Gender and medication use in Turkey: Evidence from a general population survey

Tekin Kose

doi:10.1371/journal.pone.0321590

Abstract

Gender differences in health behaviors and outcomes were commonly documented by researchers. The focus of this study was the analysis of gender differences in medication use for a general population in Turkey. It also explored a range of factors associated with medication use at the individual level. A nationally representative cross-sectional data set was obtained from the 2019 wave of the Turkish Health Survey. The sample of this study included 17,083 adults residing in different regions of Turkey. Conditional mixed-process regression models were estimated for the whole sample and subsamples by gender. The rates of prescribed and non-prescribed medication use were 40.7% and 30.2%, respectively, in the adult population of Turkey. There were significant gender differences in medication use in the Turkish case. Females were 19.4% more likely to use prescribed medication, and they were 30.8% more likely to use non-prescribed medication compared to males in Turkey. There were negative associations between prescribed and non-prescribed medication use. On average, females were 9.2% less likely to report higher levels of health status, and they were 18.4% more likely to use healthcare services. Individuals with higher levels of self-rated health status were less likely to use prescribed medication. Both prescribed and non-prescribed medication use were positively related to healthcare service use. Complementing the earlier literature, the results of the present study demonstrated that gender-specific designs should be considered by health policies on the use of medications.

Citation: Kose T (2025) Gender and medication use in Turkey: Evidence from a general population survey. PLoS ONE 20(4): e0321590. https://doi.org/10.1371/journal.pone.0321590

Editor: Laura Hannah Kelly, Public Library of Science, UNITED KINGDOM OF GREAT BRITAIN AND NORTHERN IRELAND

Received: June 16, 2023; Accepted: March 7, 2025; Published: April 9, 2025

Copyright: © 2025 Tekin Kose. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The author used the Turkey Health Survey (THS) for 2019, a secondary data set collected by TSI. The data set included information on demographics, health information, health behaviors, and health-related outcomes for children and adult participants. Participants of the survey were selected using randomized sampling methods from various regions of Turkey. The data that support the findings of this study are available from the Turkish Statistical Institute (TSI) but restrictions apply to the availability of these data, which were used under license for this study, and so are not publicly available. The data are however available from TSI upon reasonable request and with permission of TSI. Information on micro-level data sets of the TSI is provided on the website (https://www.tuik.gov.tr/Kurumsal/Mikro_Veri). In order to apply for access to micro data of the TSI, researchers are required to contact the TSI and comply with data access procedures of the TSI, as outlined on the aforementioned website. TSI can be contacted directly at the following address: Devlet Mah. Necatibey Cad. No:114, 06420 Çankaya/ANKARA; Phone: 0 (312) 454 70 00; Email: info@tuik.gov.tr.

Funding: This research is supported by The Scientific and Technological Research Council of Turkey through TÜBİTAK-ARDEB-3501 Career Development Program (Project Number: 119K676). The funding institution has no role in conduct of the study and article production. The content is solely responsibility of the author.

Competing interests: The author(s) have declared that no competing interests exist.

Introduction

Men and women differed with respect to their health behaviors and outcomes. For individuals over the age of 15, according to Eurostat data for 2019, more than 52% of females and more than 43% of males reported that they used prescribed medication within the last two weeks in 27 European Union countries [1]. Similarly, the prevalence of non-prescribed medication use was more than 27% for males and more than 37% for females in European countries [1]. Decisions about medication use were associated with biological, social, demographic, and economic factors [2,3]. The conceptual framework for health service use was provided by Andersen (1968) and it is extended by further research [4–7]. The conceptual model framework for Andersen`s behavioral model of health service use categorized factors that impact use of health services into three major groups: 1) predisposing, 2) enabling, and 3) needs [7,8]. Predisposing factors included demographics and social factors. Enabling factors were associated with the policy, financing, and organization of health services. Needs included factors such as body mass index, physical health, and mental health status indicators that lead to actions for the use of health services. Many studies provided empirical tests of the conceptual model for various cases [6–10]. Researchers also indicated that there were multidirectional relationships among the use of health services, health behaviors, health outcomes, and factors pointed out in the model [7–8].

Empirical evidence provided diversified findings on the relationships between gender and the medication use practices of individuals [11–17]. Some researchers found that women have higher total medication use rates than men [18,19]. Additional evidence indicated that women were more likely to use prescribed medications than men [11,20–24]. There were also studies that reported no gender differences in prescribed medication use [25]. A frequent finding indicated that females were more likely to use non-prescription medication [10–12,17,23,24,26–32]. However, some researchers observed that men were more likely to utilize non-prescribed medication [19,33]. Furthermore, some studies found no significant associations between gender and non-prescribed medication use [34–39]. While the extensive literature presented valuable findings, there were limitations and gaps that need to be addressed by additional studies. The mixed findings on the associations of gender with diverse types of medication use may be attributed to multiple factors, such as heterogeneities in data collection, measurement, and analysis methods, sampling differences, and variation in health policies across countries. Many studies used small data sets collected by convenience or purposive sampling methods from local or regional populations. There were also variations in survey measurement tools, questions, and data analysis methodologies. A branch of literature employed multivariate methods, whereas some utilized bivariate analysis, and most ignored simultaneous associations across variables of interest.

Despite the valuable insights on gender differences in medication use, there was no consensus reached among the earlier studies. The primary objective of this research is to make contribution to the current literature by providing empirical analysis for the Turkish case. Turkey introduced transformation programs for health care policies and currently offers universal public health insurance, and individuals were also allowed to obtain several types of private health insurance. In this context, this research had multiple objectives and contributions. First, the author employed a large, nationally representative survey from a developing country to investigate associations between gender and medication use. Second, a specific estimation framework was utilized to analyze gender differences in medication use. Third, unlike earlier studies, the statistical methodology of this research accounted for endogeneities between health status, healthcare service utilization, and medication use. Finally, findings from the Turkish case for medication use would provide additional insights for health policy decisions, and the management of health care in developing countries.

Materials and methods

Data and variables

The author utilized a cross-sectional data set from the 2019 wave of the Turkish Health Survey (THS) [40]. THS 2019 was a nationally representative survey, and it was conducted by the Turkish Statistical Institute (TSI). THS 2019 used stratified sampling techniques and covered 9,470 households from various regions of Turkey. Individuals from 8,166 different households participated in the survey, and this corresponded to a response rate of 86.2% at the household level. The survey was conducted with 17,083 participants who were at least 15 years old. It collected individual-level information on demographics, socioeconomic conditions, health conditions, health-related behavior, health care access, and health service utilization. The data collection process of this study did not require ethical approval for the use of human subjects since it uses a secondary data set. The data set was provided by the TSI through official application and approval processes.

For data analysis of this study four primary outcome variables were created: 1) prescribed medication use; 2) non-prescribed medication use; 3) healthcare service use; and 4) self-rated health status. Non-responses or missing observations were excluded from the analysis by list-wise deletion. Measures for prescribed medication use and non-prescribed medication use by individuals were based on the following survey questions:

1) “During the past two weeks, have you used any medications that were prescribed for you by a doctor? 1-Yes 2-No.” If an individual reported use of any prescribed medication in the past two weeks, then the variable prescribed medication use was equal to one; otherwise, it was zero.
2) “During the past two weeks, have you used any medications not prescribed by a doctor? 1-Yes 2-No.” If an individual reported use of any non-prescribed medication in the past two weeks, then the variable non-prescribed medication use was equal to one; otherwise, it was zero.

Healthcare service use considered whether the respondent used the following healthcare services within the last year: a) visit a hospital; b) visit a physician; c) visit a psychologist, psychotherapist, or psychiatrist. This ordered variable ranged from zero (no service use) to three (use of all three types of service). Self-rated health status was based on the corresponding survey question: “How is your health in general?” This ordered variable was recorded as follows: 1-Very bad 2-Bad 3-Fair 4-Good 5-Very good.

Data analysis considered independent variables for the demographics, socioeconomic factors, and health-related information of individuals. Table 1 presented definitions and measurement details for all variables.

Download:

Table 1. Description of variables.

https://doi.org/10.1371/journal.pone.0321590.t001

Statistical analysis

First, the author conducted a descriptive analysis for the data set of this study. Summary statistics and frequency distributions were calculated for all relevant variables. Two sample comparison tests were conducted for the proportions and means of relevant variables for females and males. Then, the author utilized an empirical analysis that accounted for endogenous relationships between the use of various healthcare services and health status. For instance, individuals may decide to use prescribed and non-prescribed medication simultaneously. An individual may consider prescribed and non-prescribed medication as substitutes and, therefore, may decide not to visit a doctor. Furthermore, the level of perceived health status may simultaneously impact decisions on medication and healthcare service use by individuals. Therefore, the empirical framework employed conditional mixed-process regression models (CMP) [41], due to the nature of the data on the health outcomes of individuals. CMP constructed a simultaneous equation system and allowed estimation of various multiple regression models with several types of variables. Additionally, CMP had the advantage of providing estimations for correlations between multiple dependent variables. Since medication use choices, health status, and healthcare service use practices of individuals were not independently determined, the author estimated a simultaneous equation framework where prescribed medication use (PMU_i), non-prescribed medication use (NPMU_i), self-rated health status (SRH_i), and health service use (HSU_i) were classified as dependent variables. The theoretical model was described by the following system of equations:

(1)

(2)

(3)

(4)

P_i, Q_i, R_i, and S_i represented the vectors of independent variables for each equation, respectively. The vectors of parameters were denoted by , , and . Disturbance terms of equations (e_i, u_i, v_i, and w_i) were assumed to be normally distributed with pair-wise correlations, ρ_st. Given the binary nature of PMU_i and NPMU_i, their corresponding regression models were estimated by the standard probit framework. Since SRH_i and HSU_i were ordered response variables, corresponding regressions were estimated by the ordered probit approach. The simultaneous equation system and correlation coefficients of error terms were estimated by employing the CMP procedures of Rodman (2011) in STATA 17 [41,42]. The systems of equations were estimated for the full sample, the female sample, and the male sample.

Results

Descriptive findings

Summary statistics of relevant variables for the full sample were displayed in Table 2. The total sample included 54.4% females. Overall, 40.7% of participants reported using prescribed medication, and 30.2% used non-prescribed medication. The average self-rated health status was 3.53 (0.81) for the total sample, and for males, their scores (3.66, 0.78) were higher, significantly different from females (3.42, 0.82) on average at the 1% significance level.

Download:

Table 2. Descriptive statistics.

https://doi.org/10.1371/journal.pone.0321590.t002

The mean of healthcare service use was 1.26, indicating that participants used more than one service on average. The mean age of participants fell into the range of 35-44 years. The average level of education was at the level of middle school. The employment rate of the sample was 38.2% and the average household income level corresponded to middle income group. Married individuals constituted 68.8% of the sample, and the average household size was more than three people. The average body mass index was 26.48 (5.07) for the total sample, and for females, their scores (26.67, 5.59) were higher, significantly different from males (26.26, 4.36) on average at the 1% significance level. More than 62% of participants reported chronic conditions, and 92.7% of the sample had health insurance. On average, participants reported lower levels of smoking and alcohol use. Overall, the sample reported elevated levels of vegetable and fruit consumption, 4 to 6 times a week. The average days of physical activity over a week were 4.55 (2.76) for the total sample, and for males, their scores (5.10, 2.62) were higher, significantly different from females (4.09, 2.79) on average at the 1% significance level.

According to Table 3, the characteristics of female and male samples displayed significant differences. The prevalence of prescribed medication use for females (46.56%) was significantly higher than that of males (33.75%). Similarly, prevalence of non-prescribed medication use was significantly higher for females (32.96%) compared to males (26.97%). Males systematically and significantly reported higher levels of self-rated health status than females. More than 26% of men did not use any healthcare services, whereas more than 43% of females used more than one service. Females were more likely to live with chronic illnesses, have health insurance, utilize multiple healthcare services, and frequently consume fruit and vegetables. Males were more likely to be educated, employed, married, have a higher level of household income, be regular smokers, and be regular drinkers. On average, males reported a lower body mass index, higher physical activity, and living with more household members.

Download:

Table 3. Characteristics of participants across gender.

https://doi.org/10.1371/journal.pone.0321590.t003

Findings for associations of gender, medication use, healthcare service use, and health status

Findings from CMP models are displayed in Table 4 for the full sample, Table 5 for females, and Table 6 for males. Robust standard errors are utilized for all regression models. According to Table 4, there are significant gender differences in medication use, self-rated health status, and healthcare service use in Turkey. Females are 19.4% (OR = 1.194 [95% CI: 1.137, 1.254]) more likely to use prescribed medication, and they are 30.8% (OR = 1.308 [95% CI: 1.247, 1.373]) more likely to use non-prescribed medication. On average, females are 9.2% (OR = 0.908 [95% CI: 0.871, 0.947]) less likely to report higher levels of health status, and they are 18.4% (OR = 1.184 [95% CI: 1.138, 1.231]) more likely to use healthcare services.

Download:

Table 4. CMP estimation results: Odds ratios for full sample.

https://doi.org/10.1371/journal.pone.0321590.t004

Download:

Table 5. CMP estimation results: Odds ratios for female sample.

https://doi.org/10.1371/journal.pone.0321590.t005

Download:

Table 6. CMP estimation results: Odds ratios for male sample.

https://doi.org/10.1371/journal.pone.0321590.t006

Estimation results indicate that there are endogenous associations among medication use, healthcare service use, and health status in Turkey. First, there are significantly negative associations between prescribed and non-prescribed medication use (atanhrhos: -0.048 [95% CI: -0.075, -0.021] for the full sample; -0.059 [95% CI: -0.094, -0.023] for females; -0.036 [95% CI: -0.078, 0.006] for males). Prescribed medication users are less likely to use non-prescribed medication, and vice versa. Second, prescribed medication use and self-rated health status display significantly negative correlations (atanhrhos: -0.214 [95% CI: -0.238, -0.189] for the full sample; -0.196 [95% CI: -0.228, -0.164] for females; -0.235 [95% CI: -0.272, -0.197] for males). Namely, individuals with higher levels of self-rated health status are less likely to use prescribed medication, and prescribed medication users are less likely to report higher levels of health. Third, prescribed medication use is positively associated with healthcare service use (atanhrhos: 0.502 [95% CI: 0.477, 0.525] for the full sample; 0.479 [95% CI: 0.447, 0.510] for females; 0.0528 [95% CI: 0.491, 0.564] for males). Individuals who utilize healthcare services are more likely to use prescribed medication, and prescribed medication users are more likely to employ healthcare services. Fourth, non-prescribed medication use and self-rated health status are not significantly related in all samples. Non-prescribed medication use is positively related to healthcare service use for the full sample (atanhrhos: 0.0260 [95% CI: 0.004, 0.047]) and males (atanhrhos: 0.0428 [95% CI: 0.009, 0.076]). Individuals who use healthcare services are more likely to utilize non-prescribed medication, and non-prescribed medication users are more likely to employ healthcare services. There are no significant associations between non-prescribed medication use and healthcare service utilization in the female sample. Finally, healthcare service use and self-rated health status display significantly negative associations (atanhrhos: -0.204 [95% CI: -0.223, -0.184] for the full sample; -0.189 [95% CI: -0.216, -0.163] for females; -0.221 [95% CI: -0.250, -0.191] for males). Specifically, individuals with higher levels of self-rated health status are less likely to use healthcare services, and healthcare service users are less likely to report higher levels of health. It should be noted that these findings are potentially influenced by nature of the data set as it covers information on healthcare service use of participants for the last two weeks. Due to limited time coverage, it would be anticipated that prescribed medicine is negatively correlated with self-rated health statuses and positively correlated with healthcare service use.

Findings for risk factors of medication use, healthcare service use, and health status

Age.

According to Tables 4–6, age level was positively related to prescribed medication use. Individuals who reported higher age levels were more likely to employ prescribed medication. Moreover, older individuals were less likely to use non-prescribed medication, and they were less likely to report higher levels of self-rated health. Table 5 indicated that use of healthcare services was not related to age level for females. According to Tables 4 and 6, there were negative associations between age and healthcare service use for males and the full sample. Older males utilized a smaller number of healthcare services compared to younger males.

Education.

Tables 4–6 demonstrated that education level was not associated with both prescribed and non-prescribed medication use for both males and females. However, it was positively associated with self-rated health status and healthcare service use in all samples. More educated individuals were more likely to report higher levels of self-rated health. Moreover, individuals with higher levels of education used more healthcare services within the previous year.

Employment.

Empirical findings, given in Tables 4–6, indicated that employed individuals were less likely to use prescribed medication and health care services in all samples. However, being employed was positively associated with non-prescribed medication use. Employed individuals were more inclined to utilize non-prescribed medication in all samples. According to Table 6, employed males reported higher levels of self-rated health status. However, Table 5 revealed that there were no significant associations between self-rated health levels and being employed for female participants in Turkey.

Household income and size.

Tables 4–6 demonstrated that household income level was not significantly associated with non-prescribed medication use or healthcare service use in all samples. According to Tables 4 and 6, prescribed medication use was not significantly associated with household income level for males and the full sample. However, according to Table 5, household income displayed positive associations with prescribed medication use in the female sample. Female respondents with higher levels of household income were more likely to utilize prescribed medication compared to those with lower levels of household income. According to Tables 4–6, there were positively significant relationships between self-rated health status and household income in all samples. Individuals with higher levels of household income reported a better self-rated health status in Turkey.

Table 4 demonstrated that household size was negatively associated with prescribed medication. Individuals living with more members within a household were less likely to utilize prescribed medication. Tables 5 and 6 indicated that although this finding was prevalent for the sample of females, there were no associations between prescribed medication use and household size for the male sample in Turkey. Empirical results from Tables 4–6 demonstrated that non-prescribed medication use was not significantly associated with household size in the Turkish case. According to Table 4, individuals residing with a higher number of members in a household were less inclined to use healthcare services. Results from Tables 5 and 6 revealed that this finding was relevant for female participants, whereas there were no significant relations between household size and the use of healthcare services for male participants in Turkey. Finally, according to Tables 4–6, self-rated health status was negatively associated with household size for all samples. Namely, participants who lived in a household with a higher number of members reported lower levels of self-rated health.

Marital status.

Findings, in Tables 4–6, indicated that married individuals were more likely to use non-prescribed medication compared to unmarried individuals in all samples. According to Table 5, married females displayed a lower likelihood of using prescribed medication compared to their unmarried counterparts. Results from Tables 4 and 6 did not indicate any associations between the utilization of prescribed medication and marital status for males or in the overall sample. According to Table 4, being married is significantly and positively related to self-rated health status in the full sample. Overall, married individuals reported better health status in Turkey. However, there were no relationships between marital status and self-rated health level in sub-samples of males and females. Tables 4 and 6 demonstrated that healthcare service use was positively associated with being married for males and the full sample. Namely, married males used more healthcare services compared to their unmarried counterparts in Turkey. On the other hand, Table 5 indicated that there was no significant difference between the healthcare service use levels of married and unmarried females.

Body mass index.

For the Turkish case, according to Tables 4–6, individuals with higher body mass indices were more likely to take prescribed and non-prescribed medication. Additionally, participants with a higher body mass index reported higher levels of self-rated health status in all samples. Table 5 revealed that body mass index was positively associated with higher levels of healthcare service use for females. Females with higher levels of body mass index used more healthcare services. However, according to Table 6, body mass index was not associated with healthcare service use levels for the sample of males in Turkey.

Chronic illness.

Empirical results indicated robust associations between having a chronic illness and the four primary health outcomes. Tables 4–6 indicated that individuals with a chronic illness were more likely to use prescribed and non-prescribed medication. Namely, the likelihood of using prescribed or non-prescribed medication was higher among participants who reported a chronic illness. Moreover, respondents with a chronic illness used more healthcare services in Turkey. According to Tables 4–6, having a chronic illness was negatively associated with reporting higher levels of health status. Respondents with a chronic illness were more likely to state that they had lower levels of self-rated health.

Health insurance.

Results from the full sample, in Table 4, indicated that having health insurance was positively associated with the use of prescribed medication. Respondents who possessed health insurance displayed a higher likelihood of using prescribed medication. Moreover, individuals with health insurance use more healthcare services in Turkey. According to Tables 4 and 5, having health insurance was not significantly associated with non-prescribed medication use for the full sample and females. However, Table 5 revealed that males with health insurance were more likely to use non-prescribed medication. Empirical findings from Tables 4–6 demonstrated that there were no significant relations between self-rated health status and having health insurance for all samples.

Tobacco and alcohol use.

Results from Table 4 revealed that smoking was negatively associated with prescribed medication use in the overall sample. Table 6 indicated that males who reported higher frequencies of tobacco use were less likely to use prescribed medication. There were no significant associations between smoking level and prescribed medication use in the female sample. According to Tables 4–6, smoking levels displayed positive associations with non-prescribed medication use in all samples. Individuals who reported higher frequencies of tobacco use exhibited lower probabilities of utilizing non-prescribed medication. Table 4 demonstrated that smoking was negatively associated with self-rated health status in the overall sample. According to Table 5, females who smoked more frequently reported lower levels of self-rated health. On the other hand, Table 6 revealed that there was no significant association between smoking level and self-rated health status for males. Smoking level displayed significantly positive associations with health service use for only females. According to Table 5, females with higher frequencies of tobacco consumption used more healthcare services in Turkey. However, Tables 4 and 6 indicated that there were no relations between smoking frequency and healthcare service utilization for the overall sample and males.

Empirical findings revealed that alcohol use level was positively associated with self-rated health status, prescribed medication use, and non-prescribed medication use in all samples. Results from Tables 4–6 displayed that individuals with higher frequencies of alcohol consumption were more likely to use prescribed and non-prescribed medications. Similarly, respondents who engaged in more frequent alcohol use had a higher propensity to report a better health status. Tables 4 and 6 indicated that there were negatively significant associations between alcohol use level and healthcare service utilization for males and the full sample. Table 6 revealed that males with higher frequencies of alcohol consumption used more healthcare services in Turkey. However, according to Table 5, alcohol use level did not display significant relations with health service use for females.

Dietary behaviors.

In the full sample results reported in Table 4, vegetable consumption level of individuals was positively related to self-rated health, prescribed medication use, and non-prescribed medication use. Individuals who consumed vegetables more frequently were more likely to utilize prescribed and non-prescribed medication. Additionally, respondents who reported higher frequencies of vegetable consumption displayed better self-rated health statuses. Table 4 also revealed that the frequency of vegetable consumption was negatively associated with healthcare service use. The higher the frequency of vegetable consumption for individuals, the lower the level of healthcare use in Turkey. According to Table 5, the vegetable consumption level of females was positively related to the likelihood of prescribed medication use. However, Table 6 indicated that there was no meaningful relationship between vegetable consumption and prescribed medication use in the male sample. According to Tables 4–6, fruit consumption level was not associated with prescribed medication use or healthcare service use in Turkey. There were significantly positive relationships between fruit consumption and the self-rated health status of individuals. Tables 4-6 revealed that individuals who consumed fruit more frequently were more likely to report higher levels of self-rated health. Results from Tables 4 and 5 demonstrated that non-prescribed medication use and fruit consumption levels were positively associated for females and the full sample. Females with higher frequencies of fruit consumption displayed higher tendencies to use non-prescribed medication. Table 6 displayed that there were no relations between fruit consumption levels and non-prescribed for males in Turkey.

Physical activity.

Empirical findings from Tables 4–6 indicated that physically active individuals were likely to report better health statuses, and they were less likely to use healthcare services and prescribed medication in all samples. According to Tables 4–6, respondents with a higher number of physically active days were less likely to utilize prescribed medication, and they used fewer healthcare services in Turkey. Moreover, individuals who spent more days with walking activity reported higher levels of self-rated health status. Finally, results from Tables 4–6 revealed that there were no significant relationships between physical activity level and non-prescribed medication use among respondents in the Turkish case.

Discussion

Health behaviors and outcomes significantly differed across genders. The main contribution of this research was the provision of empirical analysis for determinants of medication use in a developing country, Turkey, from a gender perspective. Additionally, the author made use of a nationally representative data set and an estimation method that accounted for endogeneities in health outcomes, healthcare service use, and medication use practices of individuals.

The findings confirmed that medication use practices, healthcare service use, and the self-rated health statuses of individuals displayed endogenous relationships. Prescribed and non-prescribed medication use were negatively associated. Hence, residents of Turkey treated prescribed and non-prescribed medications as substitutes. More than 40% and 30% of the general population use prescribed and non-prescribed medication, respectively, in Turkey. Both prescribed and non-prescribed medication use were significantly more prevalent in the female sample. First, these findings were in line with earlier studies that reported a higher likelihood of total medication use for females [18,19]. Second, the results complemented the line of literature that suggested females were more likely to report prescribed medication use [20–24,43]. A fraction of these large gender differences in the Turkish case may be attributed to heterogeneities in the use of different drug categories, as women were more likely to use specific products such as contraceptives. Third, results provided additional evidence for the observation that non-prescribed medication use practices were more frequently observed for females in various countries [10–12,17,23,24,26–32].

In line with earlier research, empirical findings revealed both similarities and differences for determinants of medication use across genders [9,12,22]. Age, body mass index, having a chronic illness, having health insurance, and alcohol use were positively correlated with the prescribed medication use of males and females. For the female sample only, prescribed medication use displayed positive relationships with household income and vegetable consumption. Being employed and having higher levels of physical activity were negatively related to prescribed medication use for both males and females. Married females and females living with more household members were less likely to use prescribed medication. Male smokers were less likely to use prescribed medication.

Non-prescribed medication use was positively associated with being employed, being married, having a high body mass index, having a chronic illness, smoking, and alcohol use for both females and males. For males, having health insurance was positively associated with non-prescribed medication use. Age was negatively related to non-prescribed medication use for both samples. Among females, fruit consumption level was negatively associated with prescribed medication use.

The prevalence of gender differences in medication use may be associated with lower health levels in females compared to males [13]. Another source of the gender gap may be the higher morbidity rates of women [16]. Additionally, some studies pointed out that females used more medication during their reproductive ages and visited doctors more frequently than men [16,44]. Finally, a line of literature argued that females were more prone to self-care than males, and females felt less social pressure to admit that they had pain [28,45]. Supporting these conceptual frameworks, results indicated that females were more likely to report a lower health status and employ more healthcare services.

The findings of this study had implications for public health policies within Turkey’s healthcare system. The Turkish healthcare system operates on universal health insurance coverage with complementary support from the private sector. Medication use by individuals is regulated by requirements on prescriptions and the coverage of health insurance plans. First, empirical results reported that females and males displayed different patterns of medication use. Hence, health policies on regulations and subsidization of drug use should develop interventions by considering gender differences in medication use. Second, there are differences in predictors of prescribed and non-prescribed medication use in Turkey. For instance, household income level and having health insurance are factors associated with prescribed medication use, whereas they are not significantly related to non-prescribed medication use. Public policy interventions on over-the-counter medications should account for such heterogeneities. Finally, dietary choices and physical activity are associated with the medication use decisions of females in Turkey. Health intervention programs for nutrition and drug use could be jointly constructed to address health disparities in Turkey. Overall, the findings of this research suggest that health interventions to shape medication use should incorporate gender-specific approaches. Policies that aim to reduce health disparities in developing countries should not disregard significant differences across sub-groups of the population.

Finally, this study had its own specific limitations. First, the author employed self-reported survey data, which may be subject to measurement problems and reporting biases. For instance, participants may have overstated their height, physical activity level, and healthy eating habits whereas they may have under-reported their weight, alcohol use, tobacco use and consumption of unhealthy products. Moreover, respondents may have recall bias in remembering their past behavior, which would lead to inaccuracies in the data. Survey data were also prone to social desirability bias, especially in health behavior, since participants may change their responses to match their images with social norms. Second, objective measures of health-related factors were not included in the analysis due to data limitations. Survey data provided subjective measures that potentially include framing effects, cognitive limitations of participants, and cultural interpretations. Moreover, the data set was cross-sectional and was not able to capture dynamic relationships among variables of interest over time. Specifically, the data on medication use and health service use cover limited time range of two weeks. Hence, health status, healthcare service use and medication use of participants would naturally display high correlations due to this characteristic of data collection. Third, the data set of the current study did not provide information for the use of various categories of medication. Hence, the author could not explore trends and risk factors for the use specific drug categories. Fourth, although the simultaneous equation framework of CMP considered endogenous relationships among variables for health outcomes, it did not provide causality directions due to the nature of the variables in hand. Therefore, the empirical findings of this study should still be interpreted as correlations. Additionally, like many statistical analyses, the CMP models had both linearity and normality assumptions for the variables of interest. Some health-related variables may not satisfy these assumptions, restricting the reliability of the findings. Future studies should focus on analysis of dynamic and non-linear relations among healthcare service use and health status within panel data on health outcomes. Exploration of causal pathways for medication use and analysis of specific drug categories would significantly contribute to the related research.

Acknowledgments

The author is thankful to the Turkish Statistical Institute for the provision of the data set and to the graduate research assistant of the project, Yasemin Ozturk, for her excellent research assistance. The author would like to thank the undergraduate research assistants for the project, Eda Ince and Furkan Fatih Ozgumus.

References

1. European Statistical Office (Eurostat). Medicine Use Statistics. 2019. Retrieved (08/03/2023) from: https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Medicine_use_statistics
- View Article
- Google Scholar
2. Green MA, Little E, Cooper R, Relton C, Strong M. Investigation of social, demographic and health variations in the usage of prescribed and over-the-counter medicines within a large cohort (South Yorkshire, UK). BMJ Open. 2016;6(9):e012038. pmid:27683515
- View Article
- PubMed/NCBI
- Google Scholar
3. Vogler S, Österle A, Mayer S. Inequalities in medicine use in Central Eastern Europe: an empirical investigation of socioeconomic determinants in eight countries. Int J Equity Health. 2015;14:124. pmid:26541292
- View Article
- PubMed/NCBI
- Google Scholar
4. Andersen R. A behavioral model for families’ use of health services. Center for Health Administration Studies Research Series. Chicago: University of Chicago Press; 1968.
5. Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav. 1995;36(1):1–10. pmid:7738325
- View Article
- PubMed/NCBI
- Google Scholar
6. Babitsch B, Gohl D, von Lengerke T. Re-revisiting Andersen’s behavioral model of health services use: a systematic review of studies from 1998-2011. Psychosoc Med. 2012;9:Doc11. pmid:23133505
- View Article
- PubMed/NCBI
- Google Scholar
7. Von Lengerke T, Gohl D, Babitsch B. Re-revisiting the behavioral model of health care utilization by Andersen: a review on theoretical advances and perspectives. In: Janssen C., Swart E., von Lengerke T. (eds) Health Care Utilization in Germany. 2014. Springer, New York, NY, pp.11–28.
8. Porteous T, Wyke S, Hannaford P, Bond C. Self-care behaviour for minor symptoms: can Andersen’s behavioral model of health services use help us to understand it? Int J Pharm Pract. 2015;23(1):27–35. pmid:24735228
- View Article
- PubMed/NCBI
- Google Scholar
9. Hajek A, Kretzler B, König H-H. Determinants of healthcare use based on the Andersen model: a systematic review of longitudinal studies. Healthcare (Basel). 2021;9(10):1354. pmid:34683034
- View Article
- PubMed/NCBI
- Google Scholar
10. Tavares AI, Ferreira PL, Cavadas V. Factors contributing to self-medication and consumption of non-prescribed drugs in Portugal. Int J Public Health. 2022;67:1604852. pmid:36419733
- View Article
- PubMed/NCBI
- Google Scholar
11. Al-Windi A, Elmfeldt D, Svärdsudd K. Determinants of drug utilisation in a Swedish municipality. Pharmacoepidemiol Drug Saf. 2004;13(2):97–103. pmid:14998071
- View Article
- PubMed/NCBI
- Google Scholar
12. Carrasco-Garrido P, Hernández-Barrera V, López de Andrés A, Jiménez-Trujillo I, Jiménez-García R. Sex--differences on self-medication in Spain. Pharmacoepidemiol Drug Saf. 2010;19(12):1293–9. pmid:20872823
- View Article
- PubMed/NCBI
- Google Scholar
13. Daban F, Pasarín MI, Rodríguez-Sanz M, García-Altés A, Villalbí JR, Zara C, et al. Social determinants of prescribed and non-prescribed medicine use. Int J Equity Health. 2010;9:12. pmid:20441578
- View Article
- PubMed/NCBI
- Google Scholar
14. Del Rio MC, Prada C, Alvarez FJ. The use of medication by the Spanish population. Pharmacoepidemiol Drug Saf. 1997;6(1):41–8. pmid:15073804
- View Article
- PubMed/NCBI
- Google Scholar
15. Dunnell K, Fitzpatrick J, Bunting J. Making use of official statistics in research on gender and health status: recent British data. Soc Sci Med. 1999;48(1):117–27. pmid:10048842
- View Article
- PubMed/NCBI
- Google Scholar
16. Eggen AE. Patterns of medicine use in a general population (0-80 years). The Influence of age, gender, diseases and place of residence on drug use in Norway. Pharmacoepidemiol Drug Saf. 1997;6(3):179–87. pmid:15073786
- View Article
- PubMed/NCBI
- Google Scholar
17. Mayer S, Österle A. Socioeconomic determinants of prescribed and non-prescribed medicine consumption in Austria. Eur J Public Health. 2015;25(4):597–603. pmid:25395395
- View Article
- PubMed/NCBI
- Google Scholar
18. Furu K, Straume B, Thelle DS. Legal drug use in a general population: association with gender, morbidity, health care utilization, and lifestyle characteristics. J Clin Epidemiol. 1997;50(3):341–9. pmid:9120535
- View Article
- PubMed/NCBI
- Google Scholar
19. Martín-Pérez M, Hernández Barrera V, López de Andrés A, Jiménez-Trujillo I, Jiménez-García R, Carrasco-Garrido P. Predictors of medication use in the Roma population in Spain: a population-based national study. Public Health. 2015;129(5):453–9. pmid:25795016
- View Article
- PubMed/NCBI
- Google Scholar
20. Eggen AE. Pattern of drug use in a general population--prevalence and predicting factors: the Tromsø study. Int J Epidemiol. 1994;23(6):1262–72. pmid:7721530
- View Article
- PubMed/NCBI
- Google Scholar
21. Mortensen JT, Olesen AV, Bøggild H, Olsen J, Westergård-Nielsen NC. Socioeconomic correlates of drug use based on prescription data: a population-based cross-sectional register study in Denmark 1999. Dan Med Bull. 2007;54(1):62–6. pmid:17349231
- View Article
- PubMed/NCBI
- Google Scholar
22. Neutel CI, Walop W. Drug utilization by men and women: why the differences?. Drug Information J. 2005;39(3):299–310.
- View Article
- Google Scholar
23. Nielsen MW, Hansen EH, Rasmussen NK. Prescription and non-prescription medicine use in Denmark: association with socio-economic position. Eur J Clin Pharmacol. 2003;59(8–9):677–84. pmid:14523592
- View Article
- PubMed/NCBI
- Google Scholar
24. Öztürk S, Başar D, Özen İC, Çiftçi AÖ. Socio-economic and behavioral determinants of prescription and non-prescription medicine use: the case of Turkey. Daru. 2019;27(2):735–42. pmid:31732873
- View Article
- PubMed/NCBI
- Google Scholar
25. Pappa E, Kontodimopoulos N, Papadopoulos AA, Tountas Y, Niakas D. Prescribed-drug utilization and polypharmacy in a general population in Greece: association with sociodemographic, health needs, health-services utilization, and lifestyle factors. Eur J Clin Pharmacol. 2011;67(2):185–92. pmid:21069519
- View Article
- PubMed/NCBI
- Google Scholar
26. Carrasco-Garrido P, Jiménez-García R, Barrera VH, Gil de Miguel A. Predictive factors of self-medicated drug use among the Spanish adult population. Pharmacoepidemiol Drug Saf. 2008;17(2):193–9. pmid:17654747
- View Article
- PubMed/NCBI
- Google Scholar
27. Chang F-R, Trivedi PK. Economics of self-medication: theory and evidence. Health Econ. 2003;12(9):721–39. pmid:12950092
- View Article
- PubMed/NCBI
- Google Scholar
28. Figueiras A, Caamaño F, Gestal-Otero JJ. Sociodemographic factors related to self-medication in Spain. Eur J Epidemiol. 2000;16(1):19–26. pmid:10780338
- View Article
- PubMed/NCBI
- Google Scholar
29. Lukovic JA, Miletic V, Pekmezovic T, Trajkovic G, Ratkovic N, Aleksic D, et al. Self-medication practices and risk factors for self-medication among medical students in Belgrade, Serbia. PLoS One. 2014;9(12):e114644. pmid:25503967
- View Article
- PubMed/NCBI
- Google Scholar
30. Niclós G, Olivar T, Rodilla V. Factors associated with self-medication in Spain: a cross-sectional study in different age groups. Int J Pharm Pract. 2018;26(3):258–66. pmid:28833710
- View Article
- PubMed/NCBI
- Google Scholar
31. Ocan M, Obuku EA, Bwanga F, Akena D, Richard S, Ogwal-Okeng J, et al. Household antimicrobial self-medication: a systematic review and meta-analysis of the burden, risk factors and outcomes in developing countries. BMC Public Health. 2015;15:742. pmid:26231758
- View Article
- PubMed/NCBI
- Google Scholar
32. Tripković K, Nešković A, Janković J, Odalović M. Predictors of self-medication in Serbian adult population: cross-sectional study. Int J Clin Pharm. 2018;40(3):627–34. pmid:29603073
- View Article
- PubMed/NCBI
- Google Scholar
33. Alhomoud F, Aljamea Z, Almahasnah R, Alkhalifah K, Basalelah L, Alhomoud FK. Self-medication and self-prescription with antibiotics in the Middle East-do they really happen? A systematic review of the prevalence, possible reasons, and outcomes. Int J Infect Dis. 2017;57:3–12. pmid:28111172
- View Article
- PubMed/NCBI
- Google Scholar
34. Hoai NT, Dang T. The determinants of self-medication: evidence from urban Vietnam. Soc Work Health Care. 2017;56(4):260–82. pmid:28113003
- View Article
- PubMed/NCBI
- Google Scholar
35. Ilhan MN, Durukan E, Ilhan SO, Aksakal FN, Ozkan S, Bumin MA. Self-medication with antibiotics: questionnaire survey among primary care center attendants. Pharmacoepidemiol Drug Saf. 2009;18(12):1150–7. pmid:19827009
- View Article
- PubMed/NCBI
- Google Scholar
36. Lei X, Jiang H, Liu C, Ferrier A, Mugavin J. Self-medication practice and associated factors among residents in Wuhan, China. Int J Environ Res Public Health. 2018;15(1):68. pmid:29300318
- View Article
- PubMed/NCBI
- Google Scholar
37. Pagán JA, Ross S, Yau J, Polsky D. Self-medication and health insurance coverage in Mexico. Health Policy. 2006;75(2):170–7. pmid:16338480
- View Article
- PubMed/NCBI
- Google Scholar
38. Isabel Tavares A. Self-medication model and evidence from Portugal. Int J Soc Econ. 2013;40(11):990–1009.
- View Article
- Google Scholar
39. Yuefeng L, Keqin R, Xiaowei R. Use of and factors associated with self-treatment in China. BMC Public Health. 2012;12:995. pmid:23158841
- View Article
- PubMed/NCBI
- Google Scholar
40. Turkish Statistical Institute (TSI). Turkey Health Survey Micro Data Set, 2019. Data set description is available at: https://www.tuik.gov.tr/Kurumsal/Mikro_Veri
- View Article
- Google Scholar
41. Roodman D. Fitting fully observed recursive mixed-process models with cmp. The Stata Journal. 2011;11(2):159–206.
- View Article
- Google Scholar
42. StataCorp. Stata Statistical Software: Release 17. 2021; College Station, TX: StataCorp LLC.
43. Al-Windi A. Determinants of medicine use in a Swedish primary health care practice population. Pharmacoepidemiol Drug Saf. 2005;14(1):47–51. pmid:15534902
- View Article
- PubMed/NCBI
- Google Scholar
44. Leufkens HG, Urquhart J. Variability in patterns of drug usage. J Pharm Pharmacol. 1994;46 Suppl 1:433–7. pmid:8064562
- View Article
- PubMed/NCBI
- Google Scholar
45. Antonov KI, Isacson DG. Prescription and nonprescription analgesic use in Sweden. Ann Pharmacother. 1998;32(4):485–94. pmid:9562147
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. European Statistical Office (Eurostat). Medicine Use Statistics. 2019. Retrieved (08/03/2023) from: https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Medicine_use_statistics
View Article
Google Scholar

[2] View Article

[3] Google Scholar

[ref2] 2. Green MA, Little E, Cooper R, Relton C, Strong M. Investigation of social, demographic and health variations in the usage of prescribed and over-the-counter medicines within a large cohort (South Yorkshire, UK). BMJ Open. 2016;6(9):e012038. pmid:27683515
View Article
PubMed/NCBI
Google Scholar

[5] View Article

[6] PubMed/NCBI

[7] Google Scholar

[ref3] 3. Vogler S, Österle A, Mayer S. Inequalities in medicine use in Central Eastern Europe: an empirical investigation of socioeconomic determinants in eight countries. Int J Equity Health. 2015;14:124. pmid:26541292
View Article
PubMed/NCBI
Google Scholar

[9] View Article

[10] PubMed/NCBI

[11] Google Scholar

[ref4] 4. Andersen R. A behavioral model for families’ use of health services. Center for Health Administration Studies Research Series. Chicago: University of Chicago Press; 1968.

[ref5] 5. Andersen RM. Revisiting the behavioral model and access to medical care: does it matter? J Health Soc Behav. 1995;36(1):1–10. pmid:7738325
View Article
PubMed/NCBI
Google Scholar

[14] View Article

[15] PubMed/NCBI

[16] Google Scholar

[ref6] 6. Babitsch B, Gohl D, von Lengerke T. Re-revisiting Andersen’s behavioral model of health services use: a systematic review of studies from 1998-2011. Psychosoc Med. 2012;9:Doc11. pmid:23133505
View Article
PubMed/NCBI
Google Scholar

[18] View Article

[19] PubMed/NCBI

[20] Google Scholar

[ref7] 7. Von Lengerke T, Gohl D, Babitsch B. Re-revisiting the behavioral model of health care utilization by Andersen: a review on theoretical advances and perspectives. In: Janssen C., Swart E., von Lengerke T. (eds) Health Care Utilization in Germany. 2014. Springer, New York, NY, pp.11–28.

[ref8] 8. Porteous T, Wyke S, Hannaford P, Bond C. Self-care behaviour for minor symptoms: can Andersen’s behavioral model of health services use help us to understand it? Int J Pharm Pract. 2015;23(1):27–35. pmid:24735228
View Article
PubMed/NCBI
Google Scholar

[23] View Article

[24] PubMed/NCBI

[25] Google Scholar

[ref9] 9. Hajek A, Kretzler B, König H-H. Determinants of healthcare use based on the Andersen model: a systematic review of longitudinal studies. Healthcare (Basel). 2021;9(10):1354. pmid:34683034
View Article
PubMed/NCBI
Google Scholar

[27] View Article

[28] PubMed/NCBI

[29] Google Scholar

[ref10] 10. Tavares AI, Ferreira PL, Cavadas V. Factors contributing to self-medication and consumption of non-prescribed drugs in Portugal. Int J Public Health. 2022;67:1604852. pmid:36419733
View Article
PubMed/NCBI
Google Scholar

[31] View Article

[32] PubMed/NCBI

[33] Google Scholar

[ref11] 11. Al-Windi A, Elmfeldt D, Svärdsudd K. Determinants of drug utilisation in a Swedish municipality. Pharmacoepidemiol Drug Saf. 2004;13(2):97–103. pmid:14998071
View Article
PubMed/NCBI
Google Scholar

[35] View Article

[36] PubMed/NCBI

[37] Google Scholar

[ref12] 12. Carrasco-Garrido P, Hernández-Barrera V, López de Andrés A, Jiménez-Trujillo I, Jiménez-García R. Sex--differences on self-medication in Spain. Pharmacoepidemiol Drug Saf. 2010;19(12):1293–9. pmid:20872823
View Article
PubMed/NCBI
Google Scholar

[39] View Article

[40] PubMed/NCBI

[41] Google Scholar

[ref13] 13. Daban F, Pasarín MI, Rodríguez-Sanz M, García-Altés A, Villalbí JR, Zara C, et al. Social determinants of prescribed and non-prescribed medicine use. Int J Equity Health. 2010;9:12. pmid:20441578
View Article
PubMed/NCBI
Google Scholar

[43] View Article

[44] PubMed/NCBI

[45] Google Scholar

[ref14] 14. Del Rio MC, Prada C, Alvarez FJ. The use of medication by the Spanish population. Pharmacoepidemiol Drug Saf. 1997;6(1):41–8. pmid:15073804
View Article
PubMed/NCBI
Google Scholar

[47] View Article

[48] PubMed/NCBI

[49] Google Scholar

[ref15] 15. Dunnell K, Fitzpatrick J, Bunting J. Making use of official statistics in research on gender and health status: recent British data. Soc Sci Med. 1999;48(1):117–27. pmid:10048842
View Article
PubMed/NCBI
Google Scholar

[51] View Article

[52] PubMed/NCBI

[53] Google Scholar

[ref16] 16. Eggen AE. Patterns of medicine use in a general population (0-80 years). The Influence of age, gender, diseases and place of residence on drug use in Norway. Pharmacoepidemiol Drug Saf. 1997;6(3):179–87. pmid:15073786
View Article
PubMed/NCBI
Google Scholar

[55] View Article

[56] PubMed/NCBI

[57] Google Scholar

[ref17] 17. Mayer S, Österle A. Socioeconomic determinants of prescribed and non-prescribed medicine consumption in Austria. Eur J Public Health. 2015;25(4):597–603. pmid:25395395
View Article
PubMed/NCBI
Google Scholar

[59] View Article

[60] PubMed/NCBI

[61] Google Scholar

[ref18] 18. Furu K, Straume B, Thelle DS. Legal drug use in a general population: association with gender, morbidity, health care utilization, and lifestyle characteristics. J Clin Epidemiol. 1997;50(3):341–9. pmid:9120535
View Article
PubMed/NCBI
Google Scholar

[63] View Article

[64] PubMed/NCBI

[65] Google Scholar

[ref19] 19. Martín-Pérez M, Hernández Barrera V, López de Andrés A, Jiménez-Trujillo I, Jiménez-García R, Carrasco-Garrido P. Predictors of medication use in the Roma population in Spain: a population-based national study. Public Health. 2015;129(5):453–9. pmid:25795016
View Article
PubMed/NCBI
Google Scholar

[67] View Article

[68] PubMed/NCBI

[69] Google Scholar

[ref20] 20. Eggen AE. Pattern of drug use in a general population--prevalence and predicting factors: the Tromsø study. Int J Epidemiol. 1994;23(6):1262–72. pmid:7721530
View Article
PubMed/NCBI
Google Scholar

[71] View Article

[72] PubMed/NCBI

[73] Google Scholar

[ref21] 21. Mortensen JT, Olesen AV, Bøggild H, Olsen J, Westergård-Nielsen NC. Socioeconomic correlates of drug use based on prescription data: a population-based cross-sectional register study in Denmark 1999. Dan Med Bull. 2007;54(1):62–6. pmid:17349231
View Article
PubMed/NCBI
Google Scholar

[75] View Article

[76] PubMed/NCBI

[77] Google Scholar

[ref22] 22. Neutel CI, Walop W. Drug utilization by men and women: why the differences?. Drug Information J. 2005;39(3):299–310.
View Article
Google Scholar

[79] View Article

[80] Google Scholar

[ref23] 23. Nielsen MW, Hansen EH, Rasmussen NK. Prescription and non-prescription medicine use in Denmark: association with socio-economic position. Eur J Clin Pharmacol. 2003;59(8–9):677–84. pmid:14523592
View Article
PubMed/NCBI
Google Scholar

[82] View Article

[83] PubMed/NCBI

[84] Google Scholar

[ref24] 24. Öztürk S, Başar D, Özen İC, Çiftçi AÖ. Socio-economic and behavioral determinants of prescription and non-prescription medicine use: the case of Turkey. Daru. 2019;27(2):735–42. pmid:31732873
View Article
PubMed/NCBI
Google Scholar

[86] View Article

[87] PubMed/NCBI

[88] Google Scholar

[ref25] 25. Pappa E, Kontodimopoulos N, Papadopoulos AA, Tountas Y, Niakas D. Prescribed-drug utilization and polypharmacy in a general population in Greece: association with sociodemographic, health needs, health-services utilization, and lifestyle factors. Eur J Clin Pharmacol. 2011;67(2):185–92. pmid:21069519
View Article
PubMed/NCBI
Google Scholar

[90] View Article

[91] PubMed/NCBI

[92] Google Scholar

[ref26] 26. Carrasco-Garrido P, Jiménez-García R, Barrera VH, Gil de Miguel A. Predictive factors of self-medicated drug use among the Spanish adult population. Pharmacoepidemiol Drug Saf. 2008;17(2):193–9. pmid:17654747
View Article
PubMed/NCBI
Google Scholar

[94] View Article

[95] PubMed/NCBI

[96] Google Scholar

[ref27] 27. Chang F-R, Trivedi PK. Economics of self-medication: theory and evidence. Health Econ. 2003;12(9):721–39. pmid:12950092
View Article
PubMed/NCBI
Google Scholar

[98] View Article

[99] PubMed/NCBI

[100] Google Scholar

[ref28] 28. Figueiras A, Caamaño F, Gestal-Otero JJ. Sociodemographic factors related to self-medication in Spain. Eur J Epidemiol. 2000;16(1):19–26. pmid:10780338
View Article
PubMed/NCBI
Google Scholar

[102] View Article

[103] PubMed/NCBI

[104] Google Scholar

[ref29] 29. Lukovic JA, Miletic V, Pekmezovic T, Trajkovic G, Ratkovic N, Aleksic D, et al. Self-medication practices and risk factors for self-medication among medical students in Belgrade, Serbia. PLoS One. 2014;9(12):e114644. pmid:25503967
View Article
PubMed/NCBI
Google Scholar

[106] View Article

[107] PubMed/NCBI

[108] Google Scholar

[ref30] 30. Niclós G, Olivar T, Rodilla V. Factors associated with self-medication in Spain: a cross-sectional study in different age groups. Int J Pharm Pract. 2018;26(3):258–66. pmid:28833710
View Article
PubMed/NCBI
Google Scholar

[110] View Article

[111] PubMed/NCBI

[112] Google Scholar

[ref31] 31. Ocan M, Obuku EA, Bwanga F, Akena D, Richard S, Ogwal-Okeng J, et al. Household antimicrobial self-medication: a systematic review and meta-analysis of the burden, risk factors and outcomes in developing countries. BMC Public Health. 2015;15:742. pmid:26231758
View Article
PubMed/NCBI
Google Scholar

[114] View Article

[115] PubMed/NCBI

[116] Google Scholar

[ref32] 32. Tripković K, Nešković A, Janković J, Odalović M. Predictors of self-medication in Serbian adult population: cross-sectional study. Int J Clin Pharm. 2018;40(3):627–34. pmid:29603073
View Article
PubMed/NCBI
Google Scholar

[118] View Article

[119] PubMed/NCBI

[120] Google Scholar

[ref33] 33. Alhomoud F, Aljamea Z, Almahasnah R, Alkhalifah K, Basalelah L, Alhomoud FK. Self-medication and self-prescription with antibiotics in the Middle East-do they really happen? A systematic review of the prevalence, possible reasons, and outcomes. Int J Infect Dis. 2017;57:3–12. pmid:28111172
View Article
PubMed/NCBI
Google Scholar

[122] View Article

[123] PubMed/NCBI

[124] Google Scholar

[ref34] 34. Hoai NT, Dang T. The determinants of self-medication: evidence from urban Vietnam. Soc Work Health Care. 2017;56(4):260–82. pmid:28113003
View Article
PubMed/NCBI
Google Scholar

[126] View Article

[127] PubMed/NCBI

[128] Google Scholar

[ref35] 35. Ilhan MN, Durukan E, Ilhan SO, Aksakal FN, Ozkan S, Bumin MA. Self-medication with antibiotics: questionnaire survey among primary care center attendants. Pharmacoepidemiol Drug Saf. 2009;18(12):1150–7. pmid:19827009
View Article
PubMed/NCBI
Google Scholar

[130] View Article

[131] PubMed/NCBI

[132] Google Scholar

[ref36] 36. Lei X, Jiang H, Liu C, Ferrier A, Mugavin J. Self-medication practice and associated factors among residents in Wuhan, China. Int J Environ Res Public Health. 2018;15(1):68. pmid:29300318
View Article
PubMed/NCBI
Google Scholar

[134] View Article

[135] PubMed/NCBI

[136] Google Scholar

[ref37] 37. Pagán JA, Ross S, Yau J, Polsky D. Self-medication and health insurance coverage in Mexico. Health Policy. 2006;75(2):170–7. pmid:16338480
View Article
PubMed/NCBI
Google Scholar

[138] View Article

[139] PubMed/NCBI

[140] Google Scholar

[ref38] 38. Isabel Tavares A. Self-medication model and evidence from Portugal. Int J Soc Econ. 2013;40(11):990–1009.
View Article
Google Scholar

[142] View Article

[143] Google Scholar

[ref39] 39. Yuefeng L, Keqin R, Xiaowei R. Use of and factors associated with self-treatment in China. BMC Public Health. 2012;12:995. pmid:23158841
View Article
PubMed/NCBI
Google Scholar

[145] View Article

[146] PubMed/NCBI

[147] Google Scholar

[ref40] 40. Turkish Statistical Institute (TSI). Turkey Health Survey Micro Data Set, 2019. Data set description is available at: https://www.tuik.gov.tr/Kurumsal/Mikro_Veri
View Article
Google Scholar

[149] View Article

[150] Google Scholar

[ref41] 41. Roodman D. Fitting fully observed recursive mixed-process models with cmp. The Stata Journal. 2011;11(2):159–206.
View Article
Google Scholar

[152] View Article

[153] Google Scholar

[ref42] 42. StataCorp. Stata Statistical Software: Release 17. 2021; College Station, TX: StataCorp LLC.

[ref43] 43. Al-Windi A. Determinants of medicine use in a Swedish primary health care practice population. Pharmacoepidemiol Drug Saf. 2005;14(1):47–51. pmid:15534902
View Article
PubMed/NCBI
Google Scholar

[156] View Article

[157] PubMed/NCBI

[158] Google Scholar

[ref44] 44. Leufkens HG, Urquhart J. Variability in patterns of drug usage. J Pharm Pharmacol. 1994;46 Suppl 1:433–7. pmid:8064562
View Article
PubMed/NCBI
Google Scholar

[160] View Article

[161] PubMed/NCBI

[162] Google Scholar

[ref45] 45. Antonov KI, Isacson DG. Prescription and nonprescription analgesic use in Sweden. Ann Pharmacother. 1998;32(4):485–94. pmid:9562147
View Article
PubMed/NCBI
Google Scholar

[164] View Article

[165] PubMed/NCBI

[166] Google Scholar

Figures

Abstract

Introduction

Materials and methods

Data and variables

Statistical analysis

Results

Descriptive findings

Findings for associations of gender, medication use, healthcare service use, and health status

Findings for risk factors of medication use, healthcare service use, and health status

Age.

Education.

Employment.

Household income and size.

Marital status.

Body mass index.

Chronic illness.

Health insurance.

Tobacco and alcohol use.

Dietary behaviors.

Physical activity.

Discussion

Acknowledgments

References