1 BACKGROUND

When treated with the same dosage, women show higher drug plasma levels than men for numerous drugs (Reis et al., 2009; Weiss et al., 2005; Zucker & Prendergast, 2020). This is due to sex differences in height, body weight, fat proportion, kidney function, activity of cytochrome P (CYP) 450 enzymes and smoking habits (Aichhorn, Marksteiner et al., 2006; Aichhorn, Whitworth et al., 2006; Meibohm et al., 2002; Oliveira et al., 2007; Schwartz, 2003; Soldin & Mattison, 2009). Sex differences in pharmacokinetics with a significantly higher concentration-dose ratio in women have been demonstrated for a number of psychotropic substances, for example, mirtazapine (Reis et al., 2009; Unterecker et al., 2013), olanzapine (Castberg et al., 2017; Weiss et al., 2005), venlafaxine (Hansen et al., 2017; Reis et al., 2009; Sigurdsson et al., 2015; Unterecker et al., 2012), escitalopram (Reis et al., 2009; Scherf-Clavel et al., 2019) and duloxetine (Lobo et al., 2009). While various authors have reported sex differences in the pharmacodynamic properties of psychotropic drugs, these have only been poorly studied with regard to therapeutic application. The differences in sex-dependent drug tolerability appear to be mostly due to differences in pharmacokinetic properties. Female sex hormones in particular seem to have a considerable influence on the desired and undesired effects of psychotropic drugs (Kokras et al., 2019; Soldin & Mattison, 2009; Spoletini et al., 2012).

Though reported inconsistently in current literature (D’Incau et al., 2014; Greenblatt et al., 2019), women appear to report a higher rate of adverse drug reactions (ADRs; Greil et al., 2019; Zucker & Prendergast, 2020; Zopf et al., 2008), perhaps due to higher plasma concentrations (Zucker & Prendergast, 2020), while in general, elderly patients appear more likely to suffer from ADRs (Beijer & de Blaey, 2002). However, in regards to psychotropic drugs, a number of studies have shown similar incidence rates of severe ADRs in elderly as in young patients (Greil et al., 2019; Gray et al., 2019; Singh et al., 2017). On the other hand, an increase of concentration-dose ratios with age is described for various psychotropic drugs, for example, mirtazapine (Reis et al., 2009; Unterecker et al., 2013), venlafaxine (Hansen et al., 2017; Reis et al., 2009; Sigurdsson et al., 2015; Unterecker et al., 2012), quetiapine (Aichhorn et al., 2006; Bakken et al., 2011; Castberg et al., 2017), citalopram (De Mendonça Lima et al., 2005; Reis et al., 2009; Unterecker et al. 2013), escitalopram (Reis et al., 2009), olanzapine (Castberg et al., 2017; Weiss et al., 2005), zopiclone (Gaillot et al., 1987) and sertraline (Reis et al., 2009; Unterecker et al., 2013). Since age-related changes in body composition and polypharmacy lead to different pharmacokinetics and pharmacodynamics, lower doses of psychotropic drugs in the elderly are often recommended (Hiemke et al., 2018; Katzman et al., 2014; Thürmann, 2020; Tveit et al., 2020).

In the past, women have been under-represented in clinical trials (Liu & Mager, 2016), therefore leading to a weaker understanding of sex differences in pharmacological treatment. According to Sørup et al. (2020), the influence of patient's sex on ADRs is an understudied factor. This may be especially problematic in the treatment of psychiatric conditions because women are more likely to be treated with psychotropic drugs than men (Glaeske et al., 2012).

In 2013, the U.S. Food and Drug Administration (FDA) approved label changes specifying new maximum dosing recommendations for zolpidem, being 5 mg/day for females and 10 mg/day for males (U.S. Food and Drug Administration [FDA], 2013a, 2013b). According to the FDA, higher dosages may lead to next-morning impairments, especially driving impairment (U.S. FDA, 2013a, 2013b). However, this recommendation is controversial. It has been argued that disturbed sleep in under-dosed females treated with zolpidem may impair driving even more than the intake of 10 mg/day (Greenblatt et al., 2019).

In this study, a large data set from a pharmacovigilance project is used to examine the dosage of psychotropic drugs in the treatment of major depressive disorder (MDD) in relation to sex and age in clinical routine. We selected the 10 most prescribed drugs in our dataset and additionally zolpidem due to the sex-specific recommendation made by the FDA stated above.

2 METHOD 2.1 Data source

The analysed prescription data of the present study was gathered by the project “Arzneimittelsicherheit in der Psychiatrie” (AMSP; drug safety in psychiatry). Established in 1993, AMSP is an ongoing European multi-center drug safety program which collects data on psychopharmacotherapy and severe ADRs occurring in psychiatric inpatients in a naturalistic setting. AMSP´s pharmacovigilance methods have been described in detail previously (Engel et al., 2004; Grohmann et al., 2004, 2014). In brief, AMSP consists of two principal data collections (i.e., prescription data and severe ADRs) from a total of 116 hospitals in Germany, Switzerland and Austria, as well as temporarily from Belgium and Hungary. The number of participating hospitals increased from nine in 1994 to 63 in 2015. In a cross-sectional approach, all participating hospitals record drug prescriptions for all inpatients under surveillance on two reference days per year. All drugs administered on these days are assessed along with the patients' age, sex and psychiatric diagnosis. Evaluations of the AMSP database have been approved by the Ethics Committee of the University of Munich and the Ethics Committee of the Hannover Medical School (Nr. 8100 BO S 2018). This study adheres to the Declaration of Helsinki and its later amendments. The AMSP program is a continuous observational post-marketing drug surveillance program and does not interfere with the ongoing clinical treatment of patients under surveillance.

2.2 Sample description

A total of 32,082 patients aged 18 to 89 years hospitalized between 2001 and 2015, with a primary diagnosis of MDD are investigated. The dosages of the 10 most prescribed drugs and zolpidem are analysed (descending in the frequency of prescription): mirtazapine (n = 10,431), venlafaxine (n = 8072), lorazepam (n = 7757), quetiapine (n = 6993), citalopram (n = 3909), escitalopram (n = 3842), olanzapine (n = 3441), duloxetine (n = 3070), zopiclone (n = 2936), sertraline (n = 2909) and zolpidem (n = 1873). Tricyclic antidepressants were not among the 10 most frequently prescribed drugs during the period studied. The frequency distribution according to sex and age differed only slightly; among patients over 65 years, two additional antipsychotic drugs were among the most frequently prescribed drugs: risperidone and pipamperone (see: Table S4). The sample consists of 11,887 (37.1%) males and 20,195 (62.9%) females. Note that a patient may have had more than one psychotropic drug prescribed. Table 1 shows a summary of the sample composition.

TABLE 1. Sample composition Sample composition Total number of patients 32,082 Number of males 11,887 (37.1%) Number of females 20,195 (62.9%) Age of males in years (mean ± SD) 49.5 ± 15.3 Age of females in years (mean ± SD) 51.9 ± 16.5 Number of psychotropic drugs prescribed per patient (mean ± SD) 2.6 ± 1.3 Abbreviation: SD, standard deviation.

In order to identify potential confounding variables that could account for distortions in the analysis, the following variables were considered: severity of MDD, psychiatric comorbidity, differences in calendar year and differences in age distribution.

In Figure 1, the number of male and female patients per calendar year are depicted. Substantial discrepancies in the male to female ratio over time were not observed.

Number of patients with MDD by calendar year. Number of male and female patients per calendar year between 2001 and 2015. Blue: Males. Orange: Females. Red: Male to female ratio. MDD, major depressive disorder

No substantial differences in the severity of MDD among male and female patients were found (Table 2). Table 3 shows the severity of MDD for young and elderly patients. Elderly patients were more likely to suffer from severe MDD with psychosis than younger patients (18.0% vs. 12.2%). Psychiatric comorbidity documented in 20.5% of patients did not differ essentially between women and men (see additional files for data). Further analysis is based on the assumption that differences in severity of MDD and small differences in male to female ratio over years did not lead to a significant distortion. Distribution of age (depicted in Figure 2), on the other hand, differs substantially. In particular, there were proportionally more elderly women than elderly men. Therefore, analysis was adjusted in order to consider differences in distribution of age.

TABLE 2. Severity of MDD according to sex Severity of MDD according to sex Mild MDD (“F32.0”,“F33.0”) 368 (1.2%) Male 151 (1.3%) Female 217 (1.1%) Moderate MDD (“F32.1”,“F33.1”) 8237 (25.7%) Male 3033 (25.5%) Female 5204 (25.8%) Severe MDD (“F32.2”, “F33.2”) 18,257 (56.9%) Male 6720 (56.5%) Female 11,537 (57.1%) Severe MDD with psychosis (“F32.3”, “F33.3”) 4308 (13.4%) Male 1645 (13.8%) Female 2663 (13.2%) No info or other MDD (“F32.8”,“F32.9”,“F33.4”,“F33.8,“F33.9”) 912 (2.8%) Male 338 (2.8%) Female 574 (2.8%) Note: Diagnoses according to 10th revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10). Severity of MDD is similar between male and female patients. Abbreviation: MDD, major depressive disorder. TABLE 3. Severity of MDD according to age Severity of MDD according to age Mild MDD (“F32.0”,“F33.0”) 368 (1.1%) Up to 65 ys 287 (1.3%) Over 65 ys 81 (1.2%) Moderate MDD (“F32.1”,“F33.1”) 8237 (25.7%) Up to 65 ys 6803 (26.8%) Over 65 ys 1434 (21.4%) Severe MDD (“F32.2”, “F33.2”) 18,257 (56.9%) Up to 65 ys 14,501 (57.1%) Over 65 ys 3756 (56.1%) Severe MDD with psychosis (“F32.3”, “F33.3”) 4308 (13.4%) Up to 65 ys 3104 (12.2%) Over 65 ys 1204 (18.0%) No info or other MDD (“F33.4”,“F32.9”,“F32.8”,“F33.9”,“F33.8”) 912 (2.8%) Up to 65 ys 693 (2.7%) Over 65 ys 219 (3.3%) Note: Diagnoses according to ICD-10. Overall, severity of MDD is similar between young and elderly patients (despite differences for moderate MDD and severe MDD with psychosis). Abbreviations: MDD, major depressive disorder; ys, years of age. 2.3 Statistical analysis

Statistical analysis was performed using R version 3.6.2. To assess the influence of age in dose of psychotropic drugs, the slope of a linear fit among patients older than 65 and younger than 90 years of age was computed. Significance was estimated using a t-test on the coefficient of the slope under the hypothesis to be zero. Along with the slope, the percentage of change in relation to the previous year was computed, which was calculated from the coefficient of an exponential fit. In analogy to literature on differences in age-related pharmacokinetics, the age groups of up to 65 years old and over 65 years old were selected. The elderly group served as reference. p-values were computed using a Wilcoxon test.

In order to account for sex differences in dosage, we adjusted for age by comparing prescriptions in patients with similar age. Patients were first categorized by sex and then assigned to different age groups. Mean dosage of each group for both sexes was calculated separately. The sex difference within each age group was calculated with female patients serving as reference. Finally, the total sex difference was calculated as the weighted mean of all percentage differences, by weighting with the number of patients (male and female) of each age group. Significance was determined using a paired Wilcoxon signed-rank test.

3 RESULTS 3.1 Age

Dosage in relation to age for antidepressant drugs, antipsychotic drugs and hypnotic and tranquilizing drugs are shown in Figures 3–5-3–5. Table 4 shows the mean differences of dosages in patients up to 65 years and older than 65 years.

Daily dose according to age: Antidepressant drugs. Daily prescribed doses (mg) of antidepressant drugs according to age including both male and female patients. The height of the bars corresponds to the mean daily dosage for inpatients within age groups consisting of three consecutive years. In red, a linear model fitted to patients between 66 and 89 years old is shown. Slope: Thes lope of the linear model, describing the average absolute difference of daily dose in relation to the previous year in elderly patients. Change: Average percentage change in relation to the previous year in elderly patients. p-val: p-value under the hypothesis that the slope is zero, p < 0.001 for all drugs

Daily dose according to age: Antipsychotic drugs. Daily prescribed doses (mg) of antipsychotic drugs according toage including both male and female patients. The height of the bars corresponds to the mean daily dosage for inpatients within age groups consisting of three consecutive years. In red, a linear model fitted to patients between 66 and 89 years old. Slope: The slope of the linear model, describing the average absolute difference of daily dose in relation to the previous year in elderly patients. Change: Average percentage change in relation to the previous year in elderly patients. p-val.: p-value under the hypothesis that the slope is zero, p < 0.001 for both antipsychotic drugs

Daily dose according to age: Hypnotic and Tranquilizing drugs. Daily prescribed doses (mg) of hypnotic and tranquilizing drugs according to age including both male and female patients. The height of the bars corresponds to the mean daily dosage for inpatients with in age groups consisting of three consecutive years. In red, a linear model fitted to patients between 66 and 89 years old. Slope: The slope of the linear model, describing the average absolute difference of daily dose in relation to the previous year in elderly patients. Change: Average percentage change inrelation to the previous year in elderly patients. p-val.: p-value under the hypothesis that the slope iszero, n.s. (not significant) for zopiclone; p < 0.001 for lorazepam and zolpidem

TABLE 4. Mean dosage differences in patients up to 65 and older than 65 years Drug 18ys to 65 ys 66ys to 89 ys Difference [%](means) p-value (means) N Mean dose SE N Mean dose SE [mg/d] [mg/d] Citalopram 3130 29.2 0.2 779 27.3 0.5 +7.0 <0.001 Escitalopram 3143 16.3 0.1 699 14.4 0.3 +13.4 <0.001 Sertraline 2357 102.8 1.0 552 93.6 2.0 +9.7 <0.001 Mirtazapine 7419 33.2 0.2 3012 33.6 0.3 NS NS Venlafaxine 6587 200.5 1.1 1485 175.9 2.2 +14.0 <0.001 Duloxetine 2462 79.9 0.6 608 73.0 1.2 +9.4 <0.001 Olanzapine 2601 10.2 0.1 840 8.2 0.2 +24.4 <0.001 Quetiapine 5673 215.1 2.6 1320 177.8 4.4 +21.0 <0.001 Zolpidem 1379 10.5 0.1 494 9.5 0.1 +10.4 <0.001 Zopiclone 2162 7.2 0.1 774 7.2 0.1 NS NS Lorazepam 5624 1.9 0.0 2133 1.5 0.0 +29.7 <0.001 Note: Number of patients (N), prescribed mean dosages and standard error (SE) for antidepressant, antipsychotic and hypnotic drugs in inpatients up to 65 and over 65 years old. On the right hand the difference between means as percentage change is shown, in which the elderly patient group serves as reference. p-values are calculated using a Wilcoxon test. Abbreviation: ys: years of age. 3.1.1 Age: antidepressants drugs

Among the six analysed antidepressant drugs (Figure 3), a slight increase of daily dose for each additional year of age in young patients and a decrease in elderly patients was detected. Dosage remained relatively unchanged among patients between 35 and 65 years of age. Among patients older than 65 years, a decrease between 0.65% for the noradrenergic and specific serotonergic antidepressant drug (NaSSA) mirtazapine and 1.32% for the selective serotonin reuptake inhibitor (SSRI) sertraline for every increasing year of age was observed. Dose of the SSRIs citalopram and escitalopram decreased by 0.68% and 1.13%, respectively, and 1.00% and 0.76% for the selective serotonin-norepinephrine reuptake inhibitors (SSNRIs) venlafaxine and duloxetine. All of these findings are statistically significant (p < 0.01). When comparing mean prescribed dosages between patients up to 65 years old and older than 65 years old, statistically significant differences were found for all antidepressant drugs with the exception of mirtazapine, varying from 7% for citalopram to 14% for venlafaxine (Table 4). When the comparison of dosages starts at age 30 (i.e., comparison of age groups 30-65 and 66-89), the differences between the two age groups are only slightly greater and the lack of difference in dosage for mirtazapine remains unchanged. This also applies to the hypnotic drug zoplicone; see below.

3.1.2 Age: antipsychotics drugs

Dosage of the two antipsychotic drugs olanzapine and quetiapine shows a plateau among young and middle-aged patients (Figure 4). However, a more pronounced slope than in antidepressant drugs was observed in elderly patients. Among patients over 65 years old, a decrease of 2.48% (olanzapine) and 2.83% (quetiapine) in daily prescribed dosages for each increasing year of age (p < 0.0001) was found. When comparing the mean prescribed dosages between patients up to 65 years old and older than 65 years old, statistically significant differences of 24.4% for olanzapine and 21% for quetiapine were detected (Table 4).

3.1.3 Age: hypnotic and tranquilizing drugs

Distribution of prescribed daily dosage differs substantially between the two hypnotic drugs zopiclone and zolpidem and the tranquilizing drug lorazepam (Figure 5). Dosage of zolpidem showed a slight increase in young patients, a plateau in middle-aged patients and a decrease in elderly patients (Figure 5a). This distribution is similar to the distribution of antidepressant drugs. A consistent dosage of approximately 7.2 mg zopiclone per day was seen for all age groups (Figure 5b). Dose of lorazepam showed a steady decline among all age groups (Figure 5c).

Among patients over 65 years old, daily prescribed dose decreased for each increasing year of age for zolpidem by 1.17% and lorazepam by 1.26% (p < 0.001). Zopiclone did not show any significant age-related adjustments in dosage. When comparing the mean difference in prescribed dosages between patients up to 65 years old and older than 65 years old, statistically significant differences were found for zolpidem (10.4%) and lorazepam (29.7%) but not for zopiclone (Table 4).

3.2 Sex

The influence of sex on prescribed dosages of antidepressant, antipsychotic and hypnotic and tranquilizing drugs are shown in Figures 6–8-6–8. The analysis of the dosages of the 11 drugs studied showed statistically significant interactions between age and sex for escitalopram, quetiapine and olanzapine with more prominent sex differences for dosages in the elderly compared to younger adults (p < 0.001).

Daily dose according to age and sex: Antidepressant drugs. Daily prescribed doses (mg) of antidepressant drugs according toage: males in blue and females in orange. The standard error is depicted in red. The height of the bars corresponds to the mean daily dosage for inpatients within age groups consisting of six consecutive years. p: The p-value is computed from a (paired) Wilcoxon signed-rank test under the hypothesis that there are no differences between sexes. Diff: The percentage difference between sexes. Female patients serve as reference. Prescribed dosage differences between men and females were small. p < 0.05 for escitalopram and venlafaxine; n.s. for all other antidepressant drugs

Daily dose according to age and sex: Antipsychotic drugs. Daily prescribed doses (mg) of antipsychotic drugs according to age: males in blue and females in orange. The standard error isdepicted in red. The height of the bars corresponds to the mean daily dosage for inpatients within age groups consisting of six consecutive years. p: The p-value is computed from a (paired) Wilcoxon signed-rank test under the hypothesis that there are no differences between sexes. Diff: The percentage difference between sexes. Female patients serve as reference. The prescribed dosage differences between men and females are either small or statistically insignificant: p < 0.05 for olanzapine, n.s. for quetiapine

Daily dose according to age and sex: Hypnotic and Tranquilizing drugs. Daily prescribed doses (mg) of hypnotic and tranquilizing drugs by age: males in blue and female in orange. The standard error is depicted in red. The height of the bars corresponds to the mean daily dosage for inpatients within age groups consisting of six consecutive years. p: The p-value is computed from a (paired) Wilcoxon signed-rank test under the hypothesis that there are no differences between sexes. Diff: percentage difference between sexes. Female patients served as reference

3.2.1 Sex: antidepressants drugs

Statistically significant differences in daily dose prescriptions between sexes were detected for both escitalopram and venlafaxine of which males received 5.13% and 4.59%, respectively, more than females (Figure 6). Dosage of citalopram, sertraline, mirtazapine and duloxetine did not reveal statistically significant sex differences.

3.2.2 Sex: antipsychotic drugs

Among antipsychotic drugs, statistically significant differences were detected for olanzapine of which males received 4.94% more than females (Figure 7). Dosage of quetiapine was unrelated to sex.

3.2.3 Sex: hypnotic and tranquilizing drugs

Dosage of the hypnotic drugs zolpidem and zopiclone and the tranquilizing drug lorazepam did not show any significant association to sex (Figure 8).

4 DISCUSSION 4.1 Age

The findings of the present study are discussed in context of the currently available literature regarding the influence of age and sex on plasma levels of patients taking psychotropic drugs. A comprehensive table describing primary findings of the studies referred to can be found in the supplementary material (Table S1).

4.1.1 Age: antidepressant drugs

In 2009, Reis et al. (2009) analysed plasma levels for citalopram and escitalopram at steady state, finding that elderly patients over 64 years old had 84% and 91% higher plasma levels than younger patients at the same