The Cumulative Use of Muscle Relaxants and the Risk of Alzheimer’s Disease: A Nationwide Case-Control Study

Abstract

Background:

Use of pharmacological treatments is one possible modifiable risk factor for cognitive disorders.

Objective:

To investigate if the use of muscle relaxants is associated with the risk of Alzheimer’s disease (AD).

Methods:

The study was performed in a nested case-control design. Altogether 70,718 community-dwelling residents of Finland who received AD diagnosis in 2005–2011 were included as cases (the MEDALZ study). Each case was matched with four controls without AD by age, sex, and region of residence (N = 282,858). Data was extracted from Prescription register (1995–2012), Special Reimbursement register (1972–2012), and Hospital Discharge register (1972–2012). Drug use periods were modeled with PRE2DUP-method. Defined daily dose (DDD) was used to quantify the use. Analyses were conducted for any muscle relaxant use, and drug specific analyses were done for orphenadrine and tizanidine. A five-year lag window prior to the diagnosis was used, and results analyzed with conditional logistic regression.

Results:

The use of any muscle relaxant was associated with the risk of AD, aOR (95% CI) 1.04 (1.02–1.07). Stronger associations were observed with longer use (>366 days, aOR 1.12 (1.03–1.21)) than shorter use (1–365 days aOR, 1.04 (1.02–1.06)) compared to non-users. Dose-response was not observed. Tizanidine was not associated with AD, whereas cumulative exposure of orphenadrine (≥101 DDDs) was associated with the risk of AD, aOR 1.19 (1.07–1.32).

Conclusion:

Muscle relaxant use was associated with the risk of AD and higher exposure to orphenadrine showed increased risk. Further studies on higher doses and longer durations of use are warranted.

Keywords

Alzheimer’s disease dementia muscle relaxant orphenadrine tizanidine

INTRODUCTION

The prevalence of Alzheimer’s disease (AD) is presumed to rise to 75 million by year 2030 [1]. Investigation of potential modifiable risk factors to decrease the incidence of cognitive disorders is thus essential. Use of pharmacological treatments is one possible modifiable risk factor for cognitive disorders. Muscle relaxants are relatively commonly utilized group of drugs used for muscle spasms [2]. Approximately 400,000 muscle relaxant prescriptions are dispensed in Finland each year, with tizanidine and orphenadrine being the most commonly purchased muscle relaxants [3].

Both tizanidine and orphenadrine are centrally acting. Orphenadrine is anticholinergic and tizanidine has at least mild antagonistic action towards muscarinic acetylcholine receptors [4]. Orphenadrine is considered as a high potency anticholinergic according to anticholinergic risk scales [4], and it is included in the list of Beers Criteria for Potentially Inappropriate Medication Use in Older Adults and recommended to avoid because of possible anticholinergic effects, sedation, and increased risk of fractures [5]. Based on the different anticholinergic scales, tizanidine is considered high potency anticholinergic in one of the several anticholinergic burden scales [6]. To the authors’ knowledge, there is only one, small previous study on anticholinergic use and AD analyzing muscle relaxants separately as a group [7]. The objective of this study was to investigate if the use of muscle relaxants is associated with the risk of clinically verified AD among community-dwelling persons.

MATERIALS AND METHODS

Study population

The Finnish nationwide “Medication use and Alzheimer’s disease”, MEDALZ, is a nested case-control study with community-dwelling population of Finland as the source population [8]. The study includes 70,718 community-dwelling residents who received diagnosis for AD during years 2005–2011 in Finland (AD cases). All diagnoses were clinically verified by a geriatrician or neurologist. Each case was matched with up to four controls without AD by age, sex, and region of residence (N = 282,858). Controls were assigned from a register including all residents of Finland, maintained by the Social Insurance Institution of Finland (SII). The matching was done at the date of AD diagnosis which was assigned as index date for AD case and the controls. Incidence density sampling was used and thus the odds ratios can be interpreted as incidence rate ratios from a cohort study.

Case identification

Granted special reimbursement for anti-dementia drugs (acetylcholinesterase inhibitors and memantine) was used to identify cases with verified AD diagnosis. To be entitled to special reimbursement, the diagnosis has to be done by geriatrician or neurologist. The diagnosis is according to NINCDS-ADRDA and DSM-IV criteria [9–10]. The Finnish Current Care guidelines recommend that treatment with anti-dementia drugs is initiated for everyone with mild or moderate AD [11]. The treating clinician sends the petition of the special reimbursement to the SII. The SII then assesses if the evidence is sufficient for, and if so, grants eligibility for the reimbursement.

Data sources

Data were extracted from various Finnish nationwide registers. The used registers have been described in detail previously [12]. Briefly, the Special Reimbursement Register (since 1972) was used to identify the AD cases and comorbidities that were used to as confounders, such as cardiovascular diseases and diabetes [8]. Individual with certain clinically verified by physician diagnosis are granted higher reimbursement for drug prescribed for treatment of chronic conditions. The Prescription Register (since 1995) was used to gain information of reimbursed prescription drugs (which drug, strength, amount, recurring acquisition, time of acquisition). The Hospital Discharge Register (since 1972) was used to gain information on the hospital care periods and diagnoses.

Extracted data was linked with personal identity number, which is unique for every resident of Finland. The register maintainers have approved the study protocol. Data linkage was done by the SII. All the data was pseudonymized before it was submitted to the research team. According to Finnish legislation, ethical committee approval or informed consent is not required because data is collected routinely, and treatment of the participants has not been affected

Exposure

Reimbursed dispensings of any prescribed muscle relaxants were obtained from the Prescription Register from the year 1995 until index date. The exposure observation period covered 10 to 16 years before the index date, depending on the calendar year of the index date (2005–2011). All muscle relaxants in Finland require prescription to acquire. The most commonly used products, orphenadrine and tizanidine, have been reimbursed during the whole observation period. The drug use periods, i.e., when continuous medication use started and ended, were modeled with PRE2DUP-method, which has been described previously [13]. Briefly, PRE2DUP-method uses sliding averages of daily dose in defined daily doses (DDD, the assumed average maintenance dose per day for a drug used for its main indication in adults) and modelling is conducted separately for each person and for each Anatomical Therapeutic Chemical classification code (ATC) [14–15]. Drug purchase history and variation of purchase events of an individual are taken into account and modeled mathematically to form drug use periods. The method also considers drug stockpiling and hospital care periods as during hospital care, drugs are provided by the caring unit and not recorded in the Prescription Register.

The data extracted contained all muscle relaxant products that have had marketing authorization at some point of the observation period (ATC code M03). Products containing orphenadrine or tizanidine consisted of 79% of the total muscle relaxant use (Supplementary Table 1). The orphenadrine use was retrieved from orphenadrine citrate and orphenadrine citrate/paracetamol combination products.

Overlapping drug use periods of each substance were combined to study effect of “any muscle relaxant use” during the observation period. Cumulative duration of any muscle relaxant use was calculated by summing up all combined drug use periods (e.g., sum of time periods when a person has used a drug), and classified also to three categories: non-use, 1–365 days and ≥366 days of use. Additionally, the sum of the DDDs purchased were summed together and classified to following categories: ≤15 DDDs, 16–100 DDDs and ≥101 DDDs. The dose of the muscle relaxant use was calculated to see if there is dose-response effect. This dose level calculation was done by dividing cumulative amount of DDDs by cumulative duration of use. The dose levels were classified: low dose (0.0001–0.4999 DDDs/day), medium dose (0.5000–0.9999 DDDs/day), and high dose (≥1.0000 DDDs/day). Drug specific analyses were done for orphenadrine and tizanidine. For both substances, any use during observation period and cumulative DDDs were extracted from the data, and similar duration and DDD categories were used for any muscle relaxants use.

Covariates

Covariates used in the study were extracted from several registers. Data on cardiovascular diseases (chronic hypertension, coronary heart disease, chronic heart failure, and arrhythmia) and diabetes (diabetes mellitus 1 and 2) was extracted from the Special Reimbursement register (since 1972). History of stroke, bipolar disorder/depression, and schizophrenia were extracted from Hospital Discharge Register (since 1972). Previous exposure to benzodiazepines (N05BA, N05CD, N03AE01), antipsychotics (N05A, excluding lithium), and the total number of concomitant drugs in use were extracted from the Prescription Register (since 1995) data modelled with PRE2DUP. Benzodiazepine and antipsychotic use was classified: “never use” and “any use ever” before index date. Comorbidities, benzodiazepines and antipsychotic use were extracted since start of each register and until the beginning of lag window, depending on the analysis performed (primary 5-year lag before AD diagnosis and sensitivity analyses with 3-year lag and without a lag window).

Analysis of the data

The study was performed as a nested case-control study and analyzed with conditional logistic regression due to matched design. Multivariable models were adjusted for chronic hypertension, arrhythmias, coronary heart disease, heart failure, stroke, diabetes, schizophrenia, history of bipolar disorder/depression, benzodiazepine use, antipsychotic drug use, and total number of concomitant drugs in use. A p-value of≤0.05 was statistically significant. The drug specific analyses (orphenadrine and tizanidine) were additionally adjusted for other muscle relaxant use.

The primary analysis for any muscle relaxant use was done with 5-year lag, meaning that drug use 5 years before the index date was taken into account and drug use during lag window was not considered in the analyses. This was done to avoid reverse causality (increased contact with healthcare due to prodromal symptoms of AD can affect drug exposure). Similar lag was used for covariates. Odds ratios (OR) are presented with 95% confidence intervals (CI). Sensitivity analyses were performed with 3-year lag and without a lag window (Supplementary Tables 2–5). The statistical analyses were performed with IBM SPSS Statistics 25.

RESULTS

The median age at the index date was 80.8 years (range 34–105 years). Almost 80% of the cases were over 75 years old and 65.2% were women among both cases and controls (due to matching) (Table 1). Persons with AD had higher prevalence of cardiovascular diseases, previous stroke, diabetes, schizophrenia, and history of bipolar disorder/depression (p < 0.0001). Persons with AD used more concomitant medications and had used significantly more benzodiazepines and antipsychotics compared to those without AD (p < 0.0001).

Table 1

The characteristics of the study population at the index date

	Individuals with AD % (N), N = 70,718	Individuals without AD % (N), N = 282,858	p ^a
Female gender	65.2 (46,116)	65.2 (184,459)	matched
Age, y			matched
<65	3.2 (2,278)	3.2 (9,139)
65–75	17.6 (12,414)	17.6 (49,870)
76–85	55.1 (38,994)	55.1 (155,978)
>85	24.1 (17,032)	24.0 (67,871)
Any cardiovascular disease	51.2 (36,214)	50.1 (141,678)	<0.0001
Stroke	9.7 (6,828)	8.1 (22,985)	<0.0001
Diabetes	13.4 (9,460)	11.4 (32,201)	<0.0001
Schizophrenia	2.7 (1,882)	1.7 (4,708)	<0.0001
History of bipolar disorder/depression	5.6 (3,937)	3.7 (10,588)	<0.0001
Benzodiazepine use	32.6 (23,039)	28.4 (80,382)	<0.0001
Antipsychotic use	17.3 (12,257)	8.8 (24,998)	<0.0001

^achi-square test.

Use of any muscle relaxants at least five years before the index date was slightly more common in persons with AD than in controls (17.6% versus 16.5%) (Table 2). Women were more likely to use muscle relaxant than men in both groups. The muscle relaxant users had more frequently cardiovascular disease and stroke and history of bipolar disorder/depression compared to nonusers in both groups with and without AD. Users of muscle relaxants used also more frequently benzodiazepines (38.7% versus 23.6% in AD, 36.1% versus 21.5% in non-AD group) and antipsychotics (10.9% versus 7.1% in AD, 9.4% versus 5.8% in non-AD group).

Table 2

The characteristics of individuals who had used any muscle relaxant at least 5 years before index date compared to non-users among those with or without Alzheimer’s disease (AD)

	Individuals with AD % (N), N = 70,718			Individuals without AD % (N), N = 282,858
	Muscle relaxant non-users N = (58,272)	Muscle relaxant users N = (12,446)	p ^a	Muscle relaxant non-users N = (236,231)	Muscle relaxant users N = (46,627)	p ^a
Female gender	63.8 (37,206)	71.6 (8,910)	<0.0001	64.1 (151,452)	70.8 (33,007)	<0.0001
Age, y			<0.0001			<0.0001
<65	3.0 (1,726)	4.4 (552)	2.8 (6,719)	5.2 (2,420)
65–75	17.0 (9,901)	20.2 (2,513)	17.0 (40,044)	21.1 (9,826)
76–85	55.4 (32,260)	54.1 (6,734)	55.5 (131,090)	53.4 (24,888)
>85	24.7 (14,385)	21.3 (2,647)	24.7 (28,378)	20.4 (9,493)
Any cardiovascular disease	44.8 (26,124)	47.8 (5,950)	<0.0001	42.9 (101,401)	47.4 (22,103)	<0.0001
Stroke	5.3 (3,065)	6.0 (743)	0.001	4.8 (11,300)	5.6 (2,588)	<0.0001
Diabetes	9.8 (5,689)	10.0 (1,239)	0.513	7.5 (17,636)	8.4 (3,908)	<0.0001
Schizophrenia	1.6 (909)	1.4 (170)	0.109	1.4 (3,418)	1.0 (447)	<0.0001
History of bipolar disorder/depression	3.3 (1,904)	4.7 (591)	<0.0001	2.7 (6,395)	4.3 (2,024)	<0.0001
Benzodiazepine use	23.6 (13,779)	38.7 (4,812)	<0.0001	21.5 (50,777)	36.1 (16,823)	<0.0001
Antipsychotic use	7.1 (4,154)	10.9 (1,351)	<0.0001	5.8 (13,806)	9.4 (4,379)	<0.0001

^achi-square test.

Any use of muscle relaxants at least five years before the index date was associated with risk of AD, aOR 1.04 (95% Cl 1.02–1.07) compared with non-use of muscle relaxants (Table 3). The adjusted OR for longer-term use (≥366 days) was 1.12 (95% CI 1.03–1.21) and for use of 1–365 days 1.04 (1.02–1.06) compared to non-users. The association between higher cumulative sum of DDDs (≥101, aOR 1.12, 95% CI 1.05–1.20) was of similar magnitude with long-term use, whereas cumulative sum of ≤15 DDDs yielded aOR 1.03 (95% Cl 1.00–1.06) and cumulative sum of 16–100 DDDs aOR 1.05 (95% Cl 1.01–1.09). Consistent dose-response was not observed. Low dose was associated with risk of AD, aOR 1.06 (95% Cl 1.03–1.09), whereas medium dose, aOR 1.02 (95% Cl 0.99–1.06), and high dose, aOR 1.05 (95% Cl 0.87–1.27), had non-significant association with the risk. In sensitivity analyses, 3-year lag yielded similar results as 5-year lag, except high dose use of muscle relaxants, aOR 1.21 (95% Cl 1.02–1.45) (Supplementary Table 2). Without lag window all results were similar as results in 5-year lag except high dose use of muscle relaxants, aOR 1.18 (95% Cl 1.01–1.39) (Supplementary Table 3).

Drug-specific analyses (tizanidine or orphenadrine) were in line with the main analyses with any use of muscle relaxants (Table 4, Supplementary Tables 4 and 5). Any use of orphenadrine at least five years before the index date was associated with the risk of AD, aOR 1.04 (95% Cl 1.01–1.08) as well cumulative sum of ≥101 DDDs, aOR 1.19 (95% Cl 1.07–1.32) (Table 4). The association between any use of tizanidine and AD was similar 1.03 (1.00–1.05), but no associations were observed in the DDD analyses. In sensitivity analyses orphenadrine use was associated with AD in 3-year lag time analyses but not without lag time. Use of higher (≥101) cumulative sum of DDDs was persistently associated with the risk of AD in both sensitivity analyses (3-year lag aOR 1.18 (95% Cl 1.07–1.30), without lag aOR 1.12 (95% CI 1.02–1.22)) (Supplementary Tables 4 and 5).

Table 3

Risk of Alzheimer’s disease (AD) associated with muscle relaxant use with lag window of 5 years between exposure and index date

	Individuals with AD % (N), N = 70,718	Individuals without AD % (N), N = 282,858	Unadjusted OR (95% CI)	Adjusted OR (95% CI) *
Any muscle relaxant exposure at least 5 years before the index date
No use^a	82.4 (58,272)	83.5 (236,231)	reference	reference
Use	17.6 (12,446)	16.5 (46,627)	1.08 (1.06–1.11)	1.04 (1.02–1.07)
Cumulative duration of muscle relaxant use
1–365 days	16.4 (11,596)	15.5 (43,799)	1.07 (1.05–1.10)	1.04 (1.02–1.06)
≥366 days	1.2 (850)	1.0 (2,828)	1.22 (1.13–1.31)	1.12 (1.03–1.21)
Cumulative sum of DDDs of muscle relaxants used
≤15	10.4 (7,325)	9.9 (28,068)	1.06 (1.03–1.09)	1.03 (1.00–1.06)
16 –100	5.7 (4,021)	5.3 (14,890)	1.09 (1.06–1.13)	1.05 (1.01–1.09)
≥101	1.6 (1,100)	1.3 (3,673)	1.21 (1.13–1.30)	1.12 (1.05–1.20)
Cumulative dose of muscle relaxants (DDDs/day)
0.0001–0.4999	10.8 (7,619)	10.0 (28,217)	1.09 (1.06–1.12)	1.06 (1.03–1.09)
0.5000–0.9999	6.6 (4,695)	6.3 (17,934)	1.06 (1.03–1.10)	1.02 (0.99–1.06)
≥1.0000	0.2 (132)	0.2 (475)	1.13 (0.93–1.36)	1.05 (0.87–1.27)

*Adjusted for any cardiovascular disease, stroke, diabetes, schizophrenia, history of bipolar disorder/depression, benzodiazepine use, antipsychotic drug use and total number of concomitant drugs in use. ^aReference category in all analyses. Persons who used muscle relaxants only during the lag window are excluded from non-users in this analysis. DDD, defined daily dose; OR, odds ratio; CI, confidence interval.

Table 4

Risk of Alzheimer’s disease (AD) associated with tizanidine and orphenadrine* use with lag window of 5 years between exposure and index date

	Individuals with AD % (N), N = 70,718	Individuals without AD % (N), N = 282,858	Unadjusted OR (95% CI)	Adjusted OR (95% CI) *
Tizanidine use
Non-use^a	88.2 (62,342)	88.8 (251,116)	reference	reference
Use	11.8 (8,376)	11.2 (31,742)	1.06 (1.04–1.10)	1.03 (1.00–1.05)
Tizanidine cumulative sum of DDDs
≤15	9.1 (6,457)	8.7 (24,676)	1.06 (1.03–1.09)	1.03 (1.00–1.06)
16–100	2.3 (1,630)	2.1 (6,073)	1.08 (1.03–1.15)	1.02 (0.97–1.08)
≥101	0.4 (289)	0.4 (993)	1.17 (1.03–1.34)	1.06 (0.93–1.21)
Orphenadrine use
Non-use^a	94.0 (66,465)	94.5 (267,276)	reference	reference
Use	6.0 (4,253)	5.5 (15,582)	1.10 (1.06–1.14)	1.04 (1.01–1.08)
Orphenadrine cumulative sum of DDDs
≤15	2.6 (1,856)	2.5 (7,052)	1.06 (1.01–1.12)	1.01 (0.96–1.07)
16–100	2.7 (1,941)	2.5 (7,108)	1.10 (1.04–1.16)	1.04 (0.99–1.10)
≥101	0.6 (456)	0.5 (1,422)	1.29 (1.16–1.44)	1.19 (1.07–1.32)

*Orphenadrine alone and orphenadrine combination products. **Adjusted for other muscle relaxant use, any cardiovascular disease, stroke, diabetes, schizophrenia, history of bipolar disorder/depression, benzodiazepine use, antipsychotic drug use and total number of concomitant drugs in use. ^aReference category in all analyses. DDD, defined daily dose; OR, odds ratio; CI, confidence interval.

DISCUSSION

To the authors’ knowledge, this is the first study to examine association between muscle relaxants use and a risk of AD. A previous study reported the association between anticholinergic drug use (including muscle relaxants) and AD [7]. Although the study assessed separately the risk associated with the specific drug classes, comparison to the results of the current study is problematic, as the previous study classified orphenadrine as antiparkinsonian drug and the muscle relaxant group consisted only tizanidine and methocarbamol. The total number of the users in the previous study was rather small, for tizanidine (N = 64) and orphenadrine (N = 95) during from one to 11-year exposure period and no drug specific analyses were reported.

The use of muscle relaxants and cumulative use of over > 366 days was associated with the risk of AD. This might be explained partly by use of anticholinergic muscle relaxant orphenadrine. This is supported by our finding that higher cumulative use of orphenadrine was associated with risk of AD. Orphenadrine has high anticholinergic properties due to its affinity to muscarinic receptors. Anticholinergic drugs have been associated with an increased risk of cognitive disorders in previous observational studies [7,16, 7,16], but there is not yet proof of causality between anticholinergic use and AD. Anticholinergic drugs do have the potential for such however, as these drugs target cholinergic neurons in brains. However, in present study aORs were small. The cumulative duration of use and even the high dose category were rather uncommon in our study (for example 0.2 difference for cumulative duration of use of > 366 days). Thus, further research is needed in populations with more intensive and/or longer durations of muscle relaxant use.

In this study, four out of five users were aged 75 years or older and a quarter were aged 85 years or older. This is clinically concerning as older persons are more sensitive to adverse drug effects and events of medications in general, especially those involving CNS and anticholinergic activity [17–18]. Sedation and drowsiness are the main CNS adverse drug effects reported in tizanidine users [2]. Due to the adverse drug effects of tizanidine it is categorized in the Finnish Meds75 + database maintained by Finnish Medical Agency as a drug that should be avoided in older persons [19]. Orphenadrine has adverse effects like anxiety, blurred vision, xerostomia, and dizziness [2] and is similarly categorized in Meds75 + to be avoided in older people.

We observed that users of muscle relaxants used also more frequently benzodiazepines and antipsychotics than non-users of muscle relaxants. Benzodiazepines can also be used to treat muscle spasms, so it might be that both muscle relaxants and benzodiazepines may have been attempted [2]. Antipsychotics may induce extrapyramidal symptoms, such as rigidity and muscle spasms and thus, muscle relaxants may be used for treating these adverse effects [20]. However, the association between muscle relaxant use and risk of AD remained after adjusting for benzodiazepine and antipsychotics use. The association of concomitant use of all three drug groups muscle relaxants, benzodiazepines and antipsychotics would be an interesting study topic for further research.

Current study has a number of strengths. The study is based on nationwide registers, containing largely health related information [8]. The Finnish health-care system offers universal health care services to all residents despite their socioeconomic status. The AD diagnoses were clinically verified by geriatrician or neurologist. The positive predictive value of diagnoses is high (PPV 97.1%) [21]. The exposure assessment period was long varying from 10 to 16 years. Drug exposure during the 5-year lag window was not taken into consideration to avoid reverse causality, and sensitivity analyses were done with 3-year lag window and without lag window. Drug use periods were modelled with validated PRE2DUP-method [13].

The study also has limitations. The severity of comorbidities was not included in the register data. The data lacked information on mild cognitive impairment or subjective cognitive impairments, but to address this limitation we considered drug exposure that had occurred at least five years before the outcome in the main analysis. In addition, we had no information on confounders such as lifestyle behaviors like healthy diet and physical activity. However, the adjustment for comorbidities likely captured the health impacts of these factors to some extent. As the aORs in the analyses were close to one, the observed associations might be due to residual confounding. We did not analyze all other anticholinergics drugs. However, multivariate regression models were adjusted with antipsychotic use (high potency anticholinergics). Previous study on MEDALZ cohort reported that prevalence of concomitant use of acetylcholinesterase inhibitors and anticholinergic urinary antispasmodics was 7.3% among acetylcholinesterase inhibitor users with AD [22].

In conclusion, muscle relaxant use was associated with minor risk of AD and higher exposure to orphenadrine showed increased risk possibly due to its anticholinergic effects. Further studies on higher doses and longer durations of use are warranted. Our findings are concerning as muscle relaxants were most commonly used by persons aged 75 years and especially among oldest old aged 85 years or more as these are among most vulnerable older adults.

Footnotes

ACKNOWLEDGMENTS

The authors have no acknowledgments to report.

FUNDING

The authors have no funding to report.

CONFLICT OF INTEREST

HT and AT have participated in research projects funded by grants from Janssen-Cilag and Eli Lilly to their employing institution. HT reports personal fees from Janssen-Cilag. Other authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the findings of this study are available on reasonable request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

The supplementary material is available in the electronic version of this article: .

References

Prince

, Wimo

, Guerchet

, Ali

G-C

, Wu

Y-T

, Prina

(2015). World Alzheimer Report 2015. The Global Impact of Dementia: An Analysis of Prevalence, Incidence, Cost and trends. Alzheimer’s Disease International, London.

Chou

, Peterson

, Helfand

(2004) Comparative efficacy and safety of skeletal muscle relaxants for spasticity and musculoskeletal conditions: A systematic review. J Pain Symptom Manage 28,140–175.

Official Statistics of Finland: Reimbursements of medicine expenses: Number of recipients and prescription data. Kela, Helsinki.

Durán

, Azermai

, Vander Stichele

(2013) Systematic review of anticholinergic risk scales in older adults. Eur J Clin Pharmacol 69,1485–1496.

By the 2019 American Geriatrics Society Beers Criteria^® Update Expert Panel (2019) American Geriatrics Society 2019 Updated AGS Beers Criteria^® for potentially inappropriate medication use in older adults. J Am Geriatr Soc 67,674–694.

Lozano-Ortega

, Johnston

, Cheung

, Campbell

, Dmochowski

, Ng

(2020) A review of published anticholinergic scales and measures and their applicability in database analyses. . Arch Gerontol Geriatr 87,103885.

Coupland

CAC

, Hill

, Dening

, Morriss

, Moore

, Hippisley-Cox

(2019) Anticholinergic drug exposure and the risk of dementia: A nested case-control study. JAMA Intern Med 179,1084–1093.

Tolppanen

A-M

, Taipale

, Koponen

, Lavikainen

, Tanskanen

, Tiihonen

, Hartikainen

(2016) Cohort profile: The Finnish Medication and Alzheimer’s disease (MEDALZ) study. BMJ Open 6,e012100.

McKhann

, Drachman

, Folstein

, Price

, Stadlan

(1984) Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA work group under the auspices of department of health and human services task force on Alzheimer’s disease. Neurology 34,939–944.

10.

American Psychiatric Association (1994) Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. American Psychiatric Association, Washington, DC.

11.

Working group appointed by the Finnish Medical Society Duodecim, Societas Gerontologica Fennica, Finnish Geriatricians, the Finnish Neurological Society, Finnish Psychogeriatric Association and the Finnish Association for General Practice (2017) Memory Disorders: Current Care Guidelines. Helsinki: The Finnish Medical Society Duodecim, Helsinki. https://www.kaypahoito.fi/, Accessed on July 7, 2020

12.

Tolppanen

A-M

, Taipale

, Koponen

, Lavikainen

, Tanskanen

, Tiihonen

, Hartikainen

(2013) Use of existing data sources in clinical epidemiology: Finnish health care registers in Alzheimer’s disease research – the Medication use among persons with Alzheimer’s disease (MEDALZ-2005) study. Clin Epidemiol 5,277–285.

13.

Tanskanen

, Taipale

, Koponen

, Tolppanen

, Hartikainen

, Ahonen

, Tiihonen

(2015) From prescription drug purchases to drug use periods - a second generation method (PRE2DUP). BMC Med Inform Decis Mak 15,21.

14.

WHO collaborating centre for drugs statistics methodology (2018) DDD Definitions and general considerations. https://www.whocc.no/ddd/definition_and_general_considera/, Accessed on May 7, 2020.

15.

WHO collaborating centre for drugs statistics methodology (2018) ATC Structure and principles. https://www.whocc.no/atc/structure_and_principles/, Accessed on May 7, 2020.

16.

Gray

, Hanlon

(2016) Anticholinergic medication use and dementia: Latest evidence and clinical implications. Ther Adv Drug Saf 27,217–224.

17.

Cusack , BJ (2004) Pharmacokinetics in older persons. Am J Geriatr Pharmacother 4,274–302.

18.

Mangoni

, Jackson

(2004) Age-related changes in pharmacokinetics and pharmacodynamics: Basic principles and practical applications. Brit J Clin Pharmacol 1,6–14.

19.

Finnish Medicines Agency Fimea (2020) Meds75+. 2020 http://www.fimea.fi, Accessed on May 7, 2020.

20.

Muench

, Hamer

(2010) Adverse effects of antipsychotic medications. Am Fam Physician 81,617–622.

21.

Solomon

, Ngandu

, Soininen

, Hallikainen

, Kivipelto

, Laatikainen

(2014) Validity of dementia and Alzheimer’s disease diagnoses in Finnish national registers. Alzheimers Dement 10,303–309.

22.

Torvinen-Kiiskinen , S , Taipale

, Tanskanen

, Tiihonen

, Hartikainen , S (2014) Concomitant use of acetylcholine esterase inhibitors and urinary antispasmodics among Finnish community-dwelling persons with Alzheimer disease. J Clin Psychopharmacol 34,722–727.