The role of sex, age and genetic polymorphisms of CYP enzymes on the pharmacokinetics of anticholinergic drugs

Abstract There is evidence that use of drugs with anticholinergic properties increases the risk of cognitive impairment, and increased exposure to these drugs potentiates this risk. Anticholinergic drugs are commonly used even with associated risk of adverse events. Aging, sex, and genetic polymorphisms of cytochrome P450 (CYP) enzymes are associated with alterations in pharmacokinetic processes, which increase drug exposure and may further increase the risk of adverse drug events. Due to the increasing burden of cognitive impairment in our aging population and the future of personalized medicine, the objective of this review was to provide a critical clinical perspective on age, sex, and CYP genetic polymorphisms and their role in the metabolism and exposure to anticholinergic drugs. Age‐related changes that may increase anticholinergic drug exposure include pseudocapillarization of liver sinusoidal endothelial cells, an approximate 3.5% decline in CYP content for each decade of life, and a reduction in kidney function. Sex‐related differences that may be influenced by anticholinergic drug exposure include women having delayed gastric and colonic emptying, higher gastric pH, reduced catechol‐O‐methyl transferase activity, reduced glucuronidation, and reduced renal clearance and men having larger stomachs which may affect medication absorption. The overlay of poor metabolism phenotypes for CYP2D6 and CYP2C19 may further modify anticholinergic drug exposure in a significant proportion of the population. These factors help explain findings of clinical trials that show older adults and specifically older women achieve higher plasma concentrations of anticholinergic drugs and that poor metabolizers of CYP2D6 experience increased drug exposure. Despite this knowledge neither age, sex nor CYP phenotype are routinely considered when making decisions about the use or dosing of anticholinergic medications. Future study of anticholinergic medication needs to account for age, sex and CYP polymorphisms so that we may better approach personalized medicine for optimal outcomes and avoidance of medication‐related cognitive impairment.


| INTRODUC TI ON
Subsequent to the publication of these guidelines, several new studies have identified exposure to anticholinergic agents as risk factors for mild cognitive impairment and dementia. [5][6][7] Clinical experience and research demonstrate an increased risk of adverse drug events, [8][9][10][11][12] cognitive impairment, [13][14][15][16][17][18] and mortality 12 related to the use of anticholinergic drugs in older adults. These adverse events can result in emergency department visits, 19 hospital admission, 20 or death 21 with older adults being at increased risk of these sequelae. [22][23][24] Due to variability in the anticholinergic activity of individual medications, one agent in isolation may fail to cause any noticeable effect but when two or three anticholinergic agents are combined the total anticholinergic burden can result in adverse events. 10,17,20,21,[25][26][27][28] Total medication exposure or anticholinergic burden depends upon the pharmacokinetic factors in the subject relating to the particular medication(s) consumed.
Age, sex, and genetic variation in cytochrome P450 (CYP) enzymes [29][30][31][32][33] are important factors leading to variability in drug metabolism and disposition. As a result, these factors may also contribute to variation in systemic drug exposure, response including resultant adverse events [22][23][24] and toxicity to a variety of drugs including anticholinergic medications. This review supports clinical decision-making surrounding anticholinergic medication use as we come to understand their potential risk for causing cognitive impairment and dementia, particularly in older female patients. Our analysis examines the effects of age, sex, and genetic polymorphisms of CYP2D6, CYP2C19, and CYP3A4 on the pharmacokinetics (absorption, distribution, metabolism, excretion), subsequent exposure, and pharmacologic response to anticholinergic medications. This review will provide clinicians with the pharmacokinetic considerations required when using anticholinergic medications.

| Data sources for review
The PubMed database was searched across all available dates (1950-January 2020) with the initial search terms age, sex, anticholinergic agent, and pharmacokinetics. Anticholinergic agents were considered any drugs appearing on the anticholinergic cognitive burden scale 34 as these were medications used commonly, and this anticholinergic scale is freely available for consultation.

| Role of sex on the absorption of anticholinergic medications
Some, but not all, studies showed that gastric and colonic emptying was slowed in women, potentially increasing the oral bioavailability of some drugs. [45][46][47][48][49][50][51] When stratified by age, the rate of gastric emptying for postmenopausal women and men was similar 52 and significantly faster than premenopausal (younger) women. 50 Gastric pH was higher in females 53 which may increase absorption of basic medications such as tricyclic antidepressants, many of which are quite potently anticholinergic. This difference in gastric pH was quantified by Feldman and Barnett in 1991 as a mean pH of 2.79 for women and 2.16 for men, which was due to reduced acid secretion in women. 54 The greater stomach size in men allowed for more fluid to be contained therein which may have improved both the rate and extent of dissolution of introduced oral dosage forms for men in comparison with women. By contrast, intestinal pH was not found to differ by sex. 55 Early reports suggested that the CYP3A4 substrates verapamil and midazolam had increased bioavailability in women. [57][58][59] However, in 2005 a detailed analysis of duodenal punch biopsies from 48 men and 45 women found no clinically meaningful sex difference in intestinal CYP3A4 content. 60

| Role of sex on the distribution of anticholinergic medications
In general, males are larger than females across the lifespan, with increased height, body mass index, and waist circumference. 62 Compared with men, women have increased adiposity. This difference in body composition has failed to show much difference in actual drug distribution and any differences attributable to body composition can largely be explained by differences in total body mass. 63 Distribution of drugs to the brain was dependent upon the lipophilic nature of the blood-brain barrier which favored passage and accumulation of lipophilic drugs. At this time, no statistically significant difference has been found between similarly aged women and men with respect to albumin permeability of the blood brain barrier 64 which likely can be extrapolated to at least some medications.
The brain is also protected by p-glycoprotein, which prevents drugs from accumulating in the brain by pumping them from brain capillary  71 Women had an increase in CYP3A4 activity measured as a greater clearance of CYP3A4 substrates such as the weakly anticholinergic antihypertensive medication nifedipine 72 and the weakly anticholinergic sedative alprazolam. 73 On average, the weight-normalized clearance of alprazolam to active metabolites 74 and nifedipine to inactive metabolites 75 was mainly due to CYP3A4 and was 20%-30% higher in young women than in young men. This difference applied to both parenteral and oral administration and was not explained away by p-glycoprotein activity. 76 For context, CYP3A4 activity was studied in relation to metabolism of some non-anticholinergic agents, such as midazolam to its active metabolites 77 and clindamycin. Meta-analysis suggested that women exhibited a 16% higher weight-corrected oral clearance of midazolam (p < .001) and 20% higher systemic clearance (p = .002) than men. No significant difference in the area under the curve (AUC) after oral dosing of midazolam was found but after IV administration women showed lower AUC than men (p = .02). No sexdependent differences were observed in midazolam bioavailability. 78 Clindamycin did not show any sex difference in its oral pharmacokinetics. 79 The study of midazolam and clindamycin confirmed sex variability in CYP3A4 metabolism, but failed to demonstrate any consistent sex-differences.
There was less study of sex differences in CYP2D6 and 2C19 Sex differences were demonstrated in the glucuronidation of some medications (acetaminophen) but not others (zidovudine), [84][85][86] suggesting that sex differences in drug conjugation exist and are drugdependent. To date, no anticholinergic agents have been explored with respect to glucuronidation. Clearance of some non-anticholinergic drugs by glucuronidation were shown to be increased in men in comparison with women including oxazepam, 67 temazepam, 87 and acetaminophen. 88 With regard to catechol-O-methyltransferase activity, liver tissue from female subjects exhibited approximately 25% lower activity than samples from male subjects. 89 There was a two-fold greater expression of hepatic p-glycoprotein in men compared with women 90 with unclear clinical relevance.

| Role of sex on the renal elimination of anticholinergic medications
Glomerular filtration is related to body mass. Males typically have a greater body weight than females, 62 so generally glomerular filtration is greater in males than females. This likely explains most sex-differences in renal drug clearance, though this was not observed for all drugs. Sex was found to be a significant factor in methotrexate clearance, with a 17% reduction in females after standardizing doses for body weight. 91 Some authors reasoned that for narrow therapeutic index drugs, the sex-related effect on kidney function may be clinically relevant. 91,92 Pharmacokinetic studies confirmed sex-differences in renal clearance for many drugs including the weakly anticholinergic drug digoxin, which had slower clearance in females 93 and the moderately anticholinergic drug amantadine, which had been shown to have significantly higher renal clearance in men due to putative sex differences in renal tubule secretion by organic cation transporters. 94 Sex differences in pharmacokinetics have been explored with respect to some anticholinergic medications. Results of studies that examined sex differences in anticholinergic drug pharmacokinetics as their primary objective are listed in Table 2.

| Summary of studies showing sex-differences in pharmacokinetics: Quinidine
The most commonly reported anticholinergic medication with a focus on sex-related differences was quinidine, exploring druginduced QT interval prolongation ( Table 2). [95][96][97] The findings of both Benton and Vicente 95,97 suggested that women cleared quinidine at a faster rate than men. Unexpectedly, women had a more rapid onset of ECG changes in response to drug activity than men, which was not entirely explained by increased quinidine clearance. These studies demonstrated sex-differences in quinidine pharmacokinetics; however, the mechanism of this difference was not clear. 95 123 Open label crossover study 7 men and 7 women of mean age 23 years and in good health To identify any pharmacokinetic differences between male and female volunteers in the metabolism of scopolamine when given with grapefruit juice

Methodology Results
Subjects received either 4 mg/kg of quinidine IV or a matching placebo solution over 20 min with 28 blood samples and simultaneous ECGs collected after drug/placebo infusion for each subject at predetermined time points over the following 12 h Quinidine causes QTc prolongation and T-wave morphology changes in both women and men Quinidine-induced maximum QTc (541 ± 40 ms vs. 510 ± 38 ms; p = .07) or maximum T peak -T end (216 ± 60 ms vs. 222 ± 37 ms; p = .76) was similar for men and women There was a trend toward a lower maximum serum quinidine concentration in women compared with men (2.9 ± 0.7 μg/ml vs. 3.7 ± 1.2 μg/ml; p = .07) The slope describing serum quinidine concentration versus QTc prolongation was greater in women than in men (38 ± 10 ms/μg/ml vs. 28 ± 9 ms/μg/ml; p = .02) Differences between women and men occurred primarily in the first 20 min after quinidine infusion, when serum quinidine concentrations were higher in men than women Subjects received either 4 mg/kg of quinidine IV or a matching placebo solution over 20 min. 28 blood samples and simultaneous ECGs were collected after drug/placebo infusion for each subject at predetermined time points over the following 48 h There was a trend to greater weight-adjusted clearance of quinidine in women than in men (5.2 ± 1.1 ml/min/kg vs. 4.3 ± 1.6 ml/min/kg) There was also a trend to a higher maximal plasma concentration of quinidine in men than in women (3.67 ± 0.13 μg/ml vs. 2.78 ± 0.87 μg/ml; p = .07) There were no sex-related differences in the ratio of the AUC ∞ of 3-hydroxyquinidine to the AUC ∞ of quinidine The estimated volume of distribution (V d ) at steady state was not different between the men and women There was no difference in the free fraction of quinidine in serum between men and women The free fraction of 3-hydroxyquinidine was slightly higher in women than in men (0.53 ± 0.05 μg/ml vs. 0.47 ± 0.05 μg/ml; p < .01) There were no statistically significant differences between males and females for any of the pharmacokinetic parameters-AUC (0-8 h) and C Max after the last dose were marginally significantly different between sexes 3. The population-by-sex effect was marginally significant for AUC (0-8 h) (p = .056) but not for C Max All subjects took quinidine sulfate capsules 3 mg/kg orally then ECGs and blood samples for quinidine concentrations were taken over 24 h following drug administration There were no significant differences in quinidine concentrations between men and women or in any of the pharmacokinetic variables measured The QT a , and QT c intervals were larger in females than in males Quinidine did not affect QRS duration in women but reduced QRS duration in men Participants were given doxylamine-pyridoxine 20-20 mg delayed-release tablets with 240 ml water on an empty stomach with blood sampling starting 1 h pre-dose with samples analyzed using high performance liquid chromatography-tandem mass spectrometry Females had significantly larger AUC 0-t for doxylamine compared with males A higher C Max for doxylamine was observed in females compared with males Subjects received either 8 mg of fesoterodine extended release or placebo with blood samples drawn over 36 h after drug administration and saliva samples on cotton wool collected over 24 h after drug administration No apparent differences in C Max , AUC 0-∞ , t max , or mean residual time between males and females Total plasma clearance was highest in young men and lowest in older women Elderly women experienced a 1 g decrease in salivary volume and elderly men did not 5 h after dose Elderly men experienced the greatest residual urinary volume increase 8 h after dose Each subject received at random scopolamine 0.5 mg IV, scopolamine 0.5 mg orally, or scopolamine 0.5 mg orally mixed with 150 ml fresh grapefruit juice and blood sampling occurred over the 24 h following drug administration C Max was significantly higher in males than females (6.61 ng/ml vs. 3.93 ng/ml) after IV infusion All other parameters were similar (Continues) excretion. A study of cyclobenzaprine examined sex-differences using a series of open-label, three-period, randomized, crossover studies. The first study included 24 healthy young subjects (mean age: 25.5 years), the second 18 healthy subjects (mean age: 28.7 years), and the third 12 older subjects (mean age: 71.3 years).
The primary objective was to investigate the bioavailability and pharmacokinetics of cyclobenzaprine with attention to the effects of sex, age, and hepatic insufficiency ( Table 2). There were small significant differences in the AUC and C Max for cyclobenzaprine between sexes in the older group. 98 This is most likely due to accumulation of drug in the group of older females. A study of the benzodiazepine diazepam demonstrated a shorter t 1/2 and a greater plasma clearance in men in comparison with women ( Table 2). 99 In a population of men and women receiving olanzapine for Alzheimer's disease or schizophrenia, between one and six samples were analyzed from each individual to determine sex-differences in olanzapine clearance. Sex was found to be responsible for 12% of variability in olanzapine elimination. Men cleared olanzapine 38% faster than women. 100 A natural pharmacokinetic study of anticholinergic antidepressants in older adults looked for sex-differences in serum concentrations. The ratio of absolute serum concentration in comparison with the dose-adjusted serum concentration was 1.1-to 1.5-fold higher in women than in men for clomipramine and trimipramine. This was despite a dose reduction in females who received 10%-30% lower dose but still achieved serum levels equivalent to male participants. 101 105 This may relate to a saturable metabolizing enzyme that was in a greater concentration or more active in men. Sex was correlated to paroxetine plasma concentrations in three studies that examined the effect of sex on paroxetine pharmacokinetics. In a study of 171 subjects aged ≥70 years, men had a higher paroxetine V d (461 ± 260 L) compared with women (346 ± 256 L). 106 In a study of 1677 older men and women, the serum concentration of paroxetine was 32% higher in women (86 nmol/L vs. 65 nmol/L, p < .001). 104 In a third study of 70 patients, the plasma concentration of paroxetine was higher in women across age groups (28 ng/ml vs. 16 ng/ml; p = .001). 107 The mean AUC and C Max for bupropion, a mildly anticholinergic antidepressant, were higher in women than men; however, once these parameters were standardized for body weight the statistical significance was lost. 108 For bupropion, older women had a larger V d and longer t 1/2 than young men. This does make it challenging to know how much of the effect was attributable to sex versus age. 109 Amitriptyline plasma levels were higher in women in a study of 110 inpatients receiving routine doses of amitriptyline, 110 but no significant sex-difference in serum concentration of amitriptyline was noted in the study by Reis et al. 104 Nortriptyline plasma levels were affected by sex with females experiencing higher plasma levels. 111 Desipramine was shown to have a longer elimination t 1/2 and a faster oral clearance in older men than in older women. 112 When examining risperidone plasma concentrations, the only parameter to exhibit a statistically significant difference between males and females was the plasma concentration/dose ratio. When weight was used to adjust the plasma concentration, any difference was lost. 113 Many of these psychoactive medications are metabolized by CYP2D6, and a sex-related difference in CYP2D6 activity has not consistently been identified in the literature, 114 which means there are likely other sex-dependent mechanisms contributing to these pharmacokinetics differences. In summary, while many sex-differences exist in the pharmacokinetics of psychoactive anticholinergic medications, no consistent patterns were identified. The small increases in drug exposure that were identified (most often by women) may help explain the increased experience of adverse events by women. 115 (Table 2). Total plasma clearance of fesoterodine was highest in young men and lowest in older women, but there were no apparent sex differences in C Max , AUC 0-∞ , or t Max . Interestingly, 5 h after the dose was given, older women experienced a 1 g decrease in salivary volume whereas older men did not, which provided some evidence that women were more likely to experience adverse effects (e.g. dry mouth) from this anticholinergic medication use. There was no clinically meaningful difference in any of the pharmacokinetic parameters studied based on race (mean AUC 0-t z was 70.7 ng/ml × h in white and 64.1 ng/ ml × h in black men, and mean C Max was 6.1 ng/ml in white and 5.5 ng/ml in black men). 118 Similarly, in a study of 337 individuals, darifenacin clearance was about 30% lower in females. 119

| Summary of studies showing sex-differences in pharmacokinetics: Scopolamine
An open-label crossover study of seven men and seven women of mean age 23 years and in good health was completed to identify any sex differences in pharmacokinetics in the metabolism of 0.5 mg scopolamine when given IV or orally with or without grapefruit juice. The C Max was significantly higher in males than females  for metabolism prior to elimination. 125 Drug conjugation was shown in several studies to remain fairly constant with respect to age. 126 Undeniably, numerous factors such as genetics, medication use, and frailty 127,128 can influence glucuronidation and sulfonation, but in younger and older healthy people glucuronidation and sulfonation were not statistically significantly different. In aging rat models, liver sinusoidal endothelial cells undergo pseudocapillarization, 129,130 a process characterized by loss of sinusoidal fenestrations, thickening of the endothelium, perisinusoidal collagen deposition, and basal lamina formation. 131 This process suggested that drug passages through the liver were reduced in size which, in theory, could prevent large molecules, in particular protein therapeutics and extensively protein-bound drugs, from travelling through the liver and being cleared; this was shown for liposomal doxorubicin in aged rats compared to young rats. 132 The relevance of these changes to anticholinergic drug pharmacokinetics in humans remains to be determined.

| Role of age on the renal elimination of anticholinergic medications
Renal elimination declines with age by all renal routes (glomerular filtration, tubular secretion, and passive reabsorption). 133,134 Any anticholinergic agent that is renally eliminated or has renally eliminated active metabolites is likely to accumulate in older adults in comparison to younger adults.

TA B L E 3
Details of study population, study objectives, methodology, and results of trials identified to have a primary objective of exploring age-related differences in pharmacokinetic parameters for anticholinergic medications Renal clearance was 28% lower in older men and women than younger men 3.4.4 | Role of age on blood-brain barrier function In men the V d of (R)-[11C] verapamil, a known p-glycoprotein substrate, increased with age in several cortical brain regions, strongly suggesting a progressive decrease in blood brain-barrier pglycoprotein function with age. 135 This could affect drug introduction to the brain which may affect efficacy or toxicity depending upon the anticholinergic agent used.
Studies with a primary objective of identifying age-related differences in drug pharmacokinetics were listed in Table 3. There was a trend to higher serum concentrations in older female patients with the lower dosage of fluvoxamine, but this diminished when the dosage was doubled and suggested there was an interaction between age and sex on fluvoxamine pharmacokinetics. Older subjects taking oral paroxetine had higher plasma concentrations than younger subjects. 138 In a study that examined bupropion kinetics in older adults with depression (mean age 71.5 years), clearance was 80% of that seen in younger adults 109 and t 1/2 was 34 h in comparison to most sources which report 11-14 h. 109,139 Among females, there was no significant difference between young and older groups in any of the pharmacokinetic variables for triazolam. Among males, the t 1/2 of triazolam increased. Furthermore, when age was evaluated as a continuous variable, AUC for triazolam increased significantly with age (p = .02) and clearance decreased with age (p = .02).
Further examination of cyclobenzaprine pharmacokinetics showed increased t 1/2 in older versus younger adults. 98

| Summary of studies showing age-differences in pharmacokinetics: Bladder anticholinergics
The potently anticholinergic drug oxybutynin followed the trend of increasing peak plasma levels and bioavailability with increasing age and frailty. 140 This effect was so significant that study authors suggested halving the dose of oxybutynin for older adults to achieve the same plasma levels as younger adults. AUC and C Max were increased 20% and 16% respectively when an older population was given the same dose of oxybutynin as a younger population.
Moreover, solifenacin, a newer bladder anticholinergic, had a longer t 1/2 due to slower elimination and longer time to reach C Max in older adults. This could be explained by the slowed absorption of solifenacin in older adults which increased their exposure to solifenacin by about 1.2-fold. 120 In a study of 16 young men, 16 older men and 16 older women, receiving either 8 mg of fesoterodine extended release or matching placebo, the renal clearance of fesoterodine was 28% lower in older men and women than younger men 118 (Table 3).
This increased exposure to fesoterodine in older adults may predict increased exposure of tolterodine in older adults as well, as fesoterodine and tolterodine are related compounds, with both being metabolized to the same active ingredient.

| Summary of studies showing age-differences in pharmacokinetics: Scopolamine
Healthy adult subjects were given scopolamine hydrobromide 0.5 mg IV if they were under 65 years of age and 0.3 mg if older than 65 years. These subjects then received a battery of tests of cognitive function in addition to measurement of pharmacokinetic variables. Older age was associated with slowed clearance and increased exposure to scopolamine. Age-related increases in scopolamine exposure was likely the greatest contributor to the increased sensitivity to cognitive adverse effects in older adults. The study authors hypothesized that age-related changes in CYP3A4 activity or content may have been responsible for the increased scopolamine exposure in older adults. 141

| Genetics
In addition to age and sex, it is important that we understand how genetic variation in CYP activity could influence clinical effect or toxicity as drugs that are substrates for these enzymes are frequently used by older adults. Understanding the effect of these CYP2D6 variants on pharmacokinetics is important for predicting drug effect and adverse effect.
The effect of CYP2D6 phenotype on anticholinergic medication exposure had been investigated in older adults. CYP2D6 phenotypes had been well characterized with respect to codeine pharmacokinetics. Limited activation and effect of codeine occurred in CYP2D6 PMs, and increased metabolism and toxicity was reported in UMs. 142 Nortriptyline plasma levels were mostly correlated to CYP2D6 genotype and sex. 111 In nursing home patients exposed to anticholinergic drugs, the highest serum anticholinergic activity was found in groups of CYP2D6 PMs. 143 Analysis of risperidone metabolism in 70 healthy volunteers (of whom 82.9% were either IM or EM) revealed that polymorphisms of the CYP2D6 enzyme were much more responsible than sex for variation in risperidone metabolism. CYP2D6 phenotype explained 52% of interindividual variability in risperidone pharmacokinetics. The AUC of the active moiety was found to be 28% higher in CYP2D6 PM compared with IM, EM, and UM. No other genetic markers were found to significantly affect risperidone concentrations. 144 This genetic variation in the metabolism of risperidone was of such magnitude that it could alter results when conducting bioequivalence studies. 145 Differences in dose responses should be considered as clinically relevant for any person initiated on risperidone, further supporting using the lowest possible doses at all times.
The bioavailability of tolterodine was strictly related to the genetic polymorphism of CYP2D6 and it ranged from 10% to 74%. 146 Byeon et al. investigated the relationship between CYP2D6 phenotypes and tolterodine pharmacokinetics in 46 Korean subjects.
The single dose and multiple dose C Max and AUC 0-24 of tolterodine, respectively, were significantly higher in the PM groups than in the EMs. The ratio of clearance to bioavailability of tolterodine in the EMs was 5-to 18-fold higher than PM (variant dependent) in multiple dosing studies. 147

| CYP2C19 and CYP3A4
Genetic With increased risk of hospitalization, cognitive impairment, and mortality as risks from anticholinergic drug use, improved understanding of sex, age, and genomic testing of CYP isozymes may be indicated to reduce serious anticholinergic adverse events. Rigorous pharmacokinetic analysis is a much needed and important next step to allow us to understand how dosing recommendations can be modified to treat older men and women most safely and effectively.
Studies done in the past often examined age, sex, or CYP polymorphisms alone and future work needs to account for all these factors in combination so that we may better approach personalized medicine for optimal outcomes.

D I SCLOS U R E
There are no conflicts of interest to disclose for any author.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analyzed in this study.