Full text of DOI:10.1093/gerona/56.6.M353

Journal of Gerontology: MEDICAL SCIENCES
Copyright 2001 by The Gerontological Society of America
2001, Vol. 56A, No. 6, M353–M355
Endogenous Anticholinergic Substances May Exist
During Acute Illness in Elderly Medical Patients
Jonathan M. Flacker1 and Jeanne Y. Wei2
1The Hebrew Rehabilitation Center for Aged Research and Training Institute, Boston, Massachusetts.
2The Division on Aging, Harvard Medical School, Boston, Massachusetts.
Background. The purpose of this study was to determine if serum anticholinergic activity (SACA) arises from en-
dogenous substances produced during illness.
Methods. Elderly medical inpatients (N ꢀ 612) were screened for anticholinergic medication use in the week prior to
the study by interviews of subjects and proxies and review of emergency room, hospital, and nursing home medication
administration records. Of 24 subjects without a recent anticholinergic medication history, 15 were recruited and 10
completed the study. Serum samples were obtained on Day 2 of hospital admission. SACA was measured using a radio-
nuclide displacement assay. Medications taken by subjects were assayed for central muscarinic receptor binding at ther-
apeutic concentrations.
Results. Eight of the ten subjects had SACA detectable in the serum. No medication used by these subjects had anti-
cholinergic activity at usual therapeutic concentrations.
Conclusions. Endogenous anticholinergic substances may exist during acute illness. Characterization of such sub-
stances may increase the depth of our understanding of delirium and lead to useful intervention strategies.
HE neuropathophysiology of delirium in older persons
is incompletely understood. One leading hypothesis is
METHODS
This study was conducted at the Beth Israel Deaconess
Medical Center, a tertiary-care academic hospital in Boston,
Massachusetts, and was approved by the Clinical Investiga-
tions Committee of that facility. Between April 1 and Octo-
ber 31, 1997, all persons aged 70 years or older and who
were admitted to the medical service on a weekday were
screened for inclusion in the present study. Subjects were
excluded if there was any evidence, by history or medical
record documentation, of any anticholinergic medication
use in the week prior to admission. No subjects had partici-
pated in previous studies of SACA. Medication use was de-
termined by review of outpatient records, emergency room
and inpatient medication administration records, and inter-
views with subjects and family members that included ques-
tions regarding the use of over-the-counter medications and/
or alternative medicinal products. Nursing home medication
administration sheets were also reviewed for all nursing
home residents. Subjects were also excluded if they were on
dialysis, had received a blood transfusion in the week prior
to admission, or were on an unknown study medication. To
minimize the possibility of cumulative drug effects, subjects
on six or more medications were excluded. Because this
study was designed to assess whether SACA is detectable in
persons with no anticholinergic drug exposure, rather than
the relationship between SACA and delirium, formal data
on the mental status of subjects were not collected.
T
that delirium results from an impairment of central cholin-
ergic transmission (1,2). The assay for serum anticholin-
ergic activity (SACA), first introduced by Tune and Coyle
(3) almost 2 decades ago, has given support to the cholin-
ergic deficiency hypothesis. This assay measures the ability
of a subject’s serum to block central muscarinic receptors
using a rat forebrain preparation. Higher levels of SACA
have been associated with an increased likelihood of delir-
ium in both surgical (4,5) and medical (6,7) inpatients. A
dose-response relationship between symptoms of delirium
and SACA has also been suggested (7).
The origin of SACA has generally been assumed to be
with medications or their metabolites (3). Supporting this
hypothesis is the recognition that many commonly pre-
scribed medications, such as furosemide, digoxin, and meth-
ylprednisilone, have been shown to block cholinergic recep-
tors using this assay (8). If SACA is of pathophysiologic
importance in delirium and the origin of SACA is solely from
medications, then potential clinical interventions for SACA re-
duction are limited to medication discontinuations, exchanges,
and dose adjustments.
An alternative hypothesis is that, in addition to medi-
cations, elevated SACA may arise from endogenous sub-
stances. If this hypothesis is correct, then an understanding
of the origins and nature of the endogenous source(s) of
SACA could give new insights into potential treatments
for delirium. Accordingly, we sought to measure SACA
during acute illness in elderly medical inpatients carefully
screened to ensure that they had received no anticholin-
ergic medications.
For the purposes of this study, a medication was catego-
rized as potentially anticholinergic if that medication had (i)
well-documented in vivo anticholinergic effects, (ii) dem-
onstrated in vitro anticholinergic activity, or (iii) belonged
to a class of drugs with a member that has demonstrated in
M353

M354
FLACKER AND WEI
vitro anticholinergic activity. If there was any question as to
whether a medication might have anticholinergic activity,
the medication was categorized as anticholinergic. Medica-
tions used by subjects that were categorized as anticholin-
ergic in this study were as follows: codeine, compazine,
digoxin, diphenhydramine, dexamethasone, doxepin, famo-
tidine, fentanyl, fludrocortisone, furosemide, hydrocorti-
sone, hydroxyzine, hyoscyamine, ipratropium, isosorbide,
meclizine, meperidine, methylprednisilone, morphine, ni-
fedipine, nortriptyline, oxybutynin, oxycodone, prednisone,
procainamide, prochlorperazine, promethazine, quinidine,
trazadone, trilafon, triamcinilone, and triamterene.
Informed consent was obtained from subjects and/or their
proxies on the first day after admission. Six hundred and
twelve individuals were screened. Of these potential sub-
jects, 24 had no history of any anticholinergic medication
during the hospital admission or for the week prior to ad-
mission. Of these 24 subjects, 9 subjects declined participa-
tion and 15 subjects were recruited. The analysis, however,
excludes 5 subjects who consented but did not have a serum
sample drawn due to hospital discharge (2 subjects), missed
blood draw (2 subjects), and refusal (1 subject). Thus, com-
plete data were available for 10 subjects.
ꢃl sample (mean ꢀ 0.69 [SD ꢄ 0.52] nmol/l atropine equiva-
lents/200-ꢃl sample). Medications taken by these individuals
are listed in Table 1. None of the medications that were used
by these 8 subjects had demonstrable anticholinergic activity
as measured by the anticholinergic activity assay.
DISCUSSION
Although the present study does not address the patho-
physiologic relationship between SACA and delirium, it
does present evidence that SACA is detectable in ill elderly
persons who are receiving no anticholinergic medications.
This finding challenges the hypothesis that medications are
the sole source of SACA and adds to several lines of evi-
dence that other sources of SACA exist. One study of six
hospitalized older persons showed a decline in SACA with
a resolution of delirium that was seemingly unrelated to
medication changes (6). Another study has demonstrated
that SACA declined significantly following resolution of fe-
brile illness in elderly nursing home residents, and this de-
cline also was seemingly unrelated to medication changes
(10). Interestingly, despite the accepted belief and common
observation that anticholinergic medications may contribute
to delirium, some large epidemiological studies of delirious
patients found no association between anticholinergic medi-
cation use and delirium in either elderly medical (11,12) or
postoperative surgical patients (13). Although this finding
may be due in part to improper classification of medications,
the presence of endogenous anticholinergic substances in
those patients classified as not being exposed to anticholin-
ergic substances might also obscure this relationship.
Some naturally occurring substances have been shown to
have antimuscarinic activity in vitro. For example, dynor-
phin A and myelin basic protein have been shown to inhibit
binding at muscarinic receptors by altering the receptor con-
formation (14). Protamine may also inhibit binding at mus-
carinic receptors through a similar mechanism (15). Fur-
thermore, a naturally occurring inhibitor of antagonist
binding to cholinergic receptors has been described in hu-
man brains (16). It seems unlikely that such a protective
system would have evolved in the absence of any anticho-
linergic substance to inhibit. Thus, our conclusion that clini-
cally important endogenous anticholinergic substances exist
seems reasonable.
Data Collection
On the second morning following admission, a blood
sample was drawn in a 10-cc heparinized tube. The blood
was centrifuged, and the serum was separated and stored at
ꢁ80ꢂC. Our previous experience has shown that SACA lev-
els remain stable up to 1 year following collection when
stored in this fashion. SACA was measured using the com-
petitive binding radionuclide assay described by Tune and
Coyle (2). In this assay, a radio-labeled reversible binder of
the forebrain muscarinic receptor, 3H-quinuclidinyl benzi-
late (H3-QNB; New England Nuclear, Boston, MA), was
added to a Sprague-Dawley rat forebrain membrane prepa-
ration. Separate samples of increasing dilutions of atropine
were assayed, and displacement of the H3-QNB was mea-
sured by counting residual radioactivity after each prepara-
tion was washed with buffer. The assay was then repeated,
substituting the subject’s serum for atropine to determine
the equivalents of atropine that would be as anticholinergic
as the subject’s serum. All controls and samples were run in
duplicate. The rat forebrain preparation was custom pre-
pared and stored in phosphate buffer at ꢁ80ꢂC (Analytic
Biological Services, Wilmington, DE). All medications that
were used by subjects who were found to have detectable
SACA levels were diluted to therapeutic serum concentra-
tions (9) in the phosphate buffer system and tested for anti-
cholinergic activity. The SACA testing on all subjects’ se-
rum was performed at the same time to reduce the effect of
interassay variability. The intraassay variability was 6.7%.
The SACA testing on all medications also was performed
together, but on a different day than the assay of the sub-
jects’ serum.
Table 1. Medication Use in Subjects With Serum Anticholinergic
Activity (SACA)
Age
(years) Gender Diagnosis
Medications
SACA†
87
75
94
78
F
F
F
F
Pneumonia
Pneumonia
Pneumonia
Urosepsis
Cefuroxime, insulin, glyburide
Erythromycin, cefuroxime
Cefuroxime
Ampicillin, aspirin, gentamicin,
metoprolol
1.72
1.20
0.73
0.60
81
77
F
F
C. Difficile
Delirium
Lorazepam, potassium, vancomycin
Ampicillin, ducosate, lorazepam,
L-thyroxine, valproate
0.41
0.38
94
85
F
F
Dehydration None
C. Difficile
0.24
RESULTS
Aspirin, metronidazole, multivitamin 0.23
SACA was present in 8 of the 10 subjects who had taken
no anticholinergic medications. SACA in these 8 subjects
ranged from 0.23–1.72-nmol/l atropine equivalents per 200-
Note: SACA ꢀ serum anticholinergic activity; F ꢀ female.
†Units of SACA are nmol/l atropine equivalents per 200-ꢃl sample.

ENDOGENOUS ANTICHOLINERGIC SUBSTANCES
M355
Acknowledgments
This study has several limitations. First, the serum anti-
cholinergic activity assay itself is not standardized. Previous
studies have not reported inter- and intraassay variability,
but in our laboratory, this has been up to 20% and 10%, re-
spectively, despite our efforts at standardization. In addi-
tion, we cannot completely exclude the possibility that a
metabolite of the medication(s) used by these subjects was
the source of anticholinergic activity. However, this is very
unlikely given the fact that the medications used by these
subjects were demonstrated not to have any anticholinergic
activity. Furthermore, one subject who received no medica-
tions at all still had detectable SACA. It is also important to
note that although there was no gender exclusion for this
study, all the subjects were women. Finally, while in-hospi-
tal medication use was obtained from medical records, self
and/or proxy-reported medication use was used in many
cases to identify the medications that were taken prior to ad-
mission. However, as blood samples were taken on Day 2 of
admission to the hospital, a minimum of 36 hours had
passed since the most recent possible ingestion of any po-
tentially anticholinergic medication.
It is also important to point out that “normal” levels of
SACA have not been defined, but it is generally assumed
that under conditions of health and no medication use, hu-
mans should have no detectable SACA. Systematic assess-
ments of SACA in otherwise healthy individuals not receiv-
ing medications have not been performed. Although the
presence of circulating endogenous anticholinergic sub-
stances in healthy young or older individuals would be a
significant finding on its own, no data exist to support such
a hypothesis. One study that measured SACA during rela-
tive good health 1 month following a febrile illness in older
persons found a mean SACA of 0.09- nmol/l atropine
equivalents per 200-ꢃl sample (substantially lower than the
SACA levels found in the present study), and no SACA in
14 of 22 subjects (63%), although this study did not control
for medication use (9).
This study was supported by the Hebrew Rehabilitation Center for Aged
Research and Training Institute and the National Institute on Aging (Grants
AG08812, AG04390, and AG05134). Dr. Flacker was supported in part by
a fellowship from the Medical Foundation Charles A. King Trust.
Address correspondence to Jonathan M. Flacker, MD, Grady Geriatrics
Center, Emory University Division of General Medicine, 69 Butler Street
S.E., Atlanta, GA 30303. E-mail: jflacke@emory.edu
References
1. Trzepacz P. The neuropathogenesis of delirium. Psychosomatics.
1994;4:374–391.
2. Flacker JM, Lipsitz LA. Neural mechanisms of delirium: current hypoth-
eses and evolving concepts. J Gerontol Biol Sci. 1999;54A:B239–B246.
3. Tune L, Coyle JT. Serum levels of anticholinergic drugs in treatment
of acute extrapyramidal side-effects. Arch Gen Psychiatry. 1980;37:
293–297.
4. Tune L, Holland A, Folstein MF, Damlouji NF, Gardner TJ, Coyle JT.
Association of postoperative delirium with raised serum levels of anti-
cholinergic drugs. Lancet. 1981;2:651–652.
5. Golinger RC, Peet T, Tune LE. Association of elevated plasma anticho-
linergic activity with delirium in surgical patients. Am J Psychiatry.
1987;144:1218–1220.
6. Mach JR, Dysken MW, Kuskowski M, Richelson E, Holden L, Jilk
KM. Serum anticholinergic activity in hospitalized older persons with
delirium: a preliminary study. J Am Geriatr Soc. 1995;43:491–495.
7. Flacker JM, Cummings V, Mach J, Bettin K, Kiely DK, Wei J. Serum
anticholinergic activity is associated with delirium in ill elderly per-
sons. Am J Geriatr Psychiatry. 1998;6:31–41.
8. Tune L, Carr S, Hoag E, Cooper T. Anticholinergic effects of drugs
commonly prescribed for the elderly: potential means for assessing
risk of delirium. Am J Psychiatry. 1992;149:1393–1394.
9. McEvoy GK, ed. American Hospital Formulary Service 1997 Drug
Information. Bethesda, MD: American Society of Health-System
Pharmacists; 1997.
10. Flacker JM, Lipsitz LA. Serum anticholinergic activity changes with
acute illness in elderly medical patients. J Gerontol Med Sci. 1999;
54A:M12–M16.
11. Schor JD, Levkoff SE, Lipsitz LA, et al. Risk factors for delirium in
hospitalized elderly. JAMA. 1992;267:827–831.
12. Francis J, Martin D, Kapoor WN. A prospective study of delirium in
hospitalized elderly. JAMA. 1990;263:1097–1101.
13. Marcantonio ER, Juarez G, Goldman L, et al. The relationship of post-
operative delirium with psychoactive medications. JAMA. 1994;272:
1518–1522.
Although the SACA assay was initially developed to eval-
uate the anticholinergic effects of medications, its applica-
tion to the syndrome of delirium has led to important insights
regarding that syndrome’s pathophysiology. Although medi-
cations are undoubtedly a source of anticholinergic activity
in the serum in some cases, the present study suggests that
clinically important endogenous anticholinergic substances
may also exist during acute illness. Further characterization
of such substances may increase the depth of our understand-
ing of delirium and lead to new treatment options for those
suffering from this common, morbid, and costly complica-
tion of illness.
14. Hu J, El-Fakahany EE. Allosteric interaction of dynorphin and myelin ba-
sic protein with muscarinic receptors. Pharmacology. 1993;47:351–359.
15. Hu J, Wang S, Foray C, El-Fakahany EE. Complex allosteric modula-
tion of cardiac muscarinic receptors by protamine: potential model for
putative endogenous ligands. Mol Pharamacol. 1992;42:311–324.
16. Frey WH, Emory CR, Wiebenga ME, et al. Inhibitor of binding to
muscarinic receptor is elevated in Alzheimer’s brain. Brain Res. 1994;
655:153–160.
Received May 6, 2000
Accepted June 1, 2000
Decision Editor: John E. Morley, MB, BCh