flacker JM, Cummings V, Mach JR Jr, et al: The association of Radloff LS: The CES-D Scale: a self-report depression scale forresearch in the general population. Applied Psychological Mea- serum anticholinergic activity with delirium in elderly medical patients. Am J Geriatr Psychiatry 1998; 6:31-41 12. Lasilla H, Stoehr G, Ganguli M, et al: Factors associated with the 28. Agostini JV, Leo-Summers LS, Inouye S: Cognitive and other ad- use of prescription medications in an elderly rural population: verse effects of diphenhydramine use in hospitalized older pa- The MoVIES Project. Ann Pharmacother 1996; 30:589-595 tients. Arch Intern Med 2001; 161:2091-2097 13. Stoehr G, Ganguli M, Seaberg EC, et al: Over-the-counter medi- 29. Tollefson GD, Montague-Clause J, Lancaster SP: The relationship cation use in a rural elderly community: The MoVIES Project. J of serum anticholinergic activity to mental status performance in an elderly nursing home population. J Neuropsychiatry Clin Neu- 14. Ganguli M, Reynolds CF, Gilby JE, et al: Prevalence and persis- tence of sleep complaints in a rural elderly community sample: 30. Nebes RD, Pollock BG, Mulsant BH, et al: Low-level serum anti- The MoVIES Project. J Am Geriatr Soc 1996; 44:778-784 cholinergicity as a source of baseline cognitive heterogenicity in 15. Ganguli M, Dodge HH, Chen P, et al: Ten-year incidence of de- geriatric depressed patients. Psychopharmacol Bull 1997; mentia in a rural elderly U.S. community population: The MoVIES 31. Coccaro EF, Kramer E, Zemishlany Z, et al: Pharmacologic treat- Am J Geriatr Psychiatry 11:2, March-April 2003 ment of non-cognitive behavioral disturbances in disturbances in 39. Lines C, Traub M, Raskin S, et al: Lack of sedative and cognitive elderly demented patients. Am J Psychiatry 1990; 147:1640-1645 effects of diphenhydramine and cyclobenaprine in elderly vol- 32. Tejera CA, Saravary SM, Goldman E, et al: Diphenhydramine- unteers. J Psychopharmacol 1997; ii:325-329 induced delirium in elderly hospitalized patients with mild de- 40. Tune L, Carr S, Hoag E, et al: Anticholinergic effects of drugs mentia. Psychosomatics 1994; 35:399-402 commonly prescribed for the elderly: potential means for assess- 33. Bedard MA, Pillon B, Dubois B, et al: Acute and long-term admin- ing the risk of delirium. Am J Psychiatry 1992; 149:1393-1394 istration of anticholinergics in Parkinson's disease: specific ef- 41. Berlinger W, Goldberg M, Reynold S, et al: Diphenhydramine: fects on the subcortico-frontal syndrome. Brain Cogn 1999; kinetics and psychomotor effects in elderly women. Clin Phar- 34. Nishiyama K, Momose T, Sugishita M, et al: Positron emission 42. Aarsland D, Larsen JP, Reinvang I, et al: Effects of cholinergic tomography of reversible intellectual impairment induced by blockade on language in healthy young women. Brain 1994; long-term anticholinergic therapy. J Neurol Sci 1995; 132:89-92 35. DeSmet Y, Ruberg M, Serdaru M, et al: Confusion, dementia, and 43. Tharion WJ, Mcmenemy DJ, Rauch T, et al: Antihistamine effects anticholinergics in Parkinson's disease. J Neurol Neurosurg Psy- on the central nervous system: cognitive performance and sub- jective states.Neuropsychobiology 1994; 29:97-104 36. Katz I, Sands LP, Bilker W, et al: Identification of medications that 44. Moscowitz H, Burns M: Effects of terfenadine, diphenhydramine, cause cognitive impairment in older people: the case of oxybu-tynin chloride. J Am Geriatr Soc 1998; 46:8-13 and placebo on skills performance. Cutis 1988; 42:14-18 37. Sands LP, Katz I, Doyle S: Detecting subclinical change in cog- 45. Rice V, Snyder H: The effects of Benadryl and Hismanal on psy- nitive functioning in older adults. Am J Geriatrc Psychiatry 1993; chomotor performance and perceived performance. Aviat Space 38. Sands L, Katz I, Difilippo S, et al: Identification of drug-related 46. Gengo F, Gabar C, Miller JK: The pharmacodynamics of diphen- cognitive impairment in older individuals: challenge studies with hydramine-induced drowsiness and changes in mental perfor- diphenhydramine. Am J Geriatr Psychiatry 1997; 5:156-166 mance. Clin Pharmacol Ther 1989; 45:15-21 Am J Geriatr Psychiatry 11:2, March-April 2003

Source: http://www.wpic.pitt.edu/research/dementia_epidemiology/movies/Manuscripts/Basu%202002%20Sedative.pdf