Polypharmacy Results in Functional Impairment in Mice: Novel Insights Into Age and Sex Interactions

Abstract

Males and females may respond differently to medications, yet knowledge about sexual dimorphisms in the effects of polypharmacy remains limited, particularly in aging. This study aimed to assess the effect of high Drug Burden Index (DBI) polypharmacy treatment compared to control on physical function and behavior in young and old, male and female mice. We studied whether age and sex play a role in physical function and behavior following polypharmacy treatment and whether they are paralleled by differences in serum drug levels. Young (2.5 months) and old (21.5 months), C57BL/6 mice were randomized to control or high DBI polypharmacy treatment (simvastatin, metoprolol, oxybutynin, oxycodone, and citalopram; n = 6–8/group) for 4–6 weeks. Compared to control, polypharmacy reduced physical function (grip strength, rotarod latency, gait speed, and total distance), middle zone distance (increased anxiety), and nesting score (reduced activities of daily living) in mice of both ages and sexes (p < .001). Old animals had a greater decline in nesting score (p < .05) and midzone distance (p < .001) than young animals. Grip strength declined more in males than females (p < .05). Drug levels at steady state were not significantly different between polypharmacy-treated animals of both ages and sexes. We observed polypharmacy-induced functional impairment in both age and sex groups, with age and sex interactions in the degree of impairment, which were not explained by serum drug levels. Studies of the pathogenesis of functional impairment from polypharmacy may improve management strategies in both sexes.

Polypharmacy (concurrent use of 5 or more medications) is a major public health challenge in the context of a growing aging population with multimorbidity (1). Polypharmacy affects more than 15 million Americans aged 65 years and older, and its prevalence is higher in women (56.2%) than men (43.8%) (2). Females show marked differences in the physiology of aging, pharmacokinetics, pharmacodynamics, clinical presentation, and clinical outcomes of medications compared to males (3). Despite this, efficacy and safety data for commonly used medications have traditionally been based on clinical trials conducted predominantly in young and middle-aged males, with a limited representation of females and older adults (4,5). Sex differences in the long-term benefits and harms of medications are not well understood, especially when medications are used in combination and in older people (6).

Clinical epidemiological studies have demonstrated associations between polypharmacy and adverse geriatric outcomes, such as falls, frailty, and cognitive impairment (7). Furthermore, there is a dose-dependent relationship between the Drug Burden Index (DBI) and adverse geriatric outcomes (8–11). However, interpretations of observational studies are limited by potential residual confounding and confounding by indication, which makes it difficult to distinguish the impacts of age, sex, and gender or to establish causation. In addition, there are ethical and feasibility barriers to interventional studies investigating these exposures in humans (12). The DBI is a measure of cumulative exposure to drugs with anticholinergic and sedative effects. It is calculated using the equation, Drug Burden = D/(δ + D), where D is the daily dose taken, and δ is the minimum licensed daily dose, which is used as an estimate of the DR₅₀ (daily dose required to achieve 50% of the maximal effect at steady state) (13).

Using a polypharmacy mouse model (14), we recently found that short-term (2–4 weeks) treatment with low DBI polypharmacy (DBI score ~0.5) resulted in impaired physical function in old but not in young male mice (14). We subsequently applied our polypharmacy mouse model to chronic exposure (from age 12 to 24 months) with a range of regimens and found that low DBI polypharmacy, and to a greater extent high DBI polypharmacy, caused frailty and functional impairment in aging male mice (12).

To date, no preclinical studies have investigated the effect of polypharmacy on functional outcomes in females. Sex-specific biological differences have been reported in aging animals (15). Sex differences have been observed in some preclinical studies that studied individual medications for the treatment of disease or in studies of pharmaceutical and nutritional interventions targeting aging (16). Therefore, we hypothesized that the effects of high DBI polypharmacy on physical and behavioral functions may differ with age and sex.

The primary aim of the study was to assess the changes in functional outcomes after 4–6 weeks of high DBI polypharmacy treatment compared to controls in young and old, male and female mice. Our secondary aims were to investigate age and sex interactions in functional outcomes following polypharmacy treatment. We also determined whether any age or sex interactions with function were paralleled by differences in serum drug levels.

Method

All experiments were performed in accordance with the guidelines of the National Health and Medical Research Council of Australia and approved by the Animal Ethics Committee of the Northern Sydney Local Health District, Sydney, Australia (RESP/16/348). Healthy young and old C57BL/6J (B6) mice of both sexes (young males n = 12, old males n = 16; young females n = 12, old females n = 14) were sourced and housed at the Kearns facility (Kolling Institute of Medical Research, Sydney, Australia). Mice were obtained in 5 cohorts 2–4 weeks apart. The Kearns facility obtains mice from the Animal Resource Centre in Perth, WA, Australia and breeds them for up to 10 generations to maintain genetic identity.

Animals were housed in groups of up to 5 animals per cage, maintained on a 12-hour light–dark cycle (lights on at 07:00, off at 19:00). They had ad libitum access to water and food (Rat and Mouse Premium Breeder Diet containing 23% protein, Gordon Specialty Feed, Yanderra, NSW, Australia). At age 2.5 months and 21.5 months for young and old animals, respectively, animals were individually housed and received nonmedicated control feed (Standard Meat Free Mouse and Rat Feed, Specialty Feeds, WA, Australia). At age 4 months and 23 months, animals of both ages and sexes were randomly assigned to follow either nonmedicated control feed or high DBI polypharmacy feed. The randomization process involved stratification of each animal cohort by age and sex. Within each age–sex group, animals were randomly assigned in a 1:1 ratio to either control group or high DBI polypharmacy group using a random number generator in Microsoft Excel 2019 (Microsoft Inc., Redmond, WA; young male control n = 6, young male polypharmacy n = 6, young female control n = 6, young female polypharmacy n = 6, old male control n = 7, old male polypharmacy n = 8, old female control n = 8, and old female polypharmacy n = 6). The drugs used in this study were selected from drug classes that are commonly prescribed in older people (17), which have similar pharmacokinetics and pharmacodynamics in humans and mice and are not known to be toxic when given to healthy mice. The high DBI polypharmacy regimen contained metoprolol 350 mg/kg/day, simvastatin 20 mg/kg/day, and 3 drugs with anticholinergic or sedative effects: oxycodone and oxybutynin at their minimum effective doses (5 and 27.2 mg/kg/day, respectively) and citalopram at a dose 50% higher than the minimum dose (15 mg/kg/day). Medication was administered in the chow and water as previously described (12). Body weight, food, and water intake were assessed weekly. Mortality included both mice found deceased and those euthanized on the advice of the Kearns facility veterinarian. Date of death was recorded.

Functional assessments were conducted for all animals (male and female, young and old) before and after treatment (control and polypharmacy groups). This occurred over a 3-week period at age 3 and 22 months (pretreatment) and at age 5 and 24 months (posttreatment) for young and old, respectively. Functional testing (described in detail in the following sections) was conducted in the following sequence: Week 1—open field and rotarod tests, Week 2—forelimb grip strength test, and Week 3—nesting. Forelimb grip strength and rotarod tests were conducted between 13:00 and 17:00. Open field test and nesting assessment were conducted between 09:00 and 12:00. On the day of each functional test, mice were moved to the testing room and allowed to habituate for 30 minutes before testing. All functional tests were performed under white light except the open field test, which was performed under red light. Researchers were not blinded during animal testing but were blinded during the analysis of the video of the open field test. Within 2 weeks of completing the final functional measurements, animals were euthanized. Serum was collected from the inferior vena cava for biochemistry including serum drug levels, and tissue was collected for future analyses.

Forelimb Grip Strength Test

A grip strength test was used to assess the muscular strength of the forelimbs with a Grip Strength Meter (TSE-Systems, MO). It consisted of a horizontal metal bar connected to a force transducer. To measure forelimb grip strength, the mouse was held near the base of its tail and allowed to grasp the metal bar with its forepaws. The mouse was then positioned horizontally and gently pulled back from the bar until it released its grip. The peak force for the forepaws was automatically recorded by the Grip Strength Meter. Forelimb grip strength for each mouse was measured 5 times with a 20-minute resting interval between each trial. The mouse forelimb grip strength was expressed in units of gram force (gf).

Rotarod Test

A rotarod test was used to measure motor coordination and balance. It was performed as previously reported (14) using a rotarod device (Orchid Scientific and Innovative India Pvt Ltd, India). The mice were tested in 3 trials with a 30-minute resting interval between each trial. Each mouse was placed on the rod, facing away from the assessor. The rotarod device was gradually accelerated over 300 seconds (from 4 to 40 rpm) following habituation phases (60, 20, and 0 seconds; Trials 1–3, respectively) during which the rod rotated at a fixed speed of 4 rpm. The time taken for each mouse to fall from the rod (rotarod latency) was recorded.

Open Field Test

An open field test was used to measure locomotor activity and anxiety-related behavior (18). The testing apparatus consisted of a square Perspex open field arena (50 cm wide × 50 cm long × 50 cm high). Each animal was placed in the center of the arena. Activities in the open field were recorded for 5 minutes using a digital video camera situated above the arena. Total distance traveled (meters), gait speed (meters per second), and middle zone distance traveled (meters) were measured using the ANY-maze Video Tracking System Software (ANY-maze, Stoelting Co, IL). Gait speed (m/s) was calculated by dividing the total distance (meters) by the total mobile time (seconds). Middle zone distance percentage (%) was calculated by dividing the middle zone distance traveled (meters) by the total distance traveled (meters). In the present study, the middle zone percentage was used as a measure of anxiety-related behavior (19).

Nesting

Nesting was used to measure animal’s activities of daily living, as previously described (12). Nesting materials (Bed-R’Nest 8G Irradiated 60 × 25 × 35) were sourced from Tecniplast Australia Laboratory Animal Equipment Pty Ltd (Sydney, Australia). Animals were introduced to this nesting material at the beginning of the study (2.5 and 22 months for young and old animals, respectively). Before testing, animals were acclimatized for 2 days to build a nest by removing their plastic house and replacing the nesting material with a fixed amount (approximately 8 g) of new nesting material. After 48 hours, the nesting material was replaced again with new nesting material and nests were scored 24 hours later. Using the nesting scoring method adapted from Hess et al. (20), nests were divided into 4 quarters and each quarter was scored from 0 to 5 (0: undisturbed, 1: disturbed, 2: flat, 3: cup, 4: incomplete dome, and 5: full dome). Each cage was scored by 2 independent researchers, and the total score was determined by the sum of the scores of each quarter.

Serum Drug Levels

Blood was obtained from the inferior vena cava after 6–8 weeks of treatment (steady state). Serum drug levels were determined using the Shimadzu triple quadrupole mass spectrometer coupled with ultra-high performance liquid chromatography following the procedures previously reported (21). The expected therapeutic serum/plasma range for human are tenivastatin 3.2–8 ng/mL (22), metoprolol 5–212 ng/mL (23), citalopram 50–120 ng/mL (24), oxycodone 13–87 ng/mL (25), and oxybutynin 2.1–4.2 ng/mL (26). The maximum value on the standard curve for metoprolol was 500 ng/mL and results above this value may be inaccurate.

Statistical Analysis

All statistical analyses were performed using SPSS Statistics for Windows, version 24.0 (IBM Corp, NY). The study was powered to detect a difference in the open field test, informed by the difference in locomotor activity observed between young and old male mice in our previous study (14) with more than 80% power and (α = 0.05; n = 5/group).

Baseline function/characteristics

To control for multiple comparisons across outcomes, the Benjamini Hochberg procedure was applied with a false discovery rate set to 0.1. Comparison of baseline function between control and treatment groups used the independent Student’s t test or Mann–Whitney U test for parametric and nonparametric variables, respectively. Mortality between the control and high DBI polypharmacy groups in old males and old females after randomization was compared using Kaplan–Meier analysis with log-rank test and chi-square analysis.

Treatment effect, age and sex interactions with treatment effect

For each functional outcome, a repeated measures mixed model was used to assess the effect of high DBI polypharmacy treatment group relative to controls. We assessed age and sex interactions with the effects of high DBI polypharmacy using the following interaction terms: Age × Treatment, Sex × Treatment, and Age × Sex × Treatment in the repeated measures mixed model for each outcome. All pretreatment and posttreatment trial data for each animal were included using a compound symmetry covariance to account for the within mouse correlation. Maximum likelihood estimation was applied.

Age and sex differences in serum drug levels

To evaluate the age and sex differences in serum drug levels in polypharmacy-treated animals, analysis of variance with post hoc Tukey for adjustments for multiple comparisons or Dunn’s test was applied for parametric and nonparametric variables, respectively.

Results

Animal Well-Being, Drug Intake, and Mortality

The polypharmacy diet was tolerated by the animal cohort with no significant change in body weight and food or water intake compared to control (Supplementary Table 1). No clinically significant differences were observed between polypharmacy (intervention) and control groups in serum biochemistry results, such as creatinine, urea, liver function tests, and electrolytes, at the end of the study (Supplementary Table 2). There were no significant differences in baseline functional outcomes between treatment and control groups. A total of 11 deaths occurred after randomization, 5 in old female high DBI polypharmacy group (1 died from an experimental mishap and was censored in the analysis), 2 in old male high DBI polypharmacy group, 2 in old female control group, and 2 in old male control group. No statistically significant differences in mortality were found following Kaplan–Meier or chi-square analysis between control and high DBI polypharmacy groups in old males and old females (Supplementary Figure 1).

Impact of Treatment on Functional Outcomes

Compared to controls, high DBI polypharmacy treatment resulted in significant impairment in several functional measures, which included forelimb grip strength (p < .001), rotarod latency (p < .001), gait speed (p < .001), total distance traveled in the open field (p < .001), middle zone distance percentage (p < .001), and nesting score (p < .001) in mice of both ages and sexes (Figures 1 and 3).

Figure 1.

Grip strength (A), motor coordination (rotarod) (B), gait speed (open field speed) (C), mobility (open field distance) (D), anxiety (midzone distance percentage) (E) and activities of daily living (nesting) (F) for control and high DBI polypharmacy diets in young (2.5 months) and old (22 months) male and female C57B6 mice (n = 6-8 per group). The results are the estimated means with lower and upper bound 95% confidence intervals from a repeated measured mixed model with significance based on Type III tests of fixed effects using Benjamini Hochberg procedure to adjust for multiple comparisons. White and black dots represent control and high DBI polypharmacy treated animals, respectively. p < .05 pre-specified comparisons of treatment within age and sex. ^ɑp < .05 main effect comparing all treatment groups to control groups. ^βp < .05 for age*treatment interaction (Type III tests of fixed effects). ^γp < .05 for sex*treatment interaction (Type III tests of fixed effects).

Figure 3.

Pre and post mean data for each mouse for grip strength (A), rotarod latency (B), openfield speed (C), openfield distance (D), midzone distance percentage (E) and nesting (F) for control and high Drug Burden Index (DBI) polypharmacy diets in young (2.5 months) and old (22 months) male and female C57B6 mice (n = 6-8 per group). The results are presented as line graphs with pre and post mean data ± standard error of the mean. The group mean and variances were derived from each individual variance from repeated measures for every observation from each mouse. White and black dots represent control and high DBI polypharmacy treated animals, respectively.

Impact of Age and Treatment on Functional Outcomes

Among the whole cohort, there were statistically significant age and treatment interactions in nesting score and open field midzone distance percentage. When comparing the effect of age and treatment interaction on nesting score, old animals showed a significantly greater decline in nesting score compared to young animals following high DBI treatment (p < .05), a pattern that was also seen with open field midzone distance percentage (p < .001; Figures 1 and 3).

Impact of Sex and Treatment on Functional Outcomes

We found a statistically significant sex and treatment interaction in forelimb grip strength (p < .05). Across both sexes, forelimb grip strength was negatively correlated with high DBI treatment (p < .001). When comparing the effect of sex and treatment on forelimb grip strength, there was a significantly greater decline in forelimb grip strength in males compared to females following high DBI treatment (p < .05; Figures 1 and 3).

Impact of Polypharmacy on Serum Drug Levels

No statistically significant age or sex differences were found with serum drug levels of tenivastatin, metoprolol, citalopram, oxycodone, or oxybutynin in polypharmacy-treated animals (Figure 2).

Figure 2.

Mean serum parent drug levels (except for simvastatin, where the active metabolite tenisvastatin measured) of the high DBI polypharmacy-treated animals in young (6.5 months at the time of sampling) and old (25.5 months) male and female C57B6 mice (n = 6–8/group). Serum (A) tenivastatin, (B) metoprolol, (C) citalopram, (D) oxycodone, and (E) oxybutynin. The results are presented as Box and Whisker plots with 95% confidence intervals for each group. No statistical significant difference was found comparing drug levels between either sex or age groups.

Discussion

This is the first study to evaluate sexual dimorphisms in the impact of polypharmacy on functional outcomes in young and old mice. We demonstrated that high DBI polypharmacy impaired measures of behavior and physical function in C57BL/6 mice of both ages and sexes. This study also provided novel insights into the age and sex interactions with polypharmacy-induced functional impairment. In mice treated with high DBI polypharmacy, compared to control, we observed a significant reduction in forelimb grip strength, rotarod latency, gait speed, total distance in the open field, middle zone distance percentage, and nesting score. There was a significant interaction with polypharmacy and age, whereby old animals had a significantly greater decline in nesting score (a measure of activities of daily living) and midzone distance percentage (a measure of anxiety) compared to young animals. We observed significant interaction of polypharmacy and sex, such that males had a greater reduction in forelimb grip strength compared to females. There were no differences in serum drug levels between age or sex groups for the mice receiving high DBI polypharmacy.

It is increasingly appreciated that polypharmacy is linked to adverse health outcomes in old age. Clinical studies consistently demonstrate that medications with anticholinergic and sedative effects impair physical function (8–11,27). Several animal studies have also provided insights into the detrimental effects of polypharmacy on physical function in male mice (12,14). Using a polypharmacy mouse model, Huizer-Pajkos et al. (14) reported a significant decline in locomotor activity and rotarod latency in old, but not young male mice after 2–4 weeks of low DBI polypharmacy treatment (metoprolol, paracetamol, irbesartan, simvastatin, and citalopram) compared to controls. In a subsequent study in young adult male mice, Eroli et al. (28) examined the effects of a similar low DBI polypharmacy regimen (metoprolol, paracetamol, aspirin, simvastatin, and citalopram) in young male mice and found no significant difference in locomotor activity and rotarod latency between low DBI polypharmacy and control groups after 8 weeks of treatment. They reported a significant decline in exploratory behavior (reduced horizontal movement in open field test) and spatial working memory in the polypharmacy group (28). The results from the present preclinical study are consistent with and further extend these observations to high DBI polypharmacy treatment and to female mice. We found that irrespective of age and sex, 4–6 weeks of high DBI polypharmacy caused significant impairment in mobility, balance, motor coordination, forelimb muscle strength, anxiety-related behavior, and activities of daily living.

The present study also demonstrated significant age interactions in the degree of functional decline following polypharmacy treatment, with greater impairment in activities of daily living and anxiety-related behavior in old animals. The pathophysiologic mechanisms for these age interactions are likely to be multifactorial, which include age-related changes in pharmacokinetics and pharmacodynamics (29). While we did not observe any differences in steady-state serum drug levels between age groups, this does not exclude other potentially relevant pharmacokinetic changes, such as changes in the blood–brain barrier in old age. To the best of our knowledge, none of the medications in the polypharmacy regimen are known to cause functional impairment when used as short-term monotherapy in mice. In our previous chronic polypharmacy study in aging male mice, we found that monotherapy with citalopram or metoprolol, but not with simvastatin, oxybutynin, or oxycodone, resulted in functional impairment after 9–12 months of treatment (12). This is consistent with observational studies in large cohorts, which found that statins are not associated with a decline in physical function in humans (30,31). The results of this study align with those from previous preclinical studies, which demonstrated that relatively benign and commonly prescribed medications can have a deleterious effect on physical function when used in combination (12,14). Furthermore, in the setting of polypharmacy, old individuals may experience a greater decline in physical function compared to younger, although a sufficiently high-risk polypharmacy regimen can impair function in all age groups.

In addition to age interactions, there were sex interactions in the magnitude of functional impairment caused by polypharmacy treatment. Males were more severely affected by high DBI polypharmacy treatment than females in terms of forelimb grip strength. The mechanisms underlying this sex interaction are incompletely understood. While sex differences in response to polypharmacy have not been evaluated previously in mice, sex differences in responses to some of the monotherapies in the regimen have been evaluated in mice. A study of 1–5 months of oxybutynin in a mouse model of Alzheimer’s disease found that female but not male mice showed improved behavior on the elevated plus maze (32). Oxycodone administered acutely to C57BL/6J mice aged 10–12 weeks, resulted in increased locomotor activity in the open field over 60 minutes in females at 1, 3, and 10 mg/kg and in males at 3 and 10 mg/kg (33). Simvastatin does not extend the life span in male or female mice (34).

Studies reporting sex differences in anticholinergic and sedative drug-related functional impairment in humans have provided inconsistent results (35,36). This may partly reflect the heterogeneity in the study designs, study population, and medication regimens. Furthermore, sex-specific pharmacokinetic and pharmacodynamic differences may account for the disparities in drug effects between males and females. For instance, the plasma concentration of metoprolol is generally higher in females than in males, which leads to a greater reduction in exercise heart rate and systolic blood pressure in females (37). This increased effect may be partly attributed to the lower activity of cytochrome P450 2D6 (CYP2D6) in females (38), which decreases first-pass liver metabolism and increases the bioavailability of metoprolol in females (37). These findings are consistent with the direction of the trend observed in females compared to males in serum metoprolol levels in our study. In contrast, females have higher expression of CYP3A4 enzyme than males (39). As a result, females are able to metabolize simvastatin, oxycodone, and oxybutynin (all CYP3A4 substrates) at a faster rate than males. We did not observe any differences in these drug levels in our study. Other postulated mechanisms responsible for sex differences in the effect of polypharmacy may include differences in patterns of multimorbidity, drug use, genetic, and hormonal factors between males and females. Further research is needed to better understand the pathophysiology of the observed sex differences and how sex-specific mechanisms influence drug safety and efficacy.

Strengths and Limitations of the Study

To the best of our knowledge, there have been no prior animal studies assessing the effects of high DBI polypharmacy on physical function in females. The present study is the first to provide preclinical evidence demonstrating polypharmacy causes functional impairment in young and old female mice. Although many observational studies have demonstrated that increasing DBI is associated with impaired physical function in humans (8–11,27,40–43), a major limitation in the interpretation of these studies has been residual confounding related to heterogeneity in biopsychosocial characteristics, which limit interpretation of age, sex, gender associations, as well as attribution of causality (44). The polypharmacy mouse model provides the opportunity to systematically examine the impact of polypharmacy on functional outcomes without residual confounding (12,14). In addition, the functional measures used have previously been extensively validated as reliable measures of function in mice and are comparable to those in humans (45). This makes the results of this study translatable to patients in clinical practice.

This study has several limitations. First, animals in the intervention group were treated with a specific high DBI polypharmacy regimen for 4–6 weeks and were apparently healthy, without the disease. This differs from clinical practice where polypharmacy is used to treat disease(s), and medications may have both beneficial and adverse effects on global health outcomes. The findings of this study may not necessarily be applicable to animals being treated with different medication regimens or for different durations or with the disease. Different polypharmacy regimens may have different sex–drug interactions. To further understand the impact of medications on physical function, future studies may incorporate different combinations of monotherapy and dual therapy in the study design. Second, all behavioral experiments were performed during the light phase of the light–dark cycle (09:00 to 17:00), while mice are nocturnal. Third, animals were euthanized within 2 weeks after completing the posttreatment functional measurements, which means that the long-term implications of the observed functional changes are unclear. Future longitudinal studies are needed to determine the long-term consequences of impaired physical function, which may ultimately be an important determinant of morbidity and mortality. Fourth, the study was powered to detect a difference in the open field test but not to detect differences in secondary outcomes such as serum drug levels; however, adjustment for multiple comparisons was performed. Finally, the mouse model could establish the effects of sex, but not the more complex implications of gender on outcomes of polypharmacy.

Conclusions

High DBI polypharmacy resulted in significant impairment in functional outcomes in C57BL/6 mice of both ages and sexes. There were age and sex interactions in the degree of functional impairment following polypharmacy treatment. Key questions remain about the underlying mechanisms of these functional changes and how age and sex affect susceptibility to polypharmacy. Further mechanistic studies correlating polypharmacy-induced functional impairment with underlying pathophysiological processes may lead to better understanding and improvement in management strategies in both males and females.

Funding

J.M. and this study were funded by the Penney Ageing Research Unit, Royal North Shore Hospital, Australia. H.A. was supported by grants from the US National Institute on Aging at the National Institutes of Health (R01 AG047891, P30AG021342, and P30AG021342-16S1). R.d.C. is supported by the National Institute on Aging, National Institutes of Health. D.G. is supported by the Australian National Health and Medical Research Council Dementia Leadership Fellowship.

Conflict of Interest

None declared.

Author Contributions

S.N.H. conceptualized and designed the work, supervised acquisition, analysis, and interpretation of the data, and assisted with drafting and revising the manuscript. J.M. made substantial contributions to the design, the acquisition, statistical analysis, and interpretation of data, and drafting and revising the manuscript. H.W. made substantial contributions to animal care, experiments, acquisition of data, analysis, and interpretation of the data, and drafting and revising the manuscript. G.G. and T.T made substantial contributions to animal care, experiments, acquisition of data, interpretation of the data, and revising the manuscript. H.A. and D.G. made substantial contributions to the statistical analysis of data. S.E.H, D.L.C., and R.d.C. made substantial contributions to the conception and design of the work and to the interpretation of the data. All authors revised the manuscript and have approved the submitted version.