Association between anticholinergic medication uses and the risk of pneumonia in elderly adults: a meta-analysis and systematic review

Abstract

Objective

To conduct a meta-analysis and systematic review on the association between anticholinergic medication uses and the risk of pneumonia in elderly adults.

Materials and Methods

Medical databases were searched included PubMed, Web of Science, EBSCO and Google Scholar (up to December 7, 2022). Studies evaluating association between anticholinergic medication uses and the risk of pneumonia in elderly adults were included. Studies without available data were excluded. We made meta-analysis by using adjusted odds ratio (aOR) with 95% confidence intervals (CIs) from random-effects model. The risk of bias was assessed using ROBINS-I tool and statistical heterogeneity using the I² statistic. Registration: INPLASY202330070.

Results

A total of six studies with 107,012 participants were included. Meta-analysis results showed that anticholinergic medication uses was related with an increased risk of pneumonia (aOR = 1.59; 95%CI, 1.32–1.92) and stroke-associated pneumonia (aOR = 2.02; 95%CI, 1.76–2.33). Moreover, risk estimates of pneumonia for high-potency anticholinergics (aOR = 1.96; 95%CI, 1.22–3.14) were higher than those for low-potency anticholinergics (aOR = 1.58; 95%CI, 1.27–1.97). And increased risk of pneumonia was associated with the anticholinergic medication uses within 30 days (aOR = 2.13; 95%CI, 1.33–3.43), within 90 days (aOR = 2.03; 95%CI, 1.26–3.26) and chronic use (aOR = 1.65; 95%CI, 1.09–2.51).

Conclusions

The risk of pneumonia is increased in elderly adults with anticholinergic medication, especially with higher-potency anticholinergic drugs and in the initiation phase of anticholinergic medication. Clinicians should monitor their use in older patients carefully, especially when the pneumonia-related signs and symptoms are identified.

Key message

• Anticholinergic medication could increase the risk of pneumonia in elderly adults.

• The risk of pneumonia was higher in the initiation phase of anticholinergic medication and when the older patients was medicated with higher-potency anticholinergic drugs.

• Clinicians should monitor anticholinergic drugs use in older patients carefully, especially when the pneumonia-related signs and symptoms are identified.

Keywords:

• Anticholinergic medication
• pneumonia
• elderly adults
• meta-analysis
• systematic review

Introduction

Pneumonia is one of the most common diseases and is a leading cause of hospitalizations among the elderly adults [1]. The risk factors for pneumonia are manifold. Age is one of the strongest risk factors for pneumonia. Previous studies have reported that pneumonia incidence rates increased dramatically by age group among older population (9.9 in 65–74 years, 16.9 in 75–84 years and 29.4 in ≥85 years) [2,3]. Older individuals usually have several coexisting diseases which are also risk factors for pneumonia such as chronic cardiovascular disease, Parkinson disease, dysphagia, epilepsy, chronic obstructive disease [3]. Therefore, they have a tendency for polypharmacy.

Recently, more and more attention has been paid to whether medication uses contributes to the higher risk of pneumonia in elderly people, except for the known risk factors. Anticholinergic drugs are frequently prescribed to the elderly adults. Anticholinergic medication includes drugs from a wide range of therapeutic categories which are used in a variety of diseases. Drugs with anticholinergic properties account for 23% of the ambulatory patients and nearly 60% of nursing home residents. They nonselective act on muscarinic receptor antagonist and thus cause many side effects centrally and peripherally. Pharmacokinetic and pharmacodynamic properties are changing related with age, therefore the older patients are more susceptible to anticholinergic drugs. Therefore, the American Geriatric Society 2019 have defined some anticholinergic drugs as inappropriate medications and should be avoided in elderly patients [4].

Because of the widespread use of anticholinergic medication, several epidemiological studies have investigated the association between anticholinergic medication uses and the risk of pneumonia [5–10]. These studies had been heterogeneous with regard to study design, methodologies, countries and the results were inconsistent. Overall, all of the previous investigations were observational cohort of case-control studies which lacked powerful statistical evidence to assess the relationship between the risk of pneumonia and anticholinergic medication. In the absence of randomized clinical trials, we therefore carried out present meta-analysis and systemic review to summarize the results of all existing studies and to ascertain the risk of pneumonia associated with exposure to anticholinergic medication uses.

Materials and methods

Our systematic review and meta-analysis were performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [11]. This study protocol was registered in INPLASY under registration number INPLASY202330070.

Data sources and search strategy

We performed a computerized and manual search by using the medical databases including PubMed, Web of science, EBSCO and Google Scholar with English-language restriction from inception to December 7, 2022. Meanwhile, we also checked the citations and bibliographies of included studies to identify any further pertinent studies. The detailed search strings for PubMed was provided in Table S1 (Supplemental Material), which was also adapted for searching other databases.

Inclusion and exclusion criteria

The studies were included if they fulfilled the following inclusion criteria: (I) Population: elderly adults who aged 65 or older. (II) Intervention: exposure to anticholinergic medication uses. (III) Comparator: no exposure to anticholinergic medication. (IV) Outcome: pneumonia. (V) Design: case-control study, cohort study or randomized controlled trial. (VI) Study data: results should be reported as adjusted odds ratio (aOR) or relative risk (aRR) and the corresponding 95% confidence interval (CI). (VII) Study language: only written in English.

Exclusion criteria: studies didn’t have available data for outcome measures.

Data extraction

Two investigators (Wu-Mindan and Li-Zhixuan) independently reviewed all retrieved studies met the inclusion criteria and performed data extraction with a standardized worksheet, including country, study period, study design, data source, mean participant age, sample size (number of case and control participants), proportion of female participants, selection criteria of each included studies, ascertainment of outcomes, ascertainment and exposure definition of anticholinergic medication use, anticholinergic central or general effects, effects on outcomes (aOR or aRR and the corresponding 95%CI) and confounder adjustment. Another two investigators (Zheng-Wenchuan, Zhuang-Jia) further verified the extracted information.

Risk of bias assessment

Two investigators (Wu-Mindan and Li-Zhixuan) independently assessed the risk of bias, which estimated causal effects for each eligible study, by using the Cochrane Risk of Bias in Non-randomized Studies of Interventions (ROBINS-I) tool [12] with seven assessment domains (confounding, selection of participants, classification of interventions, deviation from intended interventions, missing data, measurement of the outcomes and selection of the reported results). Each assessment domain level and overall judgement are graded as low, moderate, serious and critical risk of bias. Any discrepancies were resolved by discussion or consulting third reviewer until reaching a consensus.

Statistical analysis

We used aOR with 95%CI for the meta-analysis to assess the association between anticholinergic medication uses and the risk of pneumonia in elderly adults. Heterogeneity across the included studies was identified by using the I² statistic. I² values of 25%, 50%, and 75% was assigned as low, moderate, and high heterogeneity, respectively. If I²>50%, the Random-effect meta-analysis was performed. When high heterogeneity was found, sensitivity analysis would be conducted to determine which study contributed to the largest heterogeneity. Egger’s test and Begg’s test were used to assess potential publication bias. All meta-analyses were conducted by using Stata/SE 12.0 (StataCorp, College Station, TX) and Review Manager version 5.3 (The Cochrane Collaboration, London, United Kingdom). The significance level was set at a P value of <0.05.

Results

Studies identification and characteristic

A total of 2,514 studies were identified through our initial database search. After preliminary review and detailed full-text evaluation, six articles were included for narrative analysis. Eventually, six studies met our inclusion criteria for this meta-analysis [5–10]. Details were showed in Figure 1. The characteristics of each eligible study were summarized in Table 1. Our meta-analysis included four case-control studies and two cohort studies. Mean age of enrolled participants was between 71.7 ∼ 84. Three studies reported outcome as community-acquired pneumonia [5,6,9]. Lampela et al. [7] focused on pneumonia in persons with Alzheimer’s disease. Kose et al. [8] focused on aspiration pneumonia. Gradek et al. [10] focused on stroke-associated pneumonia.

Figure 1. PRISMA flow chart of literature search and study identification.

Table 1. General characteristics of included studies.

Author (year)	Country	Mean Age	Study Period	Sample Size, female%		Data Source	Study Design	Outcome	Setting
				Case	Control
Paul et al. (2015) [5]	Washingto, USA	77	2000-2003	1039, 49%	2022, 49%	Group Health Cooperation	Population-based case-control study	Community-acquired pneumonia	Community
Chatterjee et al. (2016) [6]	Texas, USA	76	2009-2010	291, 65.3%	1164, 65.3%	Regional Medicare Advantage Healthcare Plan	Nested case-control study	Community-acquired pneumonia	Community
Lampela et al. (2017) [7]	Finnish	83	2005-2011	12442, 52.3%	24349, 53.3%	Finnish Nationwide MEDALZ^1 Cohort	Nested case-control study	Pneumonia in Persons with Alzheimer’s Disease	Community
Kose et al. (2018) [8]	Japan	73-84	2010-2016	162, 67.3%	456, 63.4%	Hitachinaka General Hospital	Retrospective longitudinal cohort study	Aspiration pneumonia	Convalescent rehabilitation wards
Lee et al. (2020) [9]	Taiwan, China	73.5	2015-2016	32215, 53.77%	32215, 53.75%	Longitudinal Health Insurance Database	Case-control study	Pneumonia	Community
Gradek et al. (2022) [10]	Krakow, Poland	71.7	NONE	97, 53.6%	560, 53.0%	PROPOLIS^2	Hospital-based cohort study	Stroke-associated pneumonia	Department of Neurology, University Hospital, Krakow

MEDALZ: Cohort profile: The Finnish Medication and Alzheimer’s disease; PROPOLIS: Prospective Observational POLIsh Study on post-stroke delirium.

Risk of bias

The results of all assessment domains and overall risk of bias judgement were summarized in Table 2. The overall risk of bias judgement of each eligible study ranged from moderate to serious. In the evaluation of pneumonia as outcome, three studies [5,6,9] were rated as having a moderate risk of bias, another three [7,8,10] as serious. Only two studies [9,10] assessed stroke-associated pneumonia with moderate risk of bias. Of each risk of bias domains: (1) Although most studies had controlled for objectively measurable confounding factors (such as age, gender, respiratory disease, cardiac disease), there might still be some unknown confounding factors among all participants due to a retrospective design, which finally contributed to moderate risks of bias in confounding. Details of confounder adjustment were showed in Table S2. (2) Lampela et al. [7] mainly focused on the pneumonia in persons with Alzheimer’s disease; Kose et al. [8] focused on the aspiration pneumonia in elderly. Gradek et al. [10] focused on the stroke-associated pneumonia in elderly. These above three studies had a serious risk of bias in selection of participants. (3) All eligible studies defined the anticholinergic medication exposure by using Anticholinergic Drug Scale (ADS) [13] or Anticholinergic Risk Scale (ARS) [14,15], according the clinical records of prescription information, pharmacy dispensing records and self-reported interview from drug packets (further checked by interviewers), which were not easily affected by the outcome and existed objectively (details in Table S2). So, bias in classification of interventions for each study was assessed as moderate risk. (4) Many studies examined pneumonia objectively and the outcome assessors were unaware of the intervention received by study participants (details in Table S3), thus all studies were rated as having low risk of outcome measure. (5) All studies were retrospective study and lacked prospective analysis, which might cause possible selection of reported results (moderate risk).

Table 2. Summary of risk of bias assessment using ROBINS-I tool.

Author/year/country	Outcome	Bias domain							Overall risk of bias judgement
		Bias due to confounding	Bias in selection of participants	Bias in classification of interventions	Bias due to deviations from intended interventions	Bias due to missing data	Bias in measurement of outcomes	Bias in selection of the reported result
Paul/2015/USA	Pneumonia	Moderate	Moderate	Moderate	No information	Moderate	Low	Moderate	Moderate
Chatterjee/2016/USA	Pneumonia	Moderate	Moderate	Moderate	No information	Moderate	Low	Moderate	Moderate
Lampela/2017/ Finnish	Pneumonia	Moderate	Serious	Moderate	No information	Moderate	Low	Moderate	Serious
Kose /2018/ Japan	Pneumonia	Moderate	Serious	Moderate	No information	Moderate	Low	Moderate	Serious
Lee/2020/ China	Pneumonia	Moderate	Moderate	Moderate	No information	Moderate	Low	Moderate	Moderate
	Stroke-associated pneumonia	Moderate	Moderate	Moderate	No information	Moderate	Low	Moderate	Moderate
Gradek/2022/ Poland	Pneumonia	Moderate	Serious	Moderate	No information	Moderate	Low	Moderate	Serious
	Stroke-associated pneumonia	Moderate	Moderate	Moderate	No information	Moderate	Low	Moderate	Moderate

ROBINS-I tool: the Cochrane Risk of Bias in Non-randomized Studies of Interventions tool.

Results of meta-analysis

Anticholinergic medications and the risk of pneumonia

A total of 107,012 participants were selected from six studies which reported the association between anticholinergic medication uses and the risk of pneumonia [5–10]. Meta-analysis results showed that anticholinergic medication uses was related with an increased risk of pneumonia by using random-effects model (aOR = 1.59; 95%CI, 1.32–1.92; Figure 2(A)). The statistical heterogeneity was high (I²=86.6%, p = 0.000; Figure 2(A)). Further sensitivity analysis didn’t find out any study which contributed to the largest heterogeneity. The p value of Egger’s test and Begg’s test results were 0.452 and 0.975 respectively, which indicated there were no potential publication bias (Table S4).

Figure 2. Forest plots of the association between anticholinergic medications and the risk of pneumonia in elderly adults.

Subgroup meta-analyses

Anticholinergic medications and the risk of stroke-associated pneumonia

There were only two studies investigated the relationship between anticholinergic medication uses and the risk of stroke-associated pneumonia [9,10]. The meta-analysis demonstrated that anticholinergic medication uses also increased the risk of stroke-associated pneumonia (aOR = 2.02; 95%CI, 1.76–2.33; I²=0%, p = 0.958; Figure 2(B)).

Anticholinergic potency and the risk of pneumonia

The total anticholinergic load of anticholinergic drugs is important in clinical medication and is scored by ADS[13] or ARS [14, 15]. High-potency anticholinergics are defined as an anticholinergic drug with an ADS/ARS score of 3 or more. Relatively, low-potency anticholinergics are defined with ADS/ARS scores less than 3. Therefore, we further found out three studies that systematically evaluated the association between anticholinergic potency level and pneumonia [5, 7, 8]. Our meta-analysis declared that both the high- and low-potency anticholinergics were related with an elevated risk of pneumonia. Moreover, risk estimates for use of high-potency anticholinergics (aOR = 1.96; 95%CI, 1.22–3.14; I²=91.8%, p = 0.000; Figure 3) were higher than those for low-potency anticholinergics (aOR = 1.58; 95%CI, 1.27–1.97; I²=85%, p = 0.000; Figure 3).

Figure 3. Forest plots of the association between the anticholinergic potency and the risk of pneumonia in elderly adults.

Exposure periods of anticholinergic medications and the risk of pneumonia

In order to investigate whether the risk was related to the exposure periods, two studies further evaluated the risk estimates for pneumonia when anticholinergic use 30 and 90 days before the event date or chronic use [5–7]. Our meta-analysis results showed the increased risk of pneumonia was still associated with the anticholinergic medication uses within 30 days (aOR = 2.13; 95%CI, 1.33–3.43; I²=83.2%, p = 0.015; Figure 4), within 90 days (aOR = 2.03; 95%CI, 1.26–3.26; I²=84.7%, p = 0.011; Figure 4) and chronic use (aOR = 1.65; 95%CI, 1.09–2.51; I²=93.5%, p = 0.000; Figure 4). The risk estimates for pneumonia increased slightly with the shortening of anticholinergic medication exposure periods.

Figure 4. Forest plots of the association between the exposure periods of anticholinergic medications and the risk of pneumonia in elderly adults.

Sensitivity analysis

Although some subgroups exhibited a high heterogeneity, the sensitivity analyses failed to find out the source of heterogeneity. Therefore, meta-analysis with random-effect model was used to analyze those subgroups with I² high than 50%.

Publication bias

Begg’s test and Egger’s test were used to assess the publication bias when the included studies in each subgroup meta-analysis were more than three. The P value of publication bias in each meta-analysis subgroup were showed in Table S4. The figures of all the Begg’s test and Egger’s test were exhibited in Figure S1. The above results convinced us that there was no publication bias in our meta-analysis (p > 0.05).

Discussion

The main finding of our meta-analysis was that the anticholinergic medication uses were associated with increased risk of pneumonia in elderly adults, especially among drugs with high anticholinergic potency and in the early phases of anticholinergic drugs treatment. In addition, we reported that the risk of stroke-associated pneumonia was higher than that of all pneumonia. To our knowledge, this is the first meta-analysis and systematic review conducted on the relationship between anticholinergic medication uses and pneumonia for only elderly patients.

Pneumonia is a pulmonary infection caused by a group of specific pathogens. Aging is related with a progressive decline in each systemic function especially respiratory system. Moreover, the comorbidities also make elderly patients more predispose to a related risk of pneumonia [16]. Previous literatures have documented that stroke in elderly patients is associated with increased risk of pneumonia [17]. Our meta-analysis result showed that the elderly exposed with anticholinergic medication who suffered from stroke have a higher risk of pneumonia than that of the whole pneumonia. It might be attributable to difficulty swallowing, dysphagia and impaired cough reflexes caused by stroke.

Anticholinergic medication includes drugs from a wide range of therapeutic categories [9], such as antipsychotics (e.g. hlorpromazine, olanzapine, haloperidol, paroxetine, quetiapine), antihistamines (e.g. cimetidine, ranitidine, cetirizine, loratadine, pseudoephedrine), antispasmodics (e.g. metoclopramide, anisodamine, atropine) and neurological disorder medications (e.g. larbidopa–levodopa, diphenhydramine, pramipexole, memantine). Previous study reported that more than 90% of patients who aged ≥75 years are taking anticholinergic drugs [18]. However, anticholinergic drugs should be balanced to prevent potential harm in elderly patients, although they have many beneficial effects. Thus, the total anticholinergic load of anticholinergic drugs is important in clinical medication and is scored by ADS or ARS scores [14,15,19]. Several studies had investigated the association between anticholinergic potency of the medication based on ADS or ARS scores and pneumonia in older individuals but the results were inconsistent [5–8]. The present meta-analysis showed that higher ADS/ARS scores was correlate with higher pneumonia incidence in elderly. That meant high-potency anticholinergic medication had higher association with pneumonia risk than low-potency anticholinergic medication. The high heterogeneity might attribute to different anticholinergic drug categories and use pattern. What’s more, some studies also assessed the relationship between time frame for exposure to anticholinergic drugs and risk of pneumonia [5–7]. Our meta-analysis found that anticholinergic medication user within 30 days were at greatest risk of pneumonia is than user within 90 days or chronic period. Our results may support the hypothesis that pneumonia is more pronounced in the initiation phase of anticholinergic drugs. Hence, we concluded that the anticholinergic load needs to be monitored and pneumonia-related signs and symptoms should be paid attention to after initiation of anticholinergic medication in elderly patients.

Biological plausibility

Anticholinergic drugs have some adverse effects including central (e.g. cognitive, drowsiness, sedation, delirium, confusion) and peripheral (e.g. dry mouth, constipation) adverse effects. Several potential hypotheses were made to explain the increased risk of pneumonia with anticholinergic medication uses based on adverse effects [20]. Firstly, dryness of mouth could impair oropharyngeal and oesophageal bolus transport which may result in aspiration pneumonia [6,9,21,22]. Second, low levels of thick mucosal secretion could increase the bacterial growth. Depression of mucociliary transport could prolong bacterial stay in the lungs. Both of them may finally lead to respiratory infection [23]. Third, low oesophageal sphincter pressure might lead to acid reflux which could cause aspiration [24]. At last, sedation and altered mental status, one of the anticholinergic drugs central adverse effects, were related with poor pulmonary hygiene, atelectasis and aspiration which might contribute to pneumonia [25]. However, the possible mechanisms still remain speculate with regard to molecular mechanisms.

Implications for practice

Clinicians should consider our current evidence show that an increased risk of pneumonia in elderly population with exposure to anticholinergic medication. When the pneumonia-related signs and symptoms are identified, older patients should monitor carefully in treatment with anticholinergic drugs because they are much more susceptible to pneumonia than younger individuals.

Strengths and limitations

Our research has some strengths. To our knowledge, this is the first meta-analysis and systematic review conducted on the relationship between anticholinergic medication uses and pneumonia for elderly patients only. Moreover, our meta-analysis allows us to combine the results of included studies to improve the statistical power in this investigation, provide more reliable risk estimates of this association.

Our meta-analysis and systematic review also have several limitations. Firstly, there were high heterogeneities across some subgroup meta-analysis, although we used a random-effect model to address the heterogeneity between individual studies. The heterogeneity could be partly explained by different study design, country, methodologies and different anticholinergic drugs from a wide range of therapeutic categories included in each eligible study. The second obvious limitation is the small enrolled sample of study. Despite the whole enrolled studies was 6 and subjects was 106,012, not all six of included studies was suited to be used for subgroup meta-analysis we were interested in. For the comparisons between anticholinergic medication uses and stroke-associated pneumonia, only two studies were included. And with regard to the ADS/ARS scores and exposure periods of anticholinergic drugs, there were only two or three eligible studies could be utilized for these two comparisons. And therefore, future studies with a larger sample size are warranted to confirm or refute these results.

Conclusion

In conclusion, the risk of pneumonia is increased in elderly adults with anticholinergic medication, especially with higher-potency anticholinergic drugs and in the initiation phase of anticholinergic medication. Clinicians should monitor their use in older patients carefully. Further studies with larger samples or randomized clinical trials are warranted to confirm or refute these results.

Author contributions

(I) Conception and design: MD Wu, XY Zhong, QC Zhang; (II) Administrative support: QC Zhang, XY Zhong; (III) Provision of study materials: MD Wu, ZX Li, WC Zheng; (IV) Collection and assembly of data: All authors; (V) Data analysis and interpretation: MD Wu, XY Zhong, ZX Li; (VI) Manuscript writing: MD Wu, ZX Li, XY Zhong; (VII) Final approval of the manuscript: All authors; (VIII) Revision: MD Wu.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data sharing is not applicable for no new data was generated.

Additional information

Funding

None.