A Cross-Sectional Study on the Associations between Depression and Anxiety, Medication Use for These Diseases and Physical Activity Level in Spanish People with Hypertension

Hypertension (HTN) has a high prevalence in the overall population, affecting people’s mental health. Physical Activity (PA) has proven to be an effective tool to improve mental health. This study analyzed the associations between Depression and Anxiety prevalence, medication use for these disorders (antidepressants and anxiolytics) and Physical Activity Level (PAL) in people with HTN. A cross-sectional study was conducted with data from the Spanish National Health Survey 2017 (ENSE2017) with a final sample of 3228 individuals over 15 years of age with HTN who resided in Spain. Data normality was assessed through the Kolmogorov–Smirnov test. Associations between Depression and Anxiety prevalence, antidepressant and anxiolytic use and PAL were studied using a chi-square test. Possible differences between Depression and Anxiety prevalence and medication use according to the PAL were analyzed with a z-test for independent proportions. Depression or Anxiety and antidepressant and anxiolytic use odds ratios (OR) were calculated for every PAL group, taking the “Very Active” group as a reference. Risk factors were evaluated using multiple binary logistic regression. Dependency relationships were found between Depression and Anxiety prevalence, antidepressant and anxiolytic use and PAL (p < 0.001). The Inactive group displayed the highest prevalence and medication use according to their PAL (p < 0.05). Higher ORs for Depression or Anxiety and pharmacological treatments used were also found in the Inactive group compared to the other PAL groups.


Introduction
Blood pressure (BP) is caused by the blood force exerted against the vessel walls when pumped from the heart [1]. Its measures include Systolic BP (pressure in the arteries at ventricular contraction) and Diastolic BP (pressure during ventricular relaxation) which are used to determine Blood Pressure (BP), taking values below 120/80 millimetres of mercury

Design
This manuscript presents a cross-sectional study based on data from the Spanish National Health Survey 2017 (ENSE2017) [34] by using the adult questionnaire [35]. The ENSE is one of the main data collection programs performed by the MSCBS which collects data on health and health determinants in the Spanish population over 15 years of age. The ENSE is conducted by the MSCBS together with the Spanish National Institute of Statistics (INE) every five years [34]. Complete information on the objectives of the survey, its regulations, sample selection, sample calculation, data collection and processing, among other relevant aspects of its implementation can be found at https://www.sanidad.gob.es/estadEstudios/ estadisticas/encuestaNacional/encuesta2017.htm (accessed on 25 August 2022).

Participants
The ENSE2017 uses a three-stage random stratified sampling type for participant selection including census sections (first stage), family dwellings (second stage) and individuals (third stage), employing a random sampling for each of them [34].
Eligibility criteria included being younger than 70 years, having declared to have HTN (affirmative answer to item Q.25a.1 "Have you ever had hypertension?" Answers: yes, no, don't know or no answer) and presenting data on PA-related items (Q.113-Q.117). Figure 1 describes the excluded participants with their exclusion causes and the final sample for the analysis, composed of 3228 Spanish adults with HTN who were residents in Spain and aged between 15 and 69 years. adults with HTN; and (3) determine the Depression and Anxiety risks in inactives in comparison with physically active Spanish adults with HTN.

Design
This manuscript presents a cross-sectional study based on data from the Spanish National Health Survey 2017 (ENSE2017) [34] by using the adult questionnaire [35]. The ENSE is one of the main data collection programs performed by the MSCBS which collects data on health and health determinants in the Spanish population over 15 years of age. The ENSE is conducted by the MSCBS together with the Spanish National Institute of Statistics (INE) every five years [34]. Complete information on the objectives of the survey, its regulations, sample selection, sample calculation, data collection and processing, among other relevant aspects of its implementation can be found at https://www.sanidad.gob.es/estadEstudios/estadisticas/encuestaNacional/encuesta2017.htm (accessed on 25 August 2022).

Participants
The ENSE2017 uses a three-stage random stratified sampling type for participant selection including census sections (first stage), family dwellings (second stage) and individuals (third stage), employing a random sampling for each of them [34].
Eligibility criteria included being younger than 70 years, having declared to have HTN (affirmative answer to item Q.25a.1 "Have you ever had hypertension?" Answers: yes, no, don't know or no answer) and presenting data on PA-related items (Q.113-Q.117). Figure 1 describes the excluded participants with their exclusion causes and the final sample for the analysis, composed of 3228 Spanish adults with HTN who were residents in Spain and aged between 15 and 69 years.

Variables and Procedures
Sex: SEXOa in the ENSE2017, with options being male or female. Age: EDADa in the ENSE2017, in years. Depression status: item p.25a.20 in the ENSE2017, with options being "yes, not or don't know/no answer." For analyses involving this variable, one participant was excluded for answering "don't know, no answer".
Anxiety status: item p.25a.21 in the ENSE2017: with options being "yes, not or don't know/no answer". For analyses involving this variable, three participants were excluded for answering "don't know, no answer".

Variables and Procedures
Sex: SEXOa in the ENSE2017, with options being male or female. Age: EDADa in the ENSE2017, in years. Depression status: item p.25a.20 in the ENSE2017, with options being "yes, not or don't know/no answer." For analyses involving this variable, one participant was excluded for answering "don't know, no answer".
Anxiety status: item p.25a.21 in the ENSE2017: with options being "yes, not or don't know/no answer". For analyses involving this variable, three participants were excluded for answering "don't know, no answer".
PAL Groups: Groupings were based on participants' Physical Activity Index (PAI), derived from responses to items p.113-p.117 in the ENSE2017, which corresponded to PA carried out over the week and ranged from 0 to 67.5. Several factors were applied to the frequency, duration and intensity of participants' PA, using an adaptation [36] to Nes et al. formula [37]. Groupings were: Inactive (PAI = 0, reported not walking one day a week, p.117), Walker (PAI = 0, reported walking at least one day a week), Active (PAI = 1-30) and Very Active (PAI < 30).

Statistical Analysis
IBM SPSS Statistics v. 25 software (IBM Corp., Armonk, NY, USA) was used for statistical analyses. Data normality was studied using the Kolmogorov-Smirnov test. A descriptive analysis was performed to characterise the sample, using median and interquartile range (IQR) for continuous variables (age and BMI), and absolute and relative frequencies for categorical variables (marital status, PAL, Depression status, Anxiety status, antidepressant and tranquilizer use) presented for the overall population and grouped by sex. The following non-parametrical tests were conducted: the Mann-Whitney U test to analyze possible differences between men and women in the continuous variables; a z-test for independent proportions for possible differences in proportion between sexes and the Depression status, Anxiety status, antidepressant and tranquilizer use proportions according to the PAL; and chi-square for the associations between categorical variables and sex. The contingency coefficient was calculated to assess the strength of the previous relationships interpreted according to Schubert [38]. Odds ratios for Depression and Anxiety status and antidepressant and tranquilizer use were calculated according to the PAL, using the Very Active group as a reference. Finally, multiple binary logistic regressions were performed considering as dependent variables the Depression status, Anxiety status and antidepressant and tranquilizer use, using as independent variables sex, age, PAL, BMI and civil status. For all analyses, a significance level <0.05 was considered statistically significant.

Results
The Kolmogorov-Smirnov test found insufficient evidence to accept the normality of the data so parametric tests were used. Table S1 shows the descriptive analysis to characterise the sample according to age, BMI, marital status, PAL, Depression and Anxiety status and tranquilizer and antidepressant use, both in the overall population and divided by sex. Dependence relationships were found between sex and marital status, PAL, Depression status, Anxiety status and tranquilizer and antidepressant use (p < 0.001 from the chi-square test). Higher Physically Active (23.1% vs. 17.6%, p < 0.05 from z-test) and Very Active (8.5% vs. 5.0%, p < 0.05 from z-test) proportions were found in men than in women. Depression (24.4% vs. 11.4%, p < 0.05 from z-test) and Anxiety (21.7% vs. 10.5%, p < 0.05 from z-test) prevalence was also higher in women than men ( Figure 2). Women also presented higher antidepressant (15.6% vs. 6.3%, p < 0.05 from z-test) and tranquilizer (26.6% vs. 13.7%, p < 0.05 from z-test) use than men. Table S2 displays the associations between Depression and Anxiety prevalence according to the overall population, men, and women with HTN PAL. Dependency relationships were found in the overall population between the PAL and Depression and Anxiety prevalence (p < 0.001). The same results were found when dividing by men and women (p < 0.001). Depression (26.0%) and Anxiety prevalence (22.5%) was higher in the Inactive group rather than in the other PAL groups (Figure 3). This prevalence was lower as participants' PAL increased, with significant differences between Inactives and Walkers (p < 0.05), as well as between these two groups and the Active and Very Active groups (p < 0.05), but not between the Actives and Very Actives (p < 0.05). The same happened when comparing by sexes.  Table S2 displays the associations between Depression and Anxiety prevalence according to the overall population, men, and women with HTN PAL. Dependency relationships were found in the overall population between the PAL and Depression and Anxiety prevalence (p < 0.001). The same results were found when dividing by men and women (p < 0.001). Depression (26.0%) and Anxiety prevalence (22.5%) was higher in the Inactive group rather than in the other PAL groups (Figure 3). This prevalence was lower as participants' PAL increased, with significant differences between Inactives and Walkers (p < 0.05), as well as between these two groups and the Active and Very Active groups (p < 0.05), but not between the Actives and Very Actives (p < 0.05). The same happened when comparing by sexes.  Table 1 shows the dependence relationships found between PAL and tranquilizer and antidepressant use, both in the population with HTN and according to sex (p < 0.001). Inactive people had higher tranquilizer (30.9%) and antidepressant (16.7%) use compared to participants with higher PALs. In men, the group with the lowest tranquilizer (8.9%) and antidepressant use (3.3%) was the Active one; in women, this happened in the Very Active group (tranquilizers 15.8%; antidepressants 7.9%).    Table S2 displays the associations between Depression and Anxiety prevalence according to the overall population, men, and women with HTN PAL. Dependency relationships were found in the overall population between the PAL and Depression and Anxiety prevalence (p < 0.001). The same results were found when dividing by men and women (p < 0.001). Depression (26.0%) and Anxiety prevalence (22.5%) was higher in the Inactive group rather than in the other PAL groups (Figure 3). This prevalence was lower as participants' PAL increased, with significant differences between Inactives and Walkers (p < 0.05), as well as between these two groups and the Active and Very Active groups (p < 0.05), but not between the Actives and Very Actives (p < 0.05). The same happened when comparing by sexes.  Table 1 shows the dependence relationships found between PAL and tranquilizer and antidepressant use, both in the population with HTN and according to sex (p < 0.001). Inactive people had higher tranquilizer (30.9%) and antidepressant (16.7%) use compared to participants with higher PALs. In men, the group with the lowest tranquilizer (8.9%) and antidepressant use (3.3%) was the Active one; in women, this happened in the Very Active group (tranquilizers 15.8%; antidepressants 7.9%).   Table 1 shows the dependence relationships found between PAL and tranquilizer and antidepressant use, both in the population with HTN and according to sex (p < 0.001). Inactive people had higher tranquilizer (30.9%) and antidepressant (16.7%) use compared to participants with higher PALs. In men, the group with the lowest tranquilizer (8.9%) and antidepressant use (3.3%) was the Active one; in women, this happened in the Very Active group (tranquilizers 15.8%; antidepressants 7.9%). Table 2 shows how the Inactive and Walker groups had Depression and Anxiety risks significantly elevated when compared to the Very Active in the overall population. The Inactive ones were 3.35 times more likely to have Depression and 3.27 times more likely to have Anxiety compared to the Very Active ones. Walkers were 2.10 and 2.21 times more likely to have Depression and Anxiety, respectively, compared to the Very Active groups. Similar findings were found when divided according to sex (Table 2).
Increased risks of using tranquilizers (OR: 3.50, 95% CI: 2.22-5.51) and antidepressants (OR: 3.50, 95% CI: 1.87-6.53) were found in the Inactives compared to the Very Active group in the overall population, in men and women (Table 3).    Tables S3 and S4 show the binary multiple logistic regression models concerning Depression status, Anxiety status, and tranquilizer and antidepressant use. These models explained 9.6%, 8.0%, 8.1% and 8.7% (Nagelkerke R 2 ), respectively, of the variance of the above variables. Highly active and married individuals showed lower Depression, Anxiety, and tranquilizer and antidepressant use risks.

Discussion
This study aimed to examine the associations between PAL and Anxiety and Depression status in individuals aged 15-69 years with HTN in a Spanish population cohort conducted in 2017. The key finding of this study was that physical inactivity was associated with the occurrence of anxious and depressive symptomatology. HTN is the main modifiable risk factor for cardiovascular disease and mortality worldwide, and a sedentary lifestyle is a modifiable risk factor for HTN [39]. In this condition, environmental and lifestyle factors play an important role in its onset and course [40], with PA being a protective factor in both the prevention and treatment of HTN [41], offering psychological improvements and improvements related to Depression and Anxiety, among others [42]. Patients with chronic conditions like HTN may experience many negative emotions and increased risk for mental health disorders such as Anxiety and Depression [43][44][45]. Thus, several studies conclude that people with HTN are more likely to suffer from them. One study of people with AHT is associated with psychiatric comorbidity and reduced quality of life, which is influenced by many sociodemographic and clinical factors [46]. Another study found that prehypertensive patients suffered more frequently from Depression compared with other patients with HTN, with family history, smoking and alcohol consumption being important factors [47]. Additionally, a systematic review indicates that there is evidence of a bidirectional relationship between mental health and HTN, such that elderly people with HTN would be more likely to have Depression and Anxiety, resulting in poorer quality of life and higher mortality [48].
In our study, inverse associations were found between PAL and anxious and depressive symptomatology in people with HTN, in line with findings by other authors. Physical exercise has effects on physical, mental and emotional health [49][50][51]. In the case of HTN, it is associated with a variety of improvements [52][53][54]. Depressive symptoms are common in people with HTN even in the absence of comorbidities. Thus, HTN and depressive symptoms form a combination that affects quality of life and increases all-cause mortality, so efforts to influence modifiable factors to promote suitable lifestyles should be made. Alcohol consumption and obesity predicted depressive symptoms, while non-smoking and moderate PA compared with low activity appeared to buffer depressive symptoms [55]. PA and sleep duration also impact depressive symptomatology in people with HTN [56]. Therefore, it is necessary to establish screening for Depression in patients with HTN, as untreated Depression is associated with poorer quality of life, as well as higher healthcare costs [57]. Indeed, the sense of well-being associated with aerobic exercise training may be related to the release of neurotransmitters and weight loss derived from exercise [58]. Differences were also found in people with HTN when divided by sex. One paper found significantly higher anxiety scores in women, but not in Depression [59], and another found higher Depression risks in women [55]. These results could be related to several factors, one of them being the lower PAL that women generally perform [60].
Practical applications of this study include a better understanding of relationships in the analyzed variables. As PA has proven its value as a health promotion tool, stakeholders should consider exercise benefits not only at a physical level but at a psychological one as challenges in this matter exist in people with HTN. Then, a better knowledge of these factors will help to better manage HTN psychological consequences. On the one hand, implementing PA programs to help individuals with HTN and increasing their PAL could lead to increased health, as PA reduces Anxiety and Depressive symptoms [25,30,51,[61][62][63]. This, together with the fact that in the Spanish population, in line with other countries, there has been an increase in physical inactivity and sedentary habits, has led the government to promote several policies to prevent the effects of such potentially harmful habits on the population [32]. Moreover, as this study was conducted with data from the ENSE2017, the last one before the COVID-19 pandemic, it will provide a framework to compare with future health surveys. On the other hand, people with mental health issues may have greater barriers accessing or engaging PA or sports activities, therefore this must be considered when offering PA activities or programs for these populations [64,65].
This study has several limitations. On the one hand, as this is a cross-sectional study, it is impossible to analyze cause-effect relations, but only possible to identify associations. On the other hand, due to the structure of the ENSE2017, it is not possible to compare moderate and intense PA in relation to the state of mental health with the type of PA performed or to discriminate between different types or circumstances of the HTN, such as if it was well-controlled or not, as participants were not questioned about this. We also do not have information on the medication dosage, nor on other medications that could influence the mental state of the participants. Future studies should complement data collection with questionnaires and subjective perception by monitoring objective physiological data or PA dosage (duration and intensity).

Conclusions
Dependency relationships were found between Depression and Anxiety prevalence, antidepressant and anxiolytic use and PAL (p < 0.001). The Inactive group displayed the highest prevalence and medication use according to their PAL (p < 0.05). Higher ORs for Depression or Anxiety and pharmacological treatments used were also found in the Inactive group compared to the other PAL groups.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijerph20031803/s1, Table S1: Descriptive analysis; Table S2: Relationship between Depression and Anxiety Status; and Physical Activity Level; Table S3: Logarithmic binary regression model for Depression and Anxiety risk factor; Table S4: Logarithmic binary regression model for Tranquilisers and Antidepressants Use risk factor.