Abstract
Objective
The Anticholinergic Drug Scale (ADS) is a commonly used measure of anticholinergic exposure. This study describes an expanded and revised version of the ADS (rADS) and its relationship with cultured cell‐based serum anticholinergic activity (cSAA) and cognitive measures.
Study Participants
Adults aged 60 years and older with mild cognitive impairment (MCI), remitted major depressive disorder (rMDD), or both, participate in the Prevention of Alzheimer's Dementia with Cognitive Remediation plus Transcranial Direct Current Stimulation (PACt‐MD) study.
Study Design
Cross‐sectional investigation of data from the PACt‐MD study.
Measures
The rADS includes ratings for 1047 distinct products, about twice as many as the originally published scale; previously published ratings were revised for 40 drugs. Total rADS scores were calculated as sums of ratings of all drugs taken by participants; cSAA was measured in the participants' sera; cognitive performance included measures of executive function, language, processing speed, verbal memory, visuospatial memory, working memory, and an overall composite score.
Statistical Analysis
The relationship between rADS total scores and cSAA was examined using a Spearman rank correlation coefficient. Relationships between rADS total scores and cognitive performance measures were explored in multivariable linear regression models.
Results
The sample included 310 participants (mean [standard deviation] age: 72 (6) years; 61.6% were women, and 81.6% had MCI [with or without rMDD]). Total rADS scores were positively correlated with cSAA (Spearman's correlation coefficient: 0.178, p = 0.0016). Total rADS scores were not significantly associated with cognitive performance.
Conclusions
The revised scale is recommended as a replacement for the original ADS since it includes ratings for more drugs and was significantly, albeit weakly, associated with cSAA, similar to previous findings using the original ADS.
Keywords: anticholinergic medications, cognition, depression, mild cognitive impairment, muscarinic antagonists, polypharmacy
1. INTRODUCTION
Anticholinergic medications can cause a variety of adverse effects, including cognitive deficits, constipation, dry mouth, and blurred vision. The Anticholinergic Drug Scale (ADS) is a commonly used measure to quantify anticholinergic medication exposure. The ADS has been associated with anticholinergic‐attributable adverse outcomes in older adults. 1 , 2 The last article to include a full list of ADS ratings was published in 2006. 3 The 2006 study found a significant albeit weak association of ADS scores (r‐squared = 0.0741) with conventionally measured serum anticholinergic activity (SAA) in nursing home residents, with limited additional variance explained by attempting to adjust for dosages (r‐squared = 0.0747). Since then, the ADS has been expanded to include ratings for about twice as many drugs, and some ratings have been revised to reflect new or reappraised evidence. Updated versions of the ADS have been available upon request but have not been disseminated in the published literature.
A previous study reported on the relationship between cultured cell‐based serum anticholinergic activity (cSAA) assay measures and anticholinergic scales, including the original ADS, as well as cognitive outcomes in older adults with mild cognitive impairment (MCI), remitted major depressive disorder (rMDD), or both (rMDD + MCI). 4 The authors reported that total anticholinergic drug exposure as measured by the original ADS and the Anticholinergic Cognitive Burden (ACB) scale was significantly positively associated with cSAA levels (Kendall's tau [T] scores for the association between cSAA and ACB and ADS total scores were T = 0.251, p < 0.001 and T = 0.194, p < 0.001, respectively). 4 cSAA measurements have advantages over conventional SAA measurements due to the use of cells that stably express M1 receptors, as opposed to rat brain tissue that may express multiple muscarinic receptor types, using 96‐well microplates for high throughput, and pretreatment with perchloric acid to neutralize any effects of endogenous proteins in serum samples. 5
The purpose of the current study was to examine the relationship of the revised version of the ADS (rADS) with cSAA level and cognitive performance. We carried out the study in the same participants as the aforementioned study. 4 We also present changes in rating definitions in the rADS that provide better logical consistency with rating methods, emphasizing evidence of muscarinic receptor binding and incorporating anticholinergic adverse effects. We also provide justifications for changes in previously published ratings for select drugs. The primary analysis examined associations between the rADS and cSAA, while exploratory analyses examined associations between the rADS and cognitive measures.
2. METHODS
This study adheres to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines for reporting. 6
2.1. Participants
The study sample was drawn from the Prevention of Alzheimer's Dementia with Cognitive Remediation plus Transcranial Direct Current Stimulation (PACt‐MD) study. 7 This multicenter randomized controlled trial conducted in Toronto included participants with MCI, rMDD, or both, diagnosed based on the criteria of the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM‐5). Participants with MCI were at least 60 years old. Those with rMDD or rMDD + MCI were at least 65 years old. Participants with rMDD had a Montgomery Asberg Depression Rating Scale (MADRS) score of 10 or below. Of the original sample size of 379 participants, 344 had cSAA measures and 310 were included in the analysis after excluding 26 participants for incomplete or missing data; one participant for starting new medication between cognitive testing and blood sample collection; two participants for extremely outlying cSAA scores suggesting that they were invalid; four participants whose cognitive scores suggested they had dementia; and one participant who reported taking an unspecified antihistamine, since the rADS score could not be assigned with certainty. The sample size was determined to power the PACt‐MD trial. The study was approved by the Center for Addiction and Mental Health's Research Ethics Board (which acted as the board of record for all sites) and all participants provided informed consent.
2.2. Study design and measures
This is a cross‐sectional analysis using measures from the baseline study visit. The analyses used data on age, sex, education, medication use, cognitive measures, and diagnoses of MCI, rMDD, or rMDD + MCI.
2.2.1. Cultured cell‐based serum anticholinergic activity assay
The cSAA assay was utilized as previously described. 4 , 5 In brief, the assay uses cells that stably express rat M1 muscarinic receptors. Standard curves are based on the displacement of [3H]QNB at concentrations of atropine ranging from 0.0 to 500 pmol/mL in serum. A competitive inhibition model was fitted from standard curves and used to express the anticholinergic activity of test samples in pmol/mL atropine equivalents.
2.2.2. The revised Anticholinergic Drug Scale
The rADS includes ratings for 1047 distinct products, whereas the last published version of the ADS included ratings for 537 distinct products. 3 Ratings have been added for drugs observed in various study samples in which the ADS has been applied, and other anticholinergics identified from various sources identified while reviewing evidence to support ratings. Evidence was primarily reviewed by one investigator (RMC), and three geriatric psychiatrists (SC, BGP, and TKR) reviewed and provided input on the ratings. For 30 specific medications for which ratings were reconsidered in light of new definitions or evidence, evidence was summarized in a document and these investigators came to consensus on final ratings through written communication.
For the rADS, definitions of ratings were modified as follows to better reflect the approach to assigning ratings.
Level 0: No evidence of meaningful muscarinic receptor binding, or conflicting evidence that does not clearly support muscarinic receptor binding.
Level 1: Potentially anticholinergic as evidenced by receptor binding studies, but either muscarinic receptor binding is weak, or the drug rarely or never causes anticholinergic adverse effects.
Level 2: Evidence suggests meaningful muscarinic receptor binding, and occasionally causes anticholinergic adverse effects, usually at higher doses.
Level 3: Evidence suggests meaningful muscarinic receptor binding, and frequently causes anticholinergic adverse effects at typical doses.
The ratings retain a qualitative component, as the quality of available evidence varies widely across drugs. From a practical standpoint, most drugs associated with the common occurrence of clinically meaningful anticholinergic adverse effects that are also known to bind muscarinic receptors receive a rating of 3, while drugs that have some evidence of muscarinic receptor binding and occasionally cause adverse effects that might be related to this binding receive a rating of 2. The strength of muscarinic receptor binding also informed these ratings. If receptor binding data were inconsistent, which usually involved some studies suggesting weak muscarinic receptor binding and others suggesting none, we favored the findings from the majority of studies when assigning ratings.
Ratings for 40 drugs in the original ADS were updated based on evidence that had been published, identified, or reconsidered since the original ADS was developed. The original ADS was derived from the Clinician‐rated Anticholinergic Scale 8 ; it was a modified and expanded version of this scale, and the name was changed to ADS. A guiding principle in the original ADS was that the ratings from the Clinician‐rated Anticholinergic Scale should not be changed unless the authors felt that a rating strongly conflicted with the available evidence. In the current update, we did not use that same principle and revised ratings as needed for consistency with rADS rating definitions. Some new evidence also conflicts with the evidence available when original ratings were made. Many of these original ratings were based on the work of previous investigators who used an anticholinergic activity assay to evaluate whether drugs affected muscarinic receptor binding of radiolabeled QNB. 9 , 10 Subsequently, another group used similar methods to determine the anticholinergic activity of 107 medications commonly used in older adults, but used biologically relevant concentrations. 11 In the revised ADS, if the tests at biologically relevant concentrations did not confirm the anticholinergic activity identified previously, and there was no additional data on anticholinergic activity to inform the decision, we changed the drug's rating to 0, considering this inconsistent evidence that does not clearly support muscarinic receptor binding. In addition, several benzodiazepines had received ratings of 1 in the original ADS by extrapolating results for other benzodiazepines, because prior work had identified several benzodiazepines as anticholinergic but had not reported on anticholinergic activity for all benzodiazepines. 9 , 10 In the revised ADS, if there was no direct reporting of anticholinergic activity of a benzodiazepine, its rating was changed to 0. The entire rADS is provided in Table S1. Additional details of the rationale for changing ratings are provided in Table S2.
The revised scale also includes an indicator of whether drugs with a rating higher than 0 are quaternary ammonium compounds, which are hydrophilic and not expected to cross a well‐functioning blood–brain barrier. The ability of a drug to cross the blood–brain barrier and accumulate in the central nervous system (CNS) is affected by other factors, including molecule size, whether it is a p‐glycoprotein substrate, and the effects of disease and age on the integrity of this barrier. Thus, this indicator provides users with some but not all relevant information about the likelihood that a drug will penetrate the CNS and affect cognition. 12 , 13
2.2.3. Other drugs
Several other drug classifications were created for use as covariates in analyses of cognitive test results. We classified muscle relaxants, opioids, anticonvulsants, nabilone, benzodiazepines, and non‐benzodiazepine hypnotics (Z‐drugs) as sedatives if they had an ADS score of 0 or 1, to help account for sedative effects through non‐anticholinergic mechanisms. Cognitive‐enhancing drug (cholinesterase inhibitors or memantine) use was not included as a covariate because no PACt‐MD participants took them, by design. We considered including antipsychotic use as a covariate but did not include it because only one participant took an antipsychotic.
2.2.4. Cognitive measures
As previously described, all PACt‐MD participants completed the Mini‐Mental State Examination (MMSE), Montreal Cognitive Assessment (MOCA), and a battery of neuropsychological tests. 4 , 14 For seven participants who were missing scores in one or two cognitive domains, the composite scores were based on non‐missing values.
Z scores from these tests were combined to calculate an overall composite cognitive score. In addition, Z scores were calculated for specific domains of interest, including verbal memory, visual memory, working memory, processing speed, language, executive function, and visuospatial processing. 4 , 14 All Z scores were calibrated such that higher Z scores indicate better performance.
2.3. Statistical Analysis
Descriptive statistics were generated for all measures. The relationship between the rADS and cSAA was examined using linear regression models in which cSAA was the dependent variable and the total rADS score was the independent variable. Because residuals were highly skewed, linear regression models were discarded and a Spearman's rank correlation coefficient of the relationship of rADS scores with cSAA was used instead. Exploratory analyses of the relationship between total rADS scores and cognitive performance were conducted using linear regression models adjusted for age, sex, highest education (categorical), number of non‐anticholinergic sedative drugs, and MDD and MCI diagnosis. We had planned to conduct an additional analysis excluding quaternary ammonium compounds rated level 2 or 3, but no participant in our sample took these drugs. We did not correct alpha for multiple comparisons as our analyses of cognitive outcomes were considered exploratory due to the many factors that could affect cognition cross‐sectionally, particularly with a high proportion of participants diagnosed with MCI. Power analyses indicated >80% power to identify a correlation of 0.16 between rADS and cSAA and that linear regression models for cognitive outcomes with 11 predictors and r‐squares ranging from 0.1 to 0.4 had 89% to 97.5% power to detect r‐square differences of 0.03 attributable to rADS scores.
3. RESULTS
3.1. Sample characteristics
Table 1 presents the characteristics of the 310 participants in the analytic sample. They had a mean (standard deviation [SD]) age of 72 (6) years; 191 (61.6%) identified as women; 229 (73.9%) had completed a bachelor's, master's, or doctorate degree; and 153 (49.4%) had a diagnosis of MCI, 57 (18.4%) rMDD, and 100 (32.3%) rMDD+MCI. The range of rADS total scores was 0–7 and scores were skewed: 218 (70.3%) participants had a score of 0; 60 (19.4%) a score of 1; and 32 (10.3%) had scores of 2 or higher. Table 2 lists the drugs with an rADS score higher than 0 taken by the sample and the number of participants who took each. Other sedative drugs were common, taken by 63 (20.3%) participants, whereas only one participant took an antipsychotic.
TABLE 1.
Baseline characteristics of 310 participants included in the analyses.
| Age in years, mean (SD) | 72 (6) |
| Women, n (%) | 191 (61.6%) |
| Highest education, n (%) | |
| Less than 9th grade | 9 (2.9%) |
| Some high school (10th or 11th) | 9 (2.9%) |
| High school graduate | 29 (9.4%) |
| Some university (at least 1 year) | 34 (11.0%) |
| Bachelor's degree | 144 (46.5%) |
| Master's or doctoral degree | 85 (27.4%) |
| Qualifying diagnosis, n (%) | |
| MCI | 153 (49.4%) |
| MDD | 57 (18.4%) |
| MCI + MDD | 100 (32.3%) |
| MMSE score, mean (SD) a | 28.1 (1.7) |
| MoCA score, mean (SD) a | 24.6 (2.9) |
| Cognition overall composite Z‐score, mean (SD) | −0.81 (0.78) |
| Executive function Z‐score, mean (SD) | −0.62 (0.91) |
| Language Z‐score, mean (SD) | −0.85 (1.10) |
| Processing speed Z‐score, mean (SD) | −0.88 (1.17) |
| Verbal memory Z‐score, mean (SD) | −0.92 (1.24) |
| Visuospatial memory Z‐score, mean (SD) | −0.79 (1.27) |
| Working memory Z‐score, mean (SD) | −0.80 (0.93) |
| cSAA level, median (IQR) atropine equivalents (pmol/mL) | 1.14 (0.46, 2.24) |
| rADS total score, n (%) | |
| 0 | 218 (70.3%) |
| 1 | 60 (19.4%) |
| 2 | 8 (2.6%) |
| 3 | 12 (3.9%) |
| 4 | 7 (2.3%) |
| 5 or greater | 5 (1.6%) |
| Number of rADS level 1 drugs, n (%) | |
| 0 | 232 (74.8%) |
| 1 | 70 (22.6%) |
| 2 | 8 (2.6%) |
| Number of rADS level 2 drugs, n (%) | |
| 0 | 307 (99%) |
| 1 | 3 (1.0%) |
| Number of rADS level 3 drugs, n (%) | |
| 0 | 287 (92.6%) |
| 1 | 20 (6.5%) |
| 2 | 3 (1.0%) |
| Number of non‐anticholinergic sedative drugs, n (%) | |
| 0 | 247 (79.7%) |
| 1 | 46 (14.8%) |
| 2 | 14 (4.5%) |
| 3 | 3 (1.0%) |
| Antipsychotic use, n (%) | 1 (0.3%) |
Abbreviations: cSAA, cultured cell‐based serum anticholinergic activity; IQR, interquartile range; MCI, mild cognitive impairment; MDD, major depressive disorder; MMSE, Mini‐Mental State Examination, MoCA, Montreal Cognitive Assessment; rADS, Anticholinergic Drug Scale revised version; SD, standard deviation.
MoCA score was missing for one participant and MMSE score was missing for three participants.
TABLE 2.
Drugs received by the sample with revised ADS scores greater than 0.
| Medication | Number of participants taking | Original ADS score | Revised ADS score |
|---|---|---|---|
| Amitriptyline | 6 | 3 | 3 |
| Chlorthalidone | 2 | 1 | 1 |
| Citalopram | 6 | 0 | 1 |
| Clindamycin (oral) | 2 | 1 | 1 |
| Clomipramine | 1 | 3 | 3 |
| Cyclobenzaprine | 2 | 2 | 3 |
| Desipramine | 1 | 3 | 3 |
| Desloratadine | 1 | NA | 1 |
| Diphenhydramine | 3 | 3 | 3 |
| Doxepin | 3 | 3 | 3 |
| Escitalopram | 15 | 0 | 1 |
| Fentanyl | 1 | 1 | 1 |
| Fesoterodine | 3 | NA | 3 |
| Fluoxetine | 5 | 1 | 1 |
| Glycopyrronium inhaled | 2 | NA | 1 |
| Loratadine | 1 | 0 | 1 |
| Mirtazapine | 9 | 0 | 1 |
| Nortriptyline | 3 | 3 | 3 |
| Oxybutynin | 1 | 3 | 3 |
| Oxycodone | 2 | 1 | 1 |
| Paroxetine | 3 | 1 | 2 |
| Prednisone | 4 | 1 | 1 |
| Quetiapine | 1 | 0 | 3 |
| Ranitidine | 6 | 2 | 1 |
| Sertraline | 27 | 1 | 1 |
| Temazepam | 1 | 1 | 1 |
| Tolterodine | 1 | 3 | 3 |
| Tramadol | 1 | 1 | 1 |
| Triamterene | 1 | 1 | 1 |
| Trimipramine | 1 | 3 | 3 |
Abbreviations: ADS, original Anticholinergic Drug Scale; NA, not present in original ADS; rADS, revised Anticholinergic Drug Scale.
3.2. Relationship between rADS scores and cSAA
Figure 1 provides a scatterplot and the regression line of rADS total scores versus cSAA, and illustrates the presence of potentially influential outliers, including several with relatively high cSAA and low rADS scores. Spearman's correlation coefficient for the relationship between rADS scores and cSAA was 0.178 (p = 0.0016).
FIGURE 1.
Scatterplot of revised ADS total scores versus cSAA, with regression line and confidence limits†. †ADS revised total score values are jittered horizontally to better illustrate individual data points. ADS, Anticholinergic Drug Scale; cSAA, cultured cell‐based serum anticholinergic activity.
3.3. Relationship of ADS scores with cognitive performance
The results of models predicting cognitive performance with rADS scores are shown in Table 3. No significant relationships between rADS scores and cognitive performance were observed.
TABLE 3.
Results of models predicting cognitive domain performance with revised ADS total scores. a
| Dependent variables | Revised ADS total score parameter estimate | Standard Error | p‐Value |
|---|---|---|---|
| Overall composite Z‐score | 0.0036 | 0.0353 | 0.9193 |
| Executive function Z‐score | 0.0591 | 0.0452 | 0.1923 |
| Language Z‐score | 0.0331 | 0.0539 | 0.5398 |
| Processing speed Z‐score | 0.0159 | 0.0565 | 0.7781 |
| Verbal memory Z‐score | −0.0718 | 0.0603 | 0.2351 |
| Visuospatial memory Z‐score | −0.0374 | 0.0639 | 0.5582 |
| Working memory Z‐score | 0.0182 | 0.0452 | 0.6868 |
Abbreviations: ADS, Anticholinergic Drug Scale.
Models adjusted for age, sex, highest education (categorical), number of non‐anticholinergic sedative drugs, major depressive disorder, and mild cognitive impairment diagnosis.
4. DISCUSSION
This study found that rADS total scores were positively, albeit weakly, correlated with cSAA but not associated with cognitive performance. This work builds upon older studies which found that ADS scores were significantly associated with SAA but explained a small amount of variance. 3 , 15 This result also extends the findings of prior work correlating ADS scores with cSAA 4 to the rADS. They found significant positive correlations between cSAA and the original ADS (T = 0.194, p < 0.001) and ACB scale scores (T = 0.251, p < 0.001). Although the ADS revisions did not result in meaningful differences in the strength of association of total scores with cSAA in this study sample, the much more comprehensive list of rated medications in the rADS offers advantages for broader application, and the rationale for rating changes has been provided here.
Some outliers influenced model fit and the validity of linear regression models comparing rADS scores to cSAA, leading to the use of a Spearman rank correlation coefficient to examine the association of revised ADS scores with cSAA. It is notable that many participants had detectable and sometimes relatively high cSAA values but were not taking drugs classified as anticholinergic (Figure 1). This may be due to the rADS classifying some drugs with anticholinergic activity as not anticholinergic. It could also be explained by endogenous anticholinergic activity. Detectable SAA has been observed in people taking only drugs without detectable anticholinergic activity, as have increases in SAA during acute illness. 16 This has been hypothesized to be related to inflammatory cytokines, cortisol, endogenous proteins known to inhibit muscarinic binding, or other factors. 16 , 17 , 18 Although the cSAA assay uses perchloric acid pretreatment to prevent interference from plasma proteins, 5 this may not completely eliminate the effects of endogenous anticholinergic substances.
Cognitive performance was not associated with rADS scores. Many factors could affect cognitive performance in this sample of older adults with MCI and rMDD, including severity of MCI, premorbid cognitive function, and other drugs or health conditions. We considered these analyses exploratory since they are from cross‐sectional data. Prior work in these study participants did not identify significant relationships between the original ADS scores and cognition, though they found that cSAA levels were negatively associated with executive function performance. 4 The models were similar except that education was treated as a continuous rather than categorical variable, and they did not include a covariate for sedative drug exposure.
This study has several other limitations. Although extensive research was conducted to support the rADS ratings, we did not perform a systematic review; scores were assigned over many years with evolving evidence, and rADS ratings retain a qualitative component. We used an existing sample, and many of the drugs rated in the rADS were not taken by the participants in this sample. In addition, a large proportion of the participants (70%) had a total rADS score of 0, possibly due to increasing awareness of the risks of using drugs with anticholinergic properties in this population of older adults. We did not consider medication dosage or pharmacokinetics. Both factors are likely associated with cSAA. However, the ADS is typically applied without adjustment for these factors, and prior work that attempted to account for dosage found that it did not explain additional variation in SAA. 3 Finally, this study was conducted in a convenience sample of trial participants with MCI or rMDD who are at risk of cognitive impairment, in whom use of rADS levels 2 and 3 drugs was infrequent. They are not representative of older adults more generally.
Since the original ADS was developed, anticholinergic classification systems have proliferated in the literature. Prior work has highlighted that ratings on these scales, many of which use a 0–3 ordinal rating system, are the same for many drugs but different for others. A 2021 study compared 10 scales and found kappa values for agreement ranging from −0.175 to 0.708 when applied in a sample of 473 older adults with complex conditions. 19 A 2021 systematic review noted that Spearman correlations of ADS ratings with those from 10 other scales ranged from 0.5 to 0.84 when considering only drugs rated by both scales, but also noted that the number of drugs included in some scales was as low as 27. 20 The ADS included the most drugs and 103 that were unique to it, and the total number of included drugs nearly doubled in the rADS. Associations of scale scores with anticholinergic‐related adverse outcomes vary across studies. A 2018 systematic review of reviews highlighted both positive and negative findings for each of six anticholinergic drug burden scales with studies examining associations with health outcomes, including the ADS. 2 This is to be expected with the strengths and weaknesses of different designs used, mostly cross‐sectional and cohort studies, and challenges in establishing causal effects of exposure to medications from a variety of classes that share a pharmacologic effect, in contrast to comparisons of specific drugs within classes. For example, exchangeability of users and nonusers is difficult to establish, with confounding by indication posing a major challenge, and many studies are vulnerable to prevalent user bias, or reverse causality in the case of cross‐sectional studies. Given the challenges and limitations of research on this topic, it is difficult to draw conclusions about the comparative validity of the various scales assessing anticholinergic properties of drugs. However, we believe the ADS has some conceptual strengths, which extend to this revised version. One is that evidence of muscarinic receptor binding or other preclinical evidence of impaired cholinergic neurotransmission was required to consider a drug anticholinergic, rather than only clinical effects that might be caused by anticholinergic effects or other mechanisms, or other factors influencing expert opinion. Another is the explicit zero ratings which are not universal across scales but allow users to know whether evidence was at some point reviewed to assess whether that drug has anticholinergic properties. Optimal application of the ADS would include assigning ratings for not yet reviewed drugs received by a study sample using the aforementioned classifications. Limitations of the ADS include lack of consideration of dose and crude potency classifications which, while common among anticholinergic scales, may not accurately represent the relative anticholinergic effects of medications. In addition, we cannot recommend a specific threshold score above which anticholinergic burden becomes concerning. For clinical application, we suggest that levels 2 and 3 drugs should be scrutinized when considering risk of anticholinergic side effects, particularly in vulnerable patients such as older adults with cognitive disorders. We also cannot rule out that accumulation of level 1 drugs might result in clinically meaningful anticholinergic adverse effects. In research applications, we not only suggest researchers consider total scores but also examine levels 1, 2, and 3 drugs separately, since their relative effects may differ. In prior work, levels 1, 2, and 3 ratings were adequately consistent with regression coefficients in a model that regressed counts of drugs in each of these categories on traditionally measured SAA. 3 However, we do not believe that assures that the anticholinergic effects of three level 1 drugs are equivalent to that of one level 3 drug, for example. To advance the utility of the ADS in quantifying anticholinergic effects of medications, future work might reexplore the utility of accounting for drug dose or even factors that might influence the pharmacokinetics of medications.
In conclusion, this study identified a significant but weak relationship between total scores from the rADS, an expanded and revised version of the ADS, and cSAA values in a sample of older adults with MCI, rMDD, or both. Although exploratory analyses did not identify relationships between rADS scores and cognitive performance, the rADS includes ratings for almost twice as many drugs, as well as revised ratings that we believe better classify select medications. Therefore, with future studies to extend the use of rADS to other populations including participants with major neurocognitive disorder, the rADS could serve as a replacement to the original ADS.
CONFLICT OF INTEREST STATEMENT
TKR has received in‐kind equipment support for an investigator‐initiated study from Magstim, and in‐kind research accounts from Scientific Brain Training Pro. BGP received research support from the Peter & Shelagh Godsoe Endowed Chair in Late‐Life Mental Health, CAMH Foundation, and Discovery Fund, National Institute of Aging, Brain Canada, the Canadian Institutes of Health Research, the Alzheimer's Drug Discovery Foundation, and the Ontario Brain Institute. He receives honoraria from the American Geriatrics Society for book authorship. BGP, RR, and JNN have Canadian provisional No. 3,054,093. and BGP, RR, JNN, and TKR hold US provisional patent 16/490,680 for cell‐based assay and kits for assessing serum anticholinergic activity, status pending. SK reports nonfinancial support from Soterix Medical unrelated to this work. KLL receives grants from the Alzheimer's Association, Alzheimer's Drug Discovery Foundation, the Canadian Institutes of Health Research, and the Weston Brain Institute, and has a research contract with Cerevel Therapeutics. She has received personal fees from Boehringer Ingelheim, Bright Minds, Bristol Meyers Squibb, Cerevel Therapeutics, Eisai Co. Ltd., Exciva, Ironshore Pharmaceuticals, Kondor Pharma, H Lundbeck A/S, Novo Nordisk, and Praxis Therapeutics, unrelated to the current work. LM reports nonfinancial support from Brainsway Ltd. unrelated to this work, personal fees from Otsuka, none of which were for the current work. BHM holds and receives support from the Labatt Family Chair in Biology of Depression in Late‐Life Adults at the University of Toronto; he currently receives or has received within the past 3 years research support from Brain Canada, the Canadian Institutes of Health Research, the CAMH Foundation, the US National Institute of Health (NIH), Capital Solution Design LLC (software used in a study founded by CAMH Foundation), and HAPPYneuron (software used in a study founded by Brain Canada). RMC, SC, NA, WW, CEF, AJF, and NH declare no conflicts of interest. RMC is a statistical scientific editor for Pharmacotherapy.
Supporting information
Appendix S1.
ACKNOWLEDGMENTS
This work was supported by Brain Canada through the Canada Brain Research Fund, with the financial support of Health Canada and the Chagnon Family; and by the Centre for Addiction and Mental Health Disovery Fund. The sponsor had no role in the design of the study, collection or analysis of the data, or in writing the manuscript. This study used baseline data from the Prevention of Alzheimer's Dementia with Cognitive Remediation plus Transcranial Direct Current Stimulation (PACt‐MD) study (Clinical trials identifier: NCT02386670).
Carnahan RM, Chandramouleeshwaran S, Ahsan N, et al. Relationship of the revised anticholinergic drug scale with cultured cell‐based serum anticholinergic activity and cognitive measures in older adults with mild cognitive impairment or remitted depression. Pharmacotherapy. 2025;45:332‐340. doi: 10.1002/phar.70022
Ryan M. Carnahan and Susmita Chandramouleeshwaran should be considered joint first authors.
Bruce G. Pollock and Tarek K. Rajji should be considered joint senior authors.
DATA AVAILABILITY STATEMENT
Data from this clinical trial will be made available beginning October 23, 2026. To access the data and data dictionary, interested parties can send an email to pact.md@camh.ca with a summary of their research proposal. The publication committee will review all requests and grant access to the data to qualified researchers who submit a methodologically sound proposal.
REFERENCES
- 1. Lisibach A, Benelli V, Ceppi MG, Waldner‐Knogler K, Csajka C, Lutters M. Quality of anticholinergic burden scales and their impact on clinical outcomes: a systematic review. Eur J Clin Pharmacol. 2021;77(2):147‐162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Welsh TJ, van der Wardt V, Ojo G, Gordon AL, Gladman JRF. Anticholinergic drug burden tools/scales and adverse outcomes in different clinical settings: a systematic review of reviews. Drugs Aging. 2018;35(6):523‐538. [DOI] [PubMed] [Google Scholar]
- 3. Carnahan RM, Lund BC, Perry PJ, Pollock BG, Culp KR. The anticholinergic drug scale as a measure of drug‐related anticholinergic burden: associations with serum anticholinergic activity. J Clin Pharmacol. 2006;46(12):1481‐1486. [DOI] [PubMed] [Google Scholar]
- 4. Chandramouleeshwaran S, Ahsan N, Raymond R, et al. Relationships between a new cultured cell‐based serum anticholinergic activity assay and anticholinergic burden scales or cognitive performance in older adults. Am J Geriatr Psychiatry. 2021;29(12):1239‐1252. [DOI] [PubMed] [Google Scholar]
- 5. Nobrega JN, Raymond RJ, Pollock BG. An improved, high‐efficiency assay for assessing serum anticholinergic activity using cultured cells stably expressing M1 receptors. J Pharmacol Toxicol Methods. 2017;86:28‐33. [DOI] [PubMed] [Google Scholar]
- 6. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453‐1457. [DOI] [PubMed] [Google Scholar]
- 7. Rajji TK, Bowie CR, Herrmann N, et al. Slowing cognitive decline in major depressive disorder and mild cognitive impairment: a randomized clinical trial. JAMA Psychiatry. 2025;82(1):12‐21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Han L, McCusker J, Cole M, Abrahamowicz M, Primeau F, Elie M. Use of medications with anticholinergic effect predicts clinical severity of delirium symptoms in older medical inpatients. Arch Intern Med. 2001;161(8):1099‐1105. [DOI] [PubMed] [Google Scholar]
- 9. 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(10):1393‐1394. [DOI] [PubMed] [Google Scholar]
- 10. Tune LE. Anticholinergic effects of medication in elderly patients. J Clin Psychiatry. 2001;62(Suppl 21):11‐14. [PubMed] [Google Scholar]
- 11. Chew ML, Mulsant BH, Pollock BG, et al. Anticholinergic activity of 107 medications commonly used by older adults. J Am Geriatr Soc. 2008;56(7):1333‐1341. [DOI] [PubMed] [Google Scholar]
- 12. Chancellor M, Boone T. Anticholinergics for overactive bladder therapy: central nervous system effects. CNS Neurosci Ther. 2012;18(2):167‐174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Araklitis G, Robinson D, Cardozo L. Cognitive effects of anticholinergic load in women with overactive bladder. Clin Interv Aging. 2020;15:1493‐1503. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Rajji TK, Bowie CR, Herrmann N, et al. Design and rationale of the PACt‐MD randomized clinical trial: prevention of Alzheimer's dementia with cognitive remediation plus transcranial direct current stimulation in mild cognitive impairment and depression. J Alzheimers Dis. 2020;76(2):733‐751. doi: 10.3233/JAD-200141 [DOI] [PubMed] [Google Scholar]
- 15. Carnahan RM, Lund BC, Perry PJ, Culp KR, Pollock BG. The relationship of an anticholinergic rating scale with serum anticholinergic activity in elderly nursing home residents. Psychopharmacol Bull. 2002;36(4):14‐19. [PubMed] [Google Scholar]
- 16. Flacker JM, Wei JY. Endogenous anticholinergic substances may exist during acute illness in elderly medical patients. J Gerontol A Biol Sci Med Sci. 2001;56(6):M353‐M355. [DOI] [PubMed] [Google Scholar]
- 17. Flacker JM, Lipsitz LA. Serum anticholinergic activity changes with acute illness in elderly medical patients. J Gerontol A Biol Sci Med Sci. 1999;54(1):M12‐M16. [DOI] [PubMed] [Google Scholar]
- 18. Hori K, Konishi K, Tani M, et al. Serum anticholinergic activity: a possible peripheral marker of the anticholinergic burden in the central nervous system in Alzheimer's disease. Dis Markers. 2014;2014:459013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Tristancho‐Pérez Á, Villalba‐Moreno Á, Santos‐Rubio MD, Belda‐Rustarazo S, Santos‐Ramos B, Sánchez‐Fidalgo S. Concordance among 10 different anticholinergic burden scales in At‐risk older populations. J Patient Saf. 2022;18(4):e816‐e821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Al Rihani SB, Deodhar M, Darakjian LI, et al. Quantifying anticholinergic burden and sedative load in older adults with polypharmacy: a systematic review of risk scales and models. Drugs Aging. 2021;38(11):977‐994. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix S1.
Data Availability Statement
Data from this clinical trial will be made available beginning October 23, 2026. To access the data and data dictionary, interested parties can send an email to pact.md@camh.ca with a summary of their research proposal. The publication committee will review all requests and grant access to the data to qualified researchers who submit a methodologically sound proposal.