Antipsychotic Treatment Associated With
Increased Mortality Risk in Patients With
Dementia. A Registry-Based Observational
Cohort Study
Emilia Schwertner, Juraj Secnik, Sara Garcia-Ptacek, Bjorn Johansson, Katarina
Nägga, Maria Eriksdotter, Bengt Winblad and Dorota Religa
The self-archived postprint version of this journal article is available at Linköping
University Institutional Repository (DiVA):
http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-155561
N.B.: When citing this work, cite the original publication.
Schwertner, E., Secnik, J., Garcia-Ptacek, S., Johansson, B., Nägga, K., Eriksdotter, M., Winblad, B.,
Religa, D., (2019), Antipsychotic Treatment Associated With Increased Mortality Risk in Patients With
Dementia. A Registry-Based Observational Cohort Study, Journal of the American Medical Directors
Association, 20(3), 323-329.e2. https://doi.org/10.1016/j.jamda.2018.12.019
Original publication available at:
https://doi.org/10.1016/j.jamda.2018.12.019
Copyright: Elsevier
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/331450092
Antipsychotic Treatment Associated With Increased Mortality Risk in Patients
With Dementia. A Registry-Based Observational Cohort Study
Article in Journal of the American Medical Directors Association · March 2019
DOI: 10.1016/j.jamda.2018.12.019 CITATIONS 0 READS 589 8 authors, including:
Some of the authors of this publication are also working on these related projects:
ADEX; the effect of physical exercise in Alzheimer's DiseaseView project
Behavioural and Psychiatric symptoms in dementia , neuroinflammation, dementia biomarkersView project Emilia Schwertner Karolinska Institutet 3PUBLICATIONS 0CITATIONS SEE PROFILE Juraj Secnik Karolinska Institutet 5PUBLICATIONS 16CITATIONS SEE PROFILE Sara García-Ptacek Karolinska Institutet 72PUBLICATIONS 546CITATIONS SEE PROFILE Maria Eriksdotter Karolinska Institutet 260PUBLICATIONS 6,407CITATIONS SEE PROFILE
All content following this page was uploaded by Juraj Secnik on 06 March 2019. The user has requested enhancement of the downloaded file.
Original Study
Antipsychotic Treatment Associated With Increased Mortality Risk
in Patients With Dementia. A Registry-Based Observational Cohort
Study
Emilia Schwertner MSc
a, Juraj Secnik MD
a, Sara Garcia-Ptacek PhD
a,b,
Björn Johansson PhD
c,d, Katarina Nagga PhD
e,f, Maria Eriksdotter PhD
a,d,
Bengt Winblad PhD
d,g, Dorota Religa PhD
a,d,*
aCenter for Alzheimer Research, Division of Clinical Geriatrics, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Huddinge,
Sweden
bDepartment of Internal Medicine, Section for Neurology, Södersjukhuset, Stockholm, Sweden cDepartment of Molecular Medicine and Surgery Stockholm, Karolinska Institutet, Sweden dTheme Aging, Karolinska University Hospital, Huddinge, Sweden
eClinical Memory Research Unit, Department of Clinical Sciences Malmö, Lund University, Malmö, Sweden fDepartment of Acute Internal Medicine and Geriatrics, Linköping University, Linköping, Sweden
gCenter for Alzheimer Research, Division of Neurogeriatrics, Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Huddinge,
Sweden Keywords: Antipsychotics dementia mortality Alzheimer vascular dementia mixed dementia
a b s t r a c t
Objective: To assess all-cause mortality patients with dementia treated with typical and atypical anti-psychotic drugs (APDs).
Design: Registry-based cohort study.
Setting and participants: A total of 58,412 patients diagnosed with dementia and registered in the Swedish Dementia Registry were included in the study. Of the study sample, 2526 of the patients were prescribed APDs. Of these, 602 patients were prescribed typical APDs and 1833 patients were prescribed atypical APDs. Ninety-one patients were prescribed both typical and atypical APDs.
Measurements: All-cause mortality based on Swedish Cause of Death Register. Adjusted hazard ratios of mortality were calculated according to class of APDs (typical or atypical) prescribed. Final models were adjusted for age at dementia diagnosis, sex, Charlson comorbidity index, living arrangement, and Mini-Mental State Examination.
Results: In the adjusted models, use of APDs at the time of dementia diagnosis was associated with
increased mortality risk in the total cohort (hazard ratio¼ 1.4; 95% confidence interval 1.3e1.5). After
stratifying for dementia types, increased mortality risks associated with APDs were found in patients with Alzheimer’s disease, mixed dementia, unspecified dementia, and vascular dementia. Higher risk for mortality was found with typical APDs in patients with mixed and vascular dementia and with atypical APDs in patients with Alzheimer’s disease, mixed, unspecified, and vascular dementia. Furthermore, in patients with Alzheimer’s disease who had typical APDs, use lower risk of death emerged in comparison with patients with atypical APDs.
Conclusions/Implications: Both the use of atypical and typical APDs increased the risk of death in patients with dementia even after adjusting for differences in basic characteristics between groups. Although we
SveDem is supportedfinancially by the Swedish Brain Power network (http://
swedishbrainpower.se), the Swedish Associations of Local Authorities and
Re-gions, Gun och Bertil Stohnes Stiftelse, CIMED grant, Alzheimerfonden, and Swedish Research Council (Drn 2012-2291 and Drn 2016-02317), and by grants provided by the Stockholm County Council (ALF project). None of the sponsors had any involvement in the design of the study, the data collection and analysis, the writing of the report, or the decision to submit the paper for publication.
All authors have completed the ICMJE uniform disclosure form atwww.icmje. org/coi_disclosure.pdfand declare: no support from any organization for the sub-mitted work, nofinancial relationships with any organizations that might have an
interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.
This study was approved by the regional ethical review board in Stockholm, Sweden. Ethical Number: 2015/2232-31/5.
The authors declare no conflicts of interest.
* Address correspondence to Dorota Religa, PhD, Department of Neurobiology, Care Sciences and Society (NVS), H1, Division of Clinical geriatrics, Sektionen För Klinisk Geriatrik, Plan 7 141 83 Huddinge, Sweden.
E-mail address:dorota.religa@ki.se(D. Religa).
https://doi.org/10.1016/j.jamda.2018.12.019
1525-8610/Ó 2018 Published by Elsevier Inc. on behalf of AMDA e The Society for Post-Acute and Long-Term Care Medicine.
JAMDA
j o u r n a l h o m e p a g e : w w w . j a m d a . c o m
cannot rule out the influence of residual confounding, these results would seem to add to studies suggesting caution in APD prescription for patients with dementia.
Ó 2018 Published by Elsevier Inc. on behalf of AMDA e The Society for Post-Acute and Long-Term Care Medicine.
Dementia is a general term for a clinical syndrome caused by
neurodegeneration.1Survival time after dementia diagnosis has been
reported to range between 1.1 and 8.5 years2and interactions of many
different factors influence survival of a patient. In a previous study, it
was shown that patients with other dementia diagnoses had a higher
risk of death compared with patients with Alzheimer’s disease (AD).3
Other risk factors are older age, functional impairment, male sex, low body mass index, severity of dementia, institutionalization, and higher
number of medications.3e5Previous studies have also raised concerns
about the use of antipsychotic drugs (APDs), which are mainly pre-scribed to treat behavioral and psychological symptoms of dementia
(BPSD).6,7Approximately 80% of patients with dementia will suffer
from at least 1 symptom during the course of disease,8 and this
significantly decreases the quality of life of both the patients and their
carers,9often underlying the decision for institutionalization10,11and
contributing to the progression of dementia.12
Both atypical and typical APDs are used in the management of BPSD. There are numerous studies showing that APD treatment in patients with dementia may lead to serious side effects including death. A meta-analysis of 15 randomized controlled trials (RCTs) showed that short-term treatment with atypical APDs increases the
risk of death 1.54 times compared with placebo treatment.13
Subse-quently, observational studies found that typical APDs may convey
similar14or even higher risk of death relative to atypical APDs.15On
the other hand, other studies suggest that this association does not
persist when adjusting for BPSD,16use of restraints,17or other clinical
factors like severity of dementia.18 Similarly, no difference in
long-term survival was found between patients with BPSD with and
without APD treatment.4 This lack of association was further
sup-ported by recent studies with a long follow-up period19as well as in
meta-analyses of RCTs.20
Furthemore, it is not clear whether APD treatment significantly
contributes to improvement of BPSD in patients with dementia. An RCT of patients with AD showed a clear positive effect of olanzapine and risperidone in treating patients with neuropsychiatric symptoms
compared with placebo and quetiapine.21However, the benefits were
observed only among those patients who tolerated the medications. Studies suggest that whether advantages of APDs can offset their
adverse effects may depend on the individual profile of a patient
including the specific BPSD manifestations and dementia severity.22
Therefore, it is important to identify factors associated with sur-vival and mortality of patients with APD treatment. Previous studies have either concentrated on patients with AD or made no distinction between different types of dementia. To our knowledge, there is only 1
study including exclusively patients with vascular dementia (VAD).23
If the effect of APDs varies in less common diagnoses, conclusions may be impeded by a lack of research on the subject. Moreover, to our knowledge, little research exists on the mixed-pathology dementia
and unspecified (UNS) dementia diagnoses with no studies
consid-ering antipsychotic treatment. To help address these questions, we used data on patients with dementia registered in Swedish Dementia
Registry (SveDem).24The large size of the database used in the current
study enables an extensive description of the effects of typical and atypical APD treatment on mortality in patients with different de-mentia diagnoses and, therefore, enables us to develop a more comprehensive picture of the risk of death associated with APDs.
In this study, we report the risk of long-term mortality of patients with different dementia diagnoses associated with being treated with
typical and atypical APDs. Adjustments were made for the most relevant clinical and sociodemographic factors that may have an impact on the survival of patients with dementia.
Methods
We studied patients diagnosed with dementia and registered at
the time of diagnosis in the SveDem.24SveDem was established in
May 2007 with the goal to improve the quality of care of patients with dementia disorders in Sweden and contains information about the diagnostic work-up, medical treatment, and community support. Dementia was categorized into 8 diagnoses: AD, VAD, mixed de-mentia, dementia with Lewy bodies (DLB), frontotemporal dementia
(FTD), Parkinson’s disease dementia, UNS dementia (used when
spe-cific dementia diagnosis was not recognized), and other dementia
types (including various dementia disorders eg, Corticobasal degen-eration and alcohol related dementias). Patients are diagnosed and followed-up yearly in specialist units, primary care centers, or in nursing homes. Clinicians were instructed to diagnose the type of dementia according to the 10th revision of the International Statistical
Classification of Diseases and Related Health Problems, 10th Revision.25In
addition, the McKeith criteria26 for DLB, the Lund-Manchester
criteria27 for FTD, and the Movement Disorder Society Task Force
criteria28for PDD are recommended for diagnosis purposes.
We used the following Anatomical Therapeutic Chemical Classi
fica-tion System codes (see: http://www.whocc.no/atc_ddd_index) for
antipsychotics classes: typical: N05AA phenothiazines with aliphatic side-chain, N05AB phenothiazines with piperazine structure, N05AD phenothiazines with piperidine structure, N05AF thioxanthene de-rivatives; atypical: N05AE indole derivatives, N05AH diazepines,
oxazepines, thiazepines, and oxepines, and N05AX other
antipsychotics.
The study population consisted of 58,412 patients registered at the time of dementia diagnosis in SveDem between 2007 and 2015. After removing duplicates (258) from SveDem and patients with missing data on dementia diagnosis (117), we obtained a study population of
58,037 patients. The file was merged with the Swedish Prescribed
Drug Register, Swedish Cause of Death Register, and patients
regis-tered in The National Board of Health and Welfare (http://www.
socialstyrelsen.se). To decrease the risk of survival bias, we
restricted the analysis of patients who had APD prescriptions after and
within thefirst year after the dementia diagnosis was established and
patients not taking APDs. Thefinal sample contained 56,048 patients.
Survival time was calculated individually for all patients from de-mentia diagnosis to date of death or end of follow-up as of August 28, 2016.
Statistical Analyses
Baseline characteristics were described with frequencies and per-centages for categorical variables and means and standard deviations (SDs) for continuous variables. To analyze differences between
pa-tients, t tests for independent samples or analysis of variance and
c
2tests were performed for continuous and categorical variables, respectively. We used Kaplan-Meier method to estimate median sur-vival time and reverse Kaplan-Meier method to calculate median time
of follow-up.29To test the proportional hazard assumption for main
predictors, we plotted the Schoenfeld residuals as a function of time
E. Schwertner et al. / JAMDA 20 (2019) 323e329 324
and tested the hypothesis of a zero slope. Next, we ran separate Cox multivariate regression models to estimate the risk of death associated with APD use compared with nonuse for the entire cohort and
strat-ified by dementia type. From these models, we derived crude and
adjusted estimates of hazard ratio (HR) and corresponding 95% con-fidence intervals (CIs). Final models were adjusted for age at dementia diagnosis, sex, and form of residency (living alone in ordinary housing, with a family member in ordinary housing, or in a nursing home), Mini-Mental State Examination (MMSE) used to assess severity of
dementia and Charlson comorbidity index (CCI) for comorbidity.30
Subsequently, we repeated analyses to compare users of typical and atypical APDs. Previous studies suggest that PDD and DLB may pertain
to the same disease spectrum31and consequently, we merged the 2
diagnoses into 1 group called LBD. The assumption for proportional hazard was violated in the VAD group, therefore, interactions with time were used to model the association in this diagnosis. Subse-quently, we inspected residual plots and divided data into 2 epochs of
thefirst 360 days and greater than that. Separate hazard functions
were fit for each time-band. Data were analyzed with the Survival
Package for R statistical software (the R Foundation for Statistical
Computing, Vienna, Austria).32,33
Results
Description of the Study Population
The results are presented as means (standard deviation) for
continuous variables (age, MMSE, CCI, number of drugs) or number of patients, and percentage for categorical variables (dementia type, sex, form of residency). The study sample contained 59% women and 41% men. The mean age was 79.7 (7.9) years and mean MMSE score was 20.9 (5.0) points. AD (31%) was the most common dementia diagnosis
followed by UNS dementia (25%), VAD (19%), and mixed dementia (18%). LBD (4%), other types of dementia (3%), and FTD (2%) together
accounted for less than 8% of all dementia diagnoses (Table 1). In the
sample, 4% (n¼ 2526) of the patients had an APD prescription: 602
had only typical APDs and 1833 had only atypical APDs. Ninety-one patients had both typical and atypical APDs prescribed. There were differences in APD use between the different dementia diagnoses. Among patients receiving APDs, typical APD use was lowest in pa-tients with LBD (13%), followed by other (23%), VAD (24%), FTD (24%), UNS dementia (25%), AD (26%), and mixed dementia (32%). Atypical use was lowest in mixed dementia (68%), followed by AD (74%), UNS dementia (75%), VAD (76%), FTD (76%), other (77%), and LBD (87%). A description of patients with and without APD treatment is provided in
Table 2.
Mortality in Patients Receiving APDs
Kaplan-Meier survival estimates for the median (95% CI) survival time after diagnosis was 1793 (range 1773e1812) days. Median (95% CI) time of follow-up was 1469 (range 1458e1495) days. Of patients treated with APDs, 61% died at the end of the study, and of patients not receiving APDs, 41% died. Among deceased patients with APD use, 63% had received typical and 58% atypical APDs. Use of typical and atypical APDs was associated with increased mortality risk compared with no APD use in the entire cohort [HR (95% CI): typical APDs: 1.4 (1.2e1.5), atypical APDs: 1.4 (1.3e1.5) (adjusted for sex, age, MMSE, CCI, and form of residency)].
After stratifying for dementia diagnoses, statistically significant
as-sociations between APD use and increased risk of mortality compared with non-APD use were found in AD [atypical: 1.5 (1.3e1.8), mixed dementia (typical: 1.3 (1.1e1.6); atypical: 1.3 (1.1e1.6))], and UNS de-mentia [atypical: 1.3 (1.2e1.5)]. In VAD after including interactions with
Table 1
Description of Patients Included in the Study
N¼ 56,048 Total Missing AD (17 315) Mixed Dementia (10,336)
VAD (10,512) LBD (1127) FTD (816) UNS Dementia (13,813) Other (1 398) P Age mean, y (SD) 79.72 (7.87) 0 (0) 77.93 (8.19) 81.26 (6.59) 80.67 (7.25) 76.38 (7.20) 70.34 (9.28) 81.44 (7.43) 76.14 (9.84) <.001 Female sex n (%) 32996 (59) 0 (0) 11 133 (64) 6 096 (59) 5 481 (52) 705 (37) 431 (53) 8 436 (61) 714 (51) <.001 MMSE mean (SD) 20.88 (5.03) 3 975 (7) 21.28 (4.98) 20.71 (4.97) 20.87 (4.98) 21.25 (4.81) 23.14 (5.37) 20.26 (5.08) 21.28 (5.18) <.001 CCI mean (SD) 2.36 (1.77) 0 (0) 1.94 (1.49) 2.52 (1.82) 3.03 (2.05) 2.10 (1.62) 1.88 (1.48) 2.31 (1.71) 2.29 (1.65) <.001 TND mean (SD) 5.54 (3.34) 6 520 (12) 4.85 (3.07) 5.78 (3.31) 6.3 (3.51) 5.87 (3.35) 4.66 (3.28) 5.59 (3.34) 5.34 (3.50) <.001 Living arrangement: <.001 Alone, ordinary housing n (%) 23 063 (41) 469 (1) 6 928 (40) 4 697 (46) 4 294 (41) 451 (24) 255 (31) 5 894 (43) 544 (40) Ordinary housing n (%) 26 606 (48) 9 288 (54) 4 747 (46) 4 703 (45) 1 150 (62) 505 (62) 5 619 (41) 594 (43) Special residency n (%) 5 910 (11) 998 (6) 864 (8) 1 436 (14) 255 (14) 56 (7) 2 063 (15) 238 (17)
TND, total number of drugs.
Table 2
Differences Between Patients Without APDs, With APDs, and With Typical and Atypical APDs
No APDs (n¼ 53, 522) APDs (n¼ 2526) P* Typical (n¼ 601) Atypical (n¼ 1833) Py Age, mean (SD), y 79.7 (7.9) 80.5 (8.07) <.001 81.16 (7.8) 80.32 (8.2) .03 Female sex, n (%) 31 495 (59) 1501 (59) .58 370 (62) 1075 (59) .22 MMSE, mean (SD) 20.96 (5.0) 19.08 (5.5) <.001 18.74 (5.7) 19.27 (5.4) .06 CCI, mean (SD) 2.35 (1.8) 2.48 (1.8) <.001 2.64 (2.0) 2.42 (1.8) .52
Living arrangement: <.001 <.001
Ordinary housing, living alone, n (%) 22,175 (42) 888 (35) 200 (34) 661 (36) Ordinary housing, n (%) 25,752 (49) 854 (34) 177 (30) 652 (36) Nursing home, n (%) 5144 (10) 766 (31) 200 (37) 506 (28)
Deaths, n (%) 21,765 (41) 1510 (60) <.001 379 (63) 1072 (58) .05 *Comparison between APDs and no APDs.
yComparison between typical and atypical APDs.
time, adjusted HR in the first epoch for typical APDs comparing non-APD use was 1.8 (1.1e2.8), and in the second epoch 1.6 (1.2e2.1); for atypical APDs 1.5 (1.2e2.0) in the first epoch and 1.2 (1.0e1.4), in the
second epoch (Supplemental Table 1). Assuming proportional hazards:
typical: 1.6 (1.3e2.1); atypical: 1.3 (1.1e1.5), P < .001.
Additionally, in patients with AD, typical APD use was associated with lower risk of death compared with atypical APDs [0.7 (0.5e0.9)]. In patients with LBD, there was a nonsignificant trend of
increased risk of death with typical APD use compared with non-APD use: 1.4 (0.9e2.1), P ¼ .13; and atypical non-APD use 1.4 (0.9e2.2),
P¼ .19. In patients with VAD, typical APD use showed a
nonsignif-icant trend of increasing death risk compared with atypical APD use in the second epoch (1.3 (0.9e1.8), P ¼ .54) but not in the first one [1.2 (0.7e2.0), P ¼ .11]. A nonsignificant trend toward higher risk of death with typical APD use compared with atypical APD use was also
found assuming proportional hazards 1.3 (1.0-1.7), P¼ .08 (Figure 1,
Fig. 1. Cox multivariate analyses assessing all-cause mortality risk associated with APD for different dementia diagnoses. VAD: Assuming proportional hazard; Typical and atypical compared with no APDs; Typical1compared with atypical APDs; Total (typicalþ atypical) compared with no APDs.
E. Schwertner et al. / JAMDA 20 (2019) 323e329 326
crude model:Supplemental Table 2; survival curvesSupplemental
Figure 1).
Discussion
In our study, 4% of patients with newly established dementia diagnosis used APDs. Use of APDs was associated with risk for death, and these results are in line with previous studies showing higher
death risk with APD use in patients with dementia.13,14 However,
when each specific dementia diagnosis was considered, we found
differences in survival for patients with different diagnoses. More
specifically, in patients with AD, a reduced death risk for typical APDs
treatment compared with atypical APD treatment was found. More-over, typical APDs were associated with similar mortality risk as lack of APD treatment and only atypical APDs were increasing risk of death. To our knowledge, this has not been previously described. Similar to patients with AD, the increased mortality in patients with UNS de-mentia was found only with atypical APD use. It is, therefore, possible
that this group includes a number of patients with AD,24which may
contribute to this effect. Both atypical and typical APDs were associ-ated with the increased risk of death in patients with VAD and mixed
dementia. We also identified nonsignificant trends indicating
increased risk of death for typical APD use compared to atypical APD use in patients with VAD and for typical APD use compared with atypical use and no APDs in patients with LBD.
Although part of the effect of APDs could be explained by the condition of patients with APDs, the association persists even after adjusting for differences between patients with and without APD treatment. Our data show that patients with APD treatment were slightly older, had a poorer cognition, more often lived in nursing home, and had higher total amount of drugs and comorbidity index at the time of dementia diagnosis.
It has been known that patients with LBD are particularly vulner-able for severe neuroleptic sensitivity reaction (NSR), even with low
doses or atypical APDs.34 NSR usually includes increase in
Parkin-sonism, confusion, rigidity, immobility, and,finally, increased risk of
mortality.34,35Dysfunctions in dopamine D2 receptors were suggested
as a main mechanism predisposing patients with LBD for NSR.36
However, it has been shown that frequency of severe NSR differs with different atypical APDs (the highest olanzapine and the lowest with thioridazine and clozapine), which may indicate that there may
be also other mechanisms involved.34
In patients with AD and VAD, the high degree of heterogeneity of
studies leads to contradictoryfindings and as a result, conclusions are
not easily drawn (eg, a meta-analysis of observational studies indicated
no association between APDs use and mortality of patients with AD).37
However, after excluding key contributors to the heterogeneity, the pooled relative risk of a systematic meta-analysis showed 2.08 (95% CI 1.39e3.13) higher risk of death with APDs. In another study of a cohort of 45,894 patients with AD with a 12-year follow-up, APD use showed an increase of mortality rate [2.28 (2.20e2.35)] compared with no
expo-sure.38Similarly, results from a double-blind randomized clinical trial
suggested a persistent risk of increased mortality with the long-term use
of antipsychotics in patients with AD.6Contradicting our results,
sys-tematic meta-analysis37 found no difference in mortality among
pa-tients with AD between typical and atypical APD use. However, evidence from broader studies corroborates our results. A recent publication
including participants with APD use of all ages39indicates higher
mor-tality with atypical antipsychotics compared with typical. In addition, authors also found differences among individual atypical APD use. To our knowledge, there is only 1 study including only patients with VAD, and it reports no increase in mortality risk with atypical APD use
(compared with no APD treatment).23However, when stratified for
in-dividual agents, there was a trend showing that patients exposed to
quetiapine had a 1.13 times higher risk of mortality (95% CI 0.92e1.37) compared with nonexposed patients with VAD.
Several mechanisms have been discussed in conjunction with increased mortality with antipsychotic treatment. However, studies exploring potential mechanisms underpinning higher mortality for APD treatment among patients with AD and VAD are scarce. Generally, atypical APDs, which occupy dopamine D2 receptors only transiently, are thought to be associated with lower risk of extrapyramidal
symp-toms.40However, it is complicated by the fact that with increasing doses
the risk of extrapyramidal symptoms equals the risk caused by typical APDs. When the dose of APDs is considered, atypical APDs are not safer than typical antipsychotics in relation to the development of
parkin-sonism.41 Despite this, a misconception that atypical APDs are less
harmful than typical APDs is still common. Morevoer, patients with AD in our sample were younger and had lower comorbidity and MMSE score at the time of dementia diagnosis compared with patients with
VAD and mixed dementia. It is possible that it could influence a decision
to prescribe higher doses of atypical APDs to patients with AD and thereby contribute to obtained results.
Moreover, cardio- and cerebrovascular events have previously
been reported to be a serious side effect of APD treatment.42High rates
of ventricular arrhythmias and cardiac arrest have been shown to be
associated with typical APDs43 and high rates of venous
thrombo-embolism with atypical APDs.44Both typical and atypical APDs have
been associated with QT prolongation.45Because of coexisting cardiac
disease, patients with VAD and mixed dementia may be particularly
susceptible for both types of APDs. Further, Moretti et al46did notfind
any difference in the risk of stroke or myocardial infarction between atypical and typical APDs in patients with VAD. In patients with AD, only 1 study reported higher incidence of cerebrovascular adverse events (an increased risk of transient ischemic attack) associated with
atypical APDs but not typical APDs.47
Moreover, BPSD may lead to risk-taking behavior48 that adds
additional burden to the ongoing neurodegeneration. Thus, BPSDs, which motivate both typical and atypical APD use, are also a major reason for poor functioning in patients with dementia, contribute to
faster functional and cognitive decline,49 and increase the risk of
death.16However, typical APDs and atypical APDs are usually
pre-scribed for the same symptoms. We may, thus, expect that when comparing patients with different APD differences in survival may be more due to effects of drugs than symptoms they are prescribed for.
It should be emphasized that this is an observation study using registry data. Therefore, we cannot assume causal relationship be-tween APDs and mortality or obtain a conclusive and unambiguous result. Designs of this type, however, may lead to generation of new hypotheses and give solid foundation for further studies. Our study
contributes to research aiming at defining predictors of mortality of
dementia patients with antipsychotic treatment. Strength of the study is a large sample size and long follow-up time. It is, therefore, possible that sample sizes included in previous studies on individual diagnosis were underpowered to detect associations. Moreover, contrary to other studies, we studied both inpatients and outpatients. It is important to note that in our sample almost every third person with APDs was living in a special residency at the time of dementia compared with 10% of patients without APD treatment. We found that form of residency (living alone in ordinary housing, with a family member in ordinary housing, or in a special residency) was an important confounder, with special residency increasing the risk of death in all dementia diagnoses. What is more, rate of APD prescrip-tion in our study is lower compared with rates previously reported in
different countries.50 We indeed observed a decrease in APD
pre-scriptions between 2007 and 2015 in Sweden (unpublished results). SveDem was established in 2007 with the main aim to increase the quality of care. At the same time, it is a useful tool for physicians to monitor and control their own activity. This could contribute to the
changes in pharmacologic management of psychiatric symptoms in people in geriatric care. Moreover, in Sweden Selective, serotonin
reuptake inhibitors are recommended as afirst-line treatment for
ir-ritability, agitation, and anxiety. Consequently, the use of antide-pressants is common in Sweden with every second person in special
residency being treated with antidepressants.51 It is, therefore,
possible that to some extent the APD use was replaced by the antidepressant treatment.
Limitations
Our study has several limitations. It is also possible that physicians
in our cohort prescribed typical APDs more frequently“as needed” for
acute symptoms and atypical medication was used in smaller doses but for a continous period. Furthermore, we cannot rule out a scenario in which atypical APDs were prescribed in higher doses than typical APDs in patients with AD. Our study may also be subject to other
uncon-trolled confounding and indication bias. Moreover, stratification for
dementia type and class of APDs together may have led to a sample size that was too small to detect the association in some of the categories. Conclusions and Implications
Although we cannot rule out the influence of residual confounding
(eg BPSD), these results would seem to support studies that suggest cautious APD prescribing for patients with dementia.
Acknowledgments
The authors are grateful to (SveDem,www.svedem.se). We thank
all patients, caregivers, reporting units, and coordinators in SveDem as well as SveDem steering committee. Special thanks to Johan Fastbom for extracting the data from the prescribed drug registry. We also
thank the Margaretha af Ugglas’ Foundation.
References
1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. Arlington, VA: American Psychiatric Publishing; 2013. Fifth Edition (DSM-5).
2. Brodaty H, Seeher K, Gibson L. Dementia time to death: A systematic literature review on survival time and years of life lost in people with dementia. Int
Psychogeriatr 2012;24:1034e1045.
3. Garcia-Ptacek S, Farahmand B, Kåreholt I, et al. Mortality risk after dementia diagnosis by dementia type and underlying factors: A cohort of 15,209 patients
based on the Swedish Dementia Registry. J Alzheimers Dis 2014;41:467e477.
4. Nonino F, De Girolamo G, Gamberini L, Goldoni CA. Survival among elderly
Italian patients with dementia treated with atypical antipsychotics:
Observational study. Neurol Sci 2006;27:375e380.
5. García-Ptacek S, Kåreholt I, Farahmand B, et al. Body mass index and mortality in incident dementia: A cohort study on 11,398 patients from SveDem, the
Swedish Dementia Registry. J Am Med Dir Assoc 2014;15:447.e1e447.e7.
6. Ballard C, Hanney ML, Theodoulou M, et al. The dementia antipsychotic
withdrawal trial (DART-AD): Long-term follow-up of a randomised
placebo-controlled trial. Lancet Neurol 2009;8:151e157.
7. Gustafsson M, Sandman PO, Karlsson S, et al. Association between behavioral
and psychological symptoms and psychotropic drug use among old people with cognitive impairment living in geriatric care settings. Int Psychogeriatr
2013;25:1415e1423.
8. Lyketsos CG, Lopez O, Jones B, et al. Prevalence of neuropsychiatric symptoms in dementia and mild cognitive impairment: Results from the cardiovascular
health study. JAMA 2002;288:1475e1483.
9. Tan LL, Wong HB, Allen H. The impact of neuropsychiatric symptoms of
de-mentia on distress in family and professional caregivers in Singapore. Int
Psychogeriatr 2005;17:253e263.
10.Steele C, Rovner B, Chase GA, Folstein M. Psychiatric symptoms and nursing
home placement of patients with Alzheimer’s disease. Am J Psychiatry 1990;
147:1049e1051.
11.Chan DC, Kasper JD, Black BS, Rabins PV. Presence of behavioral and
psycho-logical symptoms predicts nursing home placement in community-dwelling elders with cognitive impairment in univariate but not ultivariate analysis.
J Gerontol Ser A 2003;58:548e554.
12. Scarmeas N, Brandt J, Albert M, et al. Delusions and hallucinations are associ-ated with worse outcome in Alzheimer disease. Arch Neurol 2005;62:
1601e1608.
13. Schneider LS, Dagerman KS, Insel P. Risk of death with atypical antipsychotic drug treatment for dementia: Meta-analysis of randomized placebo-controlled
trials. JAMA 2005;294:1934e1943.
14. Wang PS, Schneeweiss S, Avorn J, et al. Risk of death in elderly users of
conventional vs atypical antipsychotic medications. N Engl J Med 2005;353:
2335e2341.
15. Gill SS, Bronskill SE, Normand ST, et al. Antipsychotic drug use and mortality in
older adults with dementia. Ann Intern Med 2007;146:775e786.
16. Lopez OL, Becker JT, Chang YF, et al. The long-term effects of conventional and atypical antipsychotics in patients with probable Alzheimer’s disease. Am J
Psychiatry 2013;170:1051e1058.
17. Raivio MM, Laurila JV, Strandberg TE, et al. Neither atypical nor conventional antipsychotics increase mortality or hospital admissions among elderly pa-tients with dementia: A two-year prospective study. Am J Geriatr Psychiatry
2007;15:416e424.
18. Gardette V, Lapeyre-Mestre M, Coley N, et al. Antipsychotic use and mortality risk in community-dwelling Alzheimer’s disease patients: Evidence for a role of
dementia severity. Curr Alzheimer Res 2012;9:1106e1116.
19. Selbæk G, Aarsland D, Ballard C, et al. Antipsychotic drug use is not associated with long-term mortality risk in Norwegian nursing home patients. J Am Med
Dir Assoc 2016;17:464.e1e464.e7.
20. Hulshof TA, Zuidema SU, Ostelo RWJG, Luijendijk HJ. The mortality risk of
conventional antipsychotics in elderly patients: A systematic review and meta-analysis of randomized placebo-controlled trials. J Am Med Dir Asso
2015;16:817e824.
21. Schneider LS, Tariot PN, Dagerman KS, et al. Effectiveness of atypical antipsychotic drugs in patients with Alzheimer’s disease. N Engl J Med 2006;355:1525e1538.
22. Sturm AS, Trinkley KE, Porter K, Nahata MC. Efficacy and safety of atypical
antipsychotics for behavioral symptoms of dementia among patients residing in long-term care. Int J Clin Pharm 2018;40:135e142.
23. Sultana J, Chang CK, Hayes RD, et al. Associations between risk of mortality and atypical antipsychotic use in vascular dementia: A clinical cohort study: An-tipsychotics and mortality in vascular dementia. Int J Geriatr Psychiatry 2014;
29:1249e1254.
24. Religa D, Fereshtehnejad S-M, Cermakova P, et al. SveDem, the Swedish
dementia registrydA tool for improving the quality of diagnostics, treatment and care of dementia patients in clinical practice. PLOS One 2015;10:e0116538.
25. World Health, Organization. The ICD-10 classification of mental and
behavioural disorders: Clinical descriptions and diagnostic guidelines. Geneva:
World Health Organization; 1992.
26. McKeith IG. Consensus guidelines for the clinical and pathologic diagnosis of dementia with Lewy bodies (DLB): Report of the consortium on DLB
International Workshop. J Alzheimers Dis 2006;9:417e423.
27. Brun A, Englund B, Gustafson L, et al. Clinical and neuropathological criteria for
frontotemporal dementia. J Neurol Neurosurg Psychiatry 1994;57:416e418.
28. Emre M, Aarsland D, Brown R, Dubois B. Clinical diagnostic criteria for
de-mentia associated with Parkinson’s disease. Mov Disord 2007;22:1689e1707.
29. Schemper M, Smith TL. A note on quantifying follow-up in studies of failure
time. Control Clin Trials 1996;17:343e346.
30. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying
prognostic comorbidity in longitudinal studies: Development and validation. J Chroni Dis 1987;40:373e383.
31. Goldmann Gross R, Siderowf A, Hurtig HI. Cognitive impairment in Parkinson’s
disease and dementia with Lewy bodies: A spectrum of disease. Neurosignals
2008;16:24e34.
32. Therneau T. A Package for Survival Analysis in S. version 2.38. 2015. Available
at: https://CRAN.R321project.org/package¼survival. Accessed December 1,
2017.
33. R Core Team. R. A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. 2013. Available at:http://
www.R-project.org/. Accessed December 1, 2017.
34. Aarsland D, Perry R, Larsen JP, et al. Neuroleptic sensitivity in Parkinson’s
disease and Parkinsonian dementias. J Clin Psychiatry 2005;66:633e637.
35. McKeith I, Fairbairn A, Perry R, et al. Neuroleptic sensitivity in patients with
senile dementia of Lewy body type. BMJ 1992;305:673e678.
36. Piggott MA, Perry EK, Marshall EF, et al. Nigrostriatal dopaminergic activities in dementia with Lewy bodies in relation to neuroleptic sensitivity: Comparisons with Parkinson’s disease. Biol Psychiatry 1998;44:765e774.
37. Zhai Y, Yin S, Zhang D. Association between antipsychotic drugs and mortality in older persons with Alzheimer’s disease: A systematic review and meta-analysis. J Alzheimers Dis 2016;52:631e639.
38. Nielsen RE, Lolk A, Valentin JB, Andersen K. Cumulative dosages of
antipsychotic drugs are associated with increased mortality rate in patients
with Alzheimer’s dementia. Acta Psychiatr Scand 2016;134:314e320.
39. Martín Arias LH, Treceño Lobato C, Pérez García S, et al. Risk excess of mortality and use of antipsychotics: A caseenoncase study. Int Clin Psychopharmacol 2017;32:1e5.
40. Wirshing WC. Movement disorders associated with neuroleptic treatment.
J Clin Psychiatry 2001;62:15e18.
41. Rochon PA, Stukel TA, Sykora K, et al. Atypical Antipsychotics and
Parkin-sonism. Arch Intern Med 2005;165:1882e1888.
E. Schwertner et al. / JAMDA 20 (2019) 323e329 328
42. Kahl KG, Westhoff-Bleck M, Krüger THC. Effects of psychopharmacological treatment with antipsychotic drugs on the vascular system. Vasc Pharmacol
2018;100:20e25.
43. Liperoti R, Gambassi G, Lapane KL, et al. Conventional and atypical
antipsy-chotics and the risk of hospitalization for ventricular arrhythmias or cardiac arrest. Arch Intern Med 2005;165:696.
44. Hägg S, Jönsson AK, Spigset O. Risk of venous thromboembolism due to
antipsychotic drug therapy. Expert Opin Drug Saf 2009;8:537e547.
45. Aronow WS, Shamliyan TA. Effects of atypical antipsychotic drugs on QT
interval in patients with mental disorders. Ann Transl Med 2018;6:147. 46. Moretti R, Torre P, Antonello RM, et al. Olanzapine as a possible treatment of
behavioral symptoms in vascular dementia: Risks of cerebrovascular events.
J Neurol 2005;252:1186e1193.
47.Imfeld P, Bodmer M, Schuerch M, et al. Risk of incident stroke in patients with
Alzheimer disease or vascular dementia. Neurology 2013;81:910e919.
48.Ha J, Kim E-J, Lim S, et al. Altered risk-aversion and risk-taking behaviour in patients with Alzheimer’s disease: Risky behaviour in Alzheimer’s patients.
Psychogeriatrics 2012;12:151e158.
49.Palmer K, Lupo F, Perri R, et al. Predicting disease progression in Alzheimer’s disease: The role of neuropsychiatric syndromes on functional and cognitive decline. J Alzheimer Dis 2011;24:35e45.
50.Kirkham J, Sherman C, Velkers C, et al. Antipsychotic use in dementia. Can J
Psychiatry 2017;62:170e181.
51.Gustafsson M, Karlsson S, Gustafson Y, Lövheim H. Psychotropic drug use
among people with dementiadA six-month follow-up study. BMC Pharmacol Toxicol 2013;14:56.
Supplemental Table 1
Cox Multivariate Analyses Assessing All-Cause Mortality Risk HR (95% CI) Associated With APD in VAD in the First/After 1 Year After Diagnosis: Total (Typicalþ Atypical), Typical, and Atypical
Crude Adjusted*
>1 y
Total 2.02 (1.65e2.46) 1.61 (1.28e2.04)y Typical* 2.21 (1.50e3.26) 1.79 (1.14e2.82)y Atypical* 1.98 (1.57e2.50) 1.54 (1.17e2.02)y Typicalz 1.12 (0.71e1.74) 1.18 (0.70e1.99) <1 y
Total 1.45 (1.27e1.65) 1.24 (1.07e1.44)x Typical* 1.78 (1.38e2.29) 1.59 (1.19e2.11)y Atypical* 1.39 (1.20e1.62) 1.20 (1.01e1.42)x Typicalz 1.26 (0.94e1.69) 1.31 (0.94e1.83) *Adjusted for age, sex, MMSE, CCI, and living arrangement and Ref: No APD.
yP< .001. zRef: Atypical. xP< .005.
Supplemental Table 2
Cox Multivariate Analyses Assessing All-Cause Mortality Risk HR (95%CI) Associated With APD in VAD in the First/After 1 Year After Diagnosis: Total (Typicalþ Atypical), Typical, and Atypical
Crude Adjusted*
Total
Total* 1.68 (1.6e1.77)y 1.38 (1.30e1.46)y Typical* 1.73 (1.56e1.91)y 1.35 (1.20e1.51)y Atypical* 1.65 (1.55e1.75)y 1.39 (1.29e1.48)y Typicalz 1.05 (0.93e1.18) 0.97 (0.85e1.11) AD
Total* 1.72 (1.53e1.94)y 1.35 (1.19e1.54)y Typical* 1.47 (1.16e1.86)y 1.05 (0.81e1.36) Atypical* 1.78 (1.54e2.05)y 1.51 (1.29e1.76)*
Typicalz 0.83 (0.63e1.09) 0.7 (0.52e0.94)*
Mix
Total* 1.86 (1.65e2.09)y 1.32 (1.16e1.50)*
Typical* 1.97 (1.61e2.41)y 1.32 (1.06e1.64)*
Atypical* 1.81 (1.57e2.1)y 1.34 (1.14e1.57)*
Typicalz 1.09 (0.85e1.39) 0.98 (0.75e1.28) VADx
Total* 1.58 (1.42e1.77)y 1.33 (1.17e1.51)y Typical* 1.89 (1.53e2.33)y 1.64 (1.29e2.09)y Atypical* 1.53 (1.35e1.74)y 1.28 (1.1e1.48)y Typicalz 1.23 (0.97e1.57) 1.28 (0.97e1.70) LBD
Total* 1.12 (0.93e1.33) 1.07 (0.89e1.30) Typical* 1.48 (0.99e2.22) 1.39 (0.91e2.14) Atypical* 1.05 (0.86e1.28) 1.02 (0.83e1.26) Typicalz 1.41 (0.91e2.19) 1.37 (0.86e2.17) FTD
Total* 1.21 (0.81e1.81) 1.15 (0.69e1.94) Typical* 1.29 (0.57e2.89) 1.27 (0.52e3.11)*
Atypical* 1.04 (0.64e1.69) 1.11 (0.60e2.05) Typicalz 1.24 (0.49e3.15) 1.15 (0.40e3.33) UNS
Total* 1.59 (1.43e1.76)* 1.33 (1.17e1.50)*
Typical* 1.51 (1.23e1.86)* 1.24 (0.95e1.60) Atypical* 1.59 (1.41e1.79)* 1.33 (1.15e1.53)*
Typicalz 0.95 (0.75e1.21) 0.93 (0.7e1.24) Other
Total* 1.37 (0.98e1.92) 1.21 (0.78e1.85) Typical* 1.38 (0.69e2.78) 1.01 (0.45e2.28) Atypical* 1.36 (0.92e2.00) 1.21 (0.71e2.07) Typicalx 1.02 (0.46e2.25) 0.83 (0.32e2.19) *Adjusted for age, sex, MMSE, CCI, and living arrangemnet and Ref: No APD.
yP< .001. zRef: Atypical.
xRef: Assuming proportional hazard.
E. Schwertner et al. / JAMDA 20 (2019) 323e329 329.e1
0.00 0.25 0.50 0.75 1.00 0 1000 2000 3000 Time in days Su rv iv a lr a te
Treatment No Typical Atypical
AD 0.00 0.25 0.50 0.75 1.00 0 1000 2000 3000 Time in days Sur vi val ra te
Treatment No Typical Atypical Mix 0.00 0.25 0.50 0.75 1.00 0 1000 2000 3000 Time in days S ur vi va lra te
Treatment No Typical Atypical VAD 0.00 0.25 0.50 0.75 1.00 0 1000 2000 3000 Time in days Su rv iv al ra te
Treatment No Typical Atypical
LBD 0.00 0.25 0.50 0.75 1.00 0 1000 2000 3000 Time in days Sur viv a lr a te
Treatment No Typical Atypical
FTD 0.00 0.25 0.50 0.75 1.00 0 1000 2000 3000 Time in days Sur viv al ra te
Treatment No Typical Atypical
UNS 0.00 0.25 0.50 0.75 1.00 0 1000 2000 3000 Time in days Su rv iv al rat e
Treatment No Typical Atypical
Other
Supplemental Fig. 1. Survival functions stratified by dementia diagnoses from Cox hazard regression adjusted for age, sex, MMSE, CCI, and form of living for patients receiving typical, atypical, and no APD treatment. Y-axis, estimated survival percentages; X-axis, number of days.
E. Schwertner et al. / JAMDA 20 (2019) 323e329 329.e2
View publication stats View publication stats