Association between prescribed antidepressants and other prescribed drugs differ by gender: a nationwide register-based study in Sweden

Abstract

Background: People with depression are prescribed more drugs than people in general, partly due to comorbidity with other conditions. However, little research has been done on depression-related drug use from a gender perspective.

Aim: Examine the association between antidepressants, other types of prescribed drugs, and polypharmacy, by gender.

Methods: Data on drugs dispensed October to December 2016 to all Swedish citizens aged 18–84 years were collected from the Swedish prescribed drug register. Logistic regression analyses were performed to examine the associations between antidepressants and other drugs, by gender.

Results: For both men and women, associations were found between antidepressants and drugs for alimentary tract problems, respiratory problems, blood, nervous system, analgesics, and polypharmacy. However, for women, but not men, associations were also found for drugs for diabetes, musculoskeletal problems, dermatological problems, and systemic hormones.

Conclusions: Associations were found between antidepressants and many other types of drugs for both men and women; indicating comorbidity between depression and other conditions. Further, some of the associations between antidepressants and other drugs were found to be specific among women. Whether this indicates that men and women differ in comorbidity between depression and other conditions cannot be concluded based on this cross-sectional study. However, comorbidity impairs the possibility of recovery; in the somatic condition as well as the depression. Thus, physicians need to be aware that the association between antidepressants and other types of drugs are more common among women than men.

Keywords:

• Gender
• antidepressants
• prescribed drugs
• polypharmacy

Background and aim

Depression, and hence the use of antidepressants, is common in the general population. The point prevalence of diagnosed depression is estimated to 5–8%, more common in women than in men, and the prescription of antidepressants is even higher; in Sweden, the point prevalence is ∼9% [1–6]. However, even though depression is still the main indication for antidepressant use, it is important to consider, when using antidepressants as a proxy for depression, that antidepressants are also prescribed for other indications than depression [7]. In general, people with depression are also prescribed drugs other than antidepressants more often than are those without depression [8–10]. The reason for this can partly be explained by comorbidity between depression and other conditions, mental as well as physical [11,12]. For example, in chronic diseases like diabetes, heart disease, and asthma, the prevalence of depression is above average [13–17]. Moreover, there is a strong link between depression and medically inexplicable physical symptoms such as various pain conditions (e.g. fibromyalgia) and gastrointestinal symptoms (e.g. irritable bowel syndrome and non-ulcer dyspepsia), as well as poor sleep quality – problems that often lead to high health care utilization and consequently many prescribed drugs [18–20]. In addition, people with depression experience more severe symptoms and have more difficulty getting used to the symptoms of the physical disease, probably increasing drug use additionally [21–23]. Thus, physical disease is a risk factor for depression due to the burden of physical disease and, vice versa, depression seems to be a risk factor for physical disease. This may have several causes. For instance, indirect depression-related effects, such as unhealthy lifestyle and poor adherence to medical recommendations, are considered important [24,25]. Moreover, depression and physical diseases sometimes share a common pathogenic mechanism or common genetic or environmental risk factors [26]. Whether depression is primary or secondary to physical disease, or bidirectional, is difficult to determine. It is, however, likely that comorbidity impairs the possibility of recovery from the physical disease as well as from the depression [27].

In relation to some diseases, men and women seem to be mentally affected to different extents. Taking pain conditions as an example, depression has been found to be more common in women than in men [28]. However, the association between the use of antidepressants and the use of prescribed drugs for other conditions, has hardly been studied from a gender perspective in the general population. Moreover, although studies have consistently shown a higher use of health care services and drugs among women than among men [6,29], and among depressed than among non-depressed individuals [8,9], few studies have investigated the total drug usage among men and women using antidepressants. The use of multiple drugs (polypharmacy) is not uncomplicated, however, because it generates a greater risk for drug interactions and adverse drug reactions [30,31]. The aim of the present population-based study was to examine the association between prescribed antidepressants and other prescribed drugs, different types of drugs as well as polypharmacy, by gender.

Methods

The present cross-sectional study was based on all Swedish citizens aged 18–84 years who had been dispensed at least one prescribed drug during the period October to December 2016. The study period of 3 months was based on the Swedish regulation that drugs can be dispensed for use during a 90-day period at most. Data on drugs dispensed were collected from the Swedish Prescribed Drug Register (SPDR), a national database established in 2005, regulated by the Swedish government and maintained by the National Board of Health and Welfare [6]. The SPDR includes data on all dispensed prescriptions from the entire Swedish population (about 9.9 million inhabitants in 2016). It has been estimated that the register covers 99% of all prescriptions [32]. The SPDR contains, in addition to the dispensed drugs, for example, information on age, sex, and personal identification number (unique personal identifier given to all Swedish citizens/legal residents). All drugs in the SPDR are classified according to the anatomical therapeutic chemical (ATC) classification system [33]. In the present study, drugs were primarily divided into the anatomical main groups in the ATC system (Table 1). However, in some cases, subgroups were analyzed separately, i.e. in main group A, diabetes (A10), and in main group N, analgesics (N01-N02) and antidepressants (N06A). The ATC groups P (Antiparasitic products, foremost preventive and over-the-counter (OTC) drugs) and V (Various, foremost diagnostic aids and antidotes) were excluded from the analyses since not considered relevant in this study. Those who had been dispensed at least one antidepressant (N06A) during the 3-month period studied were classified as users of antidepressants. The same classification of users was also applied to all the other ATC groups. Analyses of polypharmacy (defined as ≥4 types of drugs other than antidepressants, i.e. at least 5 drugs including antidepressants) [34] also concerned ATC groups. The ATC groups were counted once, even if drugs from the group were dispensed more than once.

Table 1. Prevalence of prescribed antidepressants among men and women prescribed other types of drugs (ATC classification).

ATC classification	Men		Women
	Number*	Percentage proportions antidepressants	Number*	Percentage proportions antidepressants
A01-A09, A11-A16 Alimentary tract & metabolism	370,288	16.0	571,000	22.6
A10 Diabetes	202,191	10.1	137,865	18.2
B01-B06 Blood & blood forming organs	436,186	11.6	414,217	20.6
C01-C10 Cardiovascular system	812,718	9.7	775,499	16.7
D01-D10 Dermatologicals	171,916	11.3	218,242	17.7
G01-G04 Genito urinary systems & sex hormones	205,526	10.3	556,816	12.6
H01-H05 Systemic hormonal preparations	118,279	11.6	341,049	17.5
J01-J07 Anti-infectives for systemic use	198,677	8.8	343,245	14.8
L01-L04 Antineoplastic & immunomodulating agents	58,455	9.0	78,471	14.8
M01-M09 Musculo-skeletal system	217,227	10.5	290,797	19.2
N01-N02 Analgesics	276,973	16.1	465,350	23.8
N03-N05 + N07 Nervous system	306,577	30.7	465,438	36.9
R01-R07 Respiratory system	296,106	12.9	437,373	20.1
S01-S03 Sensory organs	123,560	9.2	173,574	15.7
Polypharmacy^$	322,393	17.6	475,133	26.2

Chi-square tests comparing men and women; p < .001 (all ATC groups and polypharmacy).

* Number of individuals in study population prescribed at least one drug from respective ATC groups.

^$ Defined as ≥4 types of drugs other than antidepressants.

Statistical analyses

The study population was divided into men and women and users and non-users of different types of prescribed drugs (ATC groups). Chi-square tests (significance level p = .05) were performed to examine gender differences in the prevalence of antidepressants in the study population, as well as gender differences in the prevalence of antidepressants among those prescribed other types of drugs (different ATC groups as well as polypharmacy, Table 1). Logistic regression analyses (OR; 95% CI) were performed to examine the association between antidepressants and other types of prescribed drugs (different ATC groups as well as polypharmacy, Table 2) among men and women, adjusted for age as a potential confounder. The IBM SPSS Statistics for Windows (Version 22.0. Armonk, NY: IBM Corp.) was used to perform the analyses.

Table 2. Association between prescribed antidepressants and other types of prescribed drugs (ATC classification) among men and women. Logistic regression (antidepressants vs. not antidepressants) adjusted for age.

ATC classification	Men	Women
	OR (CI)	OR (CI)
A01-A09, A11-A16 Alimentary tract & metabolism	1.67 (1.65–1.69)	1.69 (1.67–1.70)
A10 Diabetes	0.88 (0.87–0.89)	1.10 (1.08–1.11)
B01-B06 Blood & blood forming organs	1.26 (1.24–1.28)	1.42 (1.40–1.43)
C01-C10 Cardiovascular system	0.77 (0.76–0.78)	0.99 (0.98–1.00)
D01-D10 Dermatologicals	0.87 (0.86–0.88)	1.03 (1.02–1.04)
G01-G04 Genito-urinary systems & sex hormones	0.95 (0.94–0.97)	0.60 (0.59–0.60)
H01-H05 Systemic hormonal preparations	1.00 (0.98–1.02)	1.03 (1.02–1.04)
J01-J07 Anti-infectives for systemic use	0.60 (0.59–0.61)	0.80 (0.79–0.80)
L01-L04 Antineoplastic & immunomodulating agents	0.75 (0.73–0.77)	0.83 (0.81–0.85)
M01-M09 Musculoskeletal system	0.80 (0.79–0.82)	1.16 (1.15–1.17)
N01-N02 Analgesics	1.55 (1.53–1.56)	1.74 (1.72–1.75)
N03-N05 + N07 Nervous system	5.06 (5.01–5.11)	4.47 (4.44–4.51)
R01-R07 Respiratory system	1.02 (1.01–1.03)	1.25 (1.24–1.26)
S01-S03 Sensory organs	0.77 (0.76–0.79)	0.90 (0.88–0.91)
Polypharmacy*	2.44 (2.42–2.47)	2.34 (2.32–2.36)

OR: Odds Ratio; CI: 95% Confidence Interval.

*Defined as ≥4 types of drugs other than antidepressants.

Ethical considerations

The study fulfills research ethics requirements and was approved by the Regional Ethical Review Board in Uppsala, Sweden (Reg. no. 2016/235). Data were anonymized before the researchers were given access to it.

Results

The study population included 3,884,468 individuals (42.6% men and 57.4% women) and was relatively evenly distributed in terms of age (30.2% were 18–44 years of age, 34.1% 45–64 years, and 35.7% 65–84 years on 1 October 2016). In total, 15.3% of the study population was dispensed at least one antidepressant during the 3-month period under study, more women (17.4%) than men (12.4%) (p < .001). In Table 1, the prevalence of antidepressants among men and women prescribed drugs from different ATC groups can be found. As can be seen, women have consistently prescribed antidepressants more often than men were in all ATC groups (p < .001).

In Table 2, the results from the logistic regression analyses can be found. Men and women showed somewhat different patterns regarding associations between antidepressants and other types of prescribed drugs. For both men and women, there were associations between antidepressants and drugs for alimentary tract problems, drugs for the blood, analgesics, drugs for the nervous system, drugs for respiratory problems, and polypharmacy. However, for women, but not men, there were associations between antidepressants and drugs for diabetes, drugs for dermatological problems, systemic hormones, and drugs for musculoskeletal problems.

Discussion

Little research has been conducted on the relationship between depression and use of drugs for other conditions in the general population from a gender perspective. Therefore, the present large population-based study examining the association between antidepressants and other types of drugs, by gender, can contribute knowledge that is relevant to clinical practice and provides a basis for future studies on the topic. The SPDR has the advantage of offering complete data on all dispensed prescriptions and has contributed considerably to new knowledge on drug therapy [32]. Also, using drug prescription databases has advantages like that the data are free of information bias and it is a relatively low-cost method of identifying patient groups in the population.

Limitations

It is important to emphasize that the cross-sectional design employed does not allow conclusions about causality to be drawn. Moreover, although the SPDR has the advantage of offering complete data on all dispensed drugs, it does not contain indications. Antidepressants, for example, increasingly have indications in addition to depression, e.g. neuropathic pain, anxiety disorders, and eating disorders, even though depression is still the main indication [35,36]. Dispensed drugs are often used as a proxy for disease [37], but because the indications for drug treatment are not fully known, the assumptions about conditions must be restricted to hypothetical considerations. Further, drug prescribing is influenced by several non-medical factors, such as physicians’ prescribing patterns, type of practice, and patient load. Also, the depression will sometimes go undetected and, as a result, antidepressants might be under-used in the population. Furthermore, depression is sometimes treated with, for example, psychotherapy and not antidepressants [38,39]. Therefore, the drug use measured by the SPDR, i.e. prescribed and dispensed drugs, may not reflect actual morbidity. Further, the study period of 3 months may be a too short period to capture concomitant use. Above that, the register does not cover for example OTC-drugs and herbal remedies, commonly used in the population.

Prevalence of antidepressants

The prevalence of antidepressants, in the total study population as well as in all the different ATC groups studied, was higher among women than among men during the 3-month period studied. The overall greater use of antidepressants among women compared to men is well-known [6,40]. However, whether this indicates that women are more depressed than men is hard to confirm. Given the more frequent health care contacts among women compared to men, and given that the diagnostic criteria are based on symptoms reported by women, women are more often diagnosed as having depression and thus more often prescribed antidepressants [41–43]. Additionally, studies have shown that women are more likely than men are to receive a prescription when visiting healthcare services [44], which is important to bear in mind when interpreting our results. Moreover, factors other than gender affect diseases, health care utilization and drug use, factors such as educational level and socioeconomic status, which were not included in our study.

Polypharmacy

It has been found that people with depression, due to comorbidity with other conditions and a great use of health care, are prescribed and use more drugs overall than do people without depression [8,9,45]. In the present study, we found polypharmacy to be highly associated with antidepressant use for both men and women. Medical comorbidity is probably an important explanation for this association [11,12]. Polypharmacy is often used as an indicator of high morbidity, and the number of drugs may be a valuable proxy for burden of disease [37]. It is difficult to determine, however, whether depression causes conditions treated with various drugs or whether depression is a consequence of other conditions. The association between polypharmacy and antidepressants could also be explained by the fact that polypharmacy increases the risk of adverse drug reactions that may be the cause of the depression. Polypharmacy can also be a health risk due to the increased risk of prescribing cascades, drug interactions, and non-adherence [46–49]. That is, when prescribing and evaluating drug treatments, physicians need to be aware of the association between depression and polypharmacy. Worth noting is the fact that, in the present study, polypharmacy was based on ATC groups and not specific drugs classes. Thus, for many individuals in the study population, the number of drugs was probably considerably larger than the ATC groups assessed. In addition, the total drug use is certainly underestimated, as many people use medications other than prescribed drugs, such as OTC drugs, herbal remedies, drugs used in hospitals, and those purchase on the Internet. On the other hand, for a variety of reasons, a certain percentage of all dispensed drugs will never be used (i.e. non-adherence), resulting in an overestimation of drug use [50].

Antidepressants and other types of drugs

In the present study, we found associations between antidepressants and many other types of prescribed drugs, and it was more common among women than among men. For example, for both men and women, an association was found between antidepressants and drugs for alimentary tract problems. It is known that many gastrointestinal problems are related to depression [18,19]. Also, in depression it is common for physical symptoms to derive from the gastrointestinal system [13]. Many of the drugs prescribed for gastrointestinal problems can be bought OTC and a lot of people probably use those drugs without prescription why the association might be even stronger than found in this present study. Furthermore, this ATC group also includes, in addition to drugs for gastrointestinal problems, anti-obesity drugs. Some previous studies have found women at increased risk for depression related to obesity, which may hypothetically be associated with the elevated importance of weight in the identities of women compared to men [51,52].

In relation to different pain conditions, particularly chronic pain, there is a well-known association with depression [15,53,54]. Whether pain causes depression or depression causes pain is, as for other conditions, often unclear, and both explanations are likely. We found a strong association between antidepressants and prescribed analgesics among both men and women. However, we also found an association for women, but not men, concerning drugs for musculoskeletal problems (including anti-inflammatory drugs related to pain conditions) and systemic hormones (including corticosteroids, often used in pain conditions). Other studies have found that women seem to be more mentally affected by pain conditions than men are [55]. Explanations for the strong association between depression and pain in women have been discussed in terms of gender differences in morbidity. Women are more often diagnosed with diseases related to pain, for example, rheumatic diseases and migraine [55]. These conditions are often severe and chronic and, therefore, related to depression. Further, women seem to have a greater sensitization to pain compared to men, and differences in work, lifestyle, and stressors may also contribute to the gender differences [56,57]. Moreover, because women seek health care more often than men do [29], they may also be more likely to have OTC analgesics prescribed for them.

Previous research has shown an association between depression and diabetes type 1 as well as type 2 [11,58]. However, few studies have focused on gender. One previous study found depression in diabetes to be associated with female gender [59], which corresponds with our study, where an association was found between antidepressants and drugs for diabetes in women, but not in men. Women, more than men, seem to have many concerns related to diabetes self-management in daily life, which may cause depression [59]. Depression decreases adherence to treatment regimens, and diabetes adherence is essential to complication prevention and mortality reduction [60,61]. Thus, physicians need to pay attention to depression in relation to diabetes, especially in women. In previous studies, asthma and other obstructive respiratory diseases have also been found to be related to depression [62,63]. Just as for many of the other types of drugs studied, we found an association between antidepressants and drugs for respiratory conditions. The reason for this association may be related to the symptoms of the disease or to the treatment [64]. In previous studies patients, especially women, reported that their asthma interfered with social activities, sleep, and life in general [65]. Just as for diabetes, self-management and adherence are essential to treatment outcomes in obstructive respiratory diseases [66]. Therefore, an awareness of depression in relation to respiratory diseases is of great importance.

For men as well as women, the association between antidepressants and other drugs for the nervous system was pronounced. Comorbidity between mental health problems is well-known and a cause of more severe symptoms [67]. The conditions treated with drug classes included in this ATC group, such as anxiolytics, hypnotics, antipsychotics, antiepileptics, and drugs for Parkinson’s disease, have all been found to be related to depression [67–70]. In other studies, comorbidity between depression and other mental health problems, such as anxiety disorders and sleeping disorders, has repeatedly been found to be more frequent in women [70,71]. However, in our study we did not distinguish the different drug classes and further studies analyzing the association between antidepressants and the specific drug classes in this ATC group could reveal some gender differences that we are unaware of today.

Dermatological problems seriously affect self-image, self-esteem, and the ability to form relationships with others, and in several studies, they have been found to be related to depression [72]. Although often seen as insignificant in comparison to other diseases, the effect of dermatological problems on mental health is comparable to that of pain, diabetes, and asthma [73]. In the present study, an association between dermatological and antidepressants was found for women but not for men. Other studies have also found women to be more mentally affected by dermatological diseases than men are [74]. The gender difference would seem to be related to appearance, as symptoms from visible body parts, such as the face and hands, were more often related to mental health problems among women than among men [75]. Women with dermatological diseases have also been found to experience great complications in their relations with others [76]. Because dermatological problems greatly interfere with women’s daily life, the association with depression is important to take seriously.

Conclusions

This large population-based study found that antidepressants were often associated with other types of prescribed drugs, as well as polypharmacy among both men and women. Some of these associations were found to be specific among women, for example, drugs related to pain conditions, dermatological conditions, and diabetes. Whether this indicates that men and women differ in comorbidity between depression and other conditions cannot be determined on the basis of the present cross-sectional study. However, the study does contribute knowledge that is important for clinical practice and that can serve as a basis for future studies on the topic aimed at establishing a causal relationship. Comorbidity between depression and other conditions impairs the possibility of recovery, and decrease adherence to medical recommendations, and polypharmacy are associated with adverse drug reactions and other risks that may cause depression. Therefore, physicians should be aware that comorbidity might differ by gender.

Acknowledgements

We thank Hans Högberg, University of Gävle, for assistance with the statistical tests.

Disclosure statement

No potential conflict of interest was reported by the authors.