This study aims to understand the spread of (significant) depression in the Austrian population. After a sample description and a general depression status check of Austrian men and women, 6 independent variables will be tested as possibly significant depression predictors. These six variables are then grouped into three pairs, with each pair corresponding to a specific hypothesis.

First, participant distribution by gender:

## 
##   Male Female 
##    993   1361

368 more women than men took part in the study.


Next, checking for age distribution:

## 
##   0–20  21–40  41–60  61–80 81–100 
##     66    455    830    793    210

Most participants were betwen 40 and 80 years old.

Now I want to see how Men vs Women feel about their mental well-being in Austria.


First, feeling of depression (in percent):

##                                  
##                                   Male Female
##   None or almost none of the time 74.2   63.7
##   Some of the time                23.5   32.4
##   Most of the time                 1.6    2.9
##   All or almost all of the time    0.6    1.0

As assumed, men feel more confident about their mental well-being, or at least they do so in reporting.


Now, feeling of loneliness (in percent):

##                                  
##                                   Male Female
##   None or almost none of the time 80.7   74.9
##   Some of the time                15.1   20.1
##   Most of the time                 2.6    3.7
##   All or almost all of the time    1.6    1.2

Same as before, men are less willing to admit to loneliness (or just feel less lonely) -> Nevertheless, it is interesting that for the highest degree of loneliness (all the time), men seem to be more prevalent.

Next to feeling lonely and feeling depressed, 6 other variables constitute to the so called CES_D8 Depression scale: Enjoying life, being happy (both with reversed scales of course), feeling everything is an effort, having restless sleep, feeling sad, and not being able to get going.

According to Briggs et al. (2018), at a cut-off score of 9, the CES_D8 accurately identifies someone as clinically depressed. Let’s see how many people from the Austrian sample count as clinically depressed (in percent):

##                
##                 Male Female
##   depressed     11.7   17.4
##   not depressed 88.3   82.6

This shows that most people fall into the category of significantly not depressed in Austria.
If the study sample is representative, 11.7% of Austrian men and 17.4% of Austrian women count as clinically depressed. This supports prevalent literature, which suggests higher depression in women than men.

Important cronbach-alpha check to see, if the depression variables succeed in measuring the same underlying construct (internal consistency) - i.e. how well they measure depression.

## 
## Cronbach's alpha for the 'DataAUT[, c("fltdpr_num", "flteeff_num", "slprl_num", "wrhpp_num", ' '    "fltlnl_num", "enjlf_num", "fltsd_num", "cldgng_num")]' data-set
## 
## Items: 8
## Sample units: 2354
## alpha: 0.801
This resulted in a standardized alpha value of 0,81, meaning that the depression variables reliably measure depression.

Checking for one of the independent variables now. First, trust in other people. For this, people’s trust is grouped into 3 categories (low, medium, high):

## 
##    Low Trust Medium Trust   High Trust 
##          318          918         1032

This shows that most Austrian people have high or medium trust in other people. There is no real visible difference between the genders.


Hypotheses:

Two independent variables were chosen from three different categories. All of them are expected to negatively correlate with the depression scale (because of the framing), meaning that an increase in these independent variables leads to a decrease on the depression scale.

Hypothesis 1 is supported by literature from Lombardo et al. (2018), Gigantesco et al. (2019) and Gallagher et al (2022).

Hypothesis 2 is supported by literature from Fermin et al. (2022).

Hypothesis 3 is supported by literature from Frasquilho et al. (2016) and Poses & Revilla (2022).


All hypotheses are tested using a multivariate linear regression model.

##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 17.79532    0.54212   32.83  < 2e-16 ***
## ctrlife     -0.38793    0.06335   -6.12  1.2e-09 ***
## stflife     -1.24049    0.05982  -20.74  < 2e-16 ***
## ppltrst     -0.01634    0.05200   -0.31   0.7535    
## pplhlp       0.00784    0.05610    0.14   0.8889    
## stfeco      -0.14139    0.05144   -2.75   0.0061 ** 
## stfdem       0.10853    0.04367    2.49   0.0131 *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Results:

The multivariate test resulted in following coefficients: Ctrlife (-0,388***), stflife (-1,240***), ppltrst (-0,016), pplhlp (0,008), stfeco (-0,141**), stfdem (0,109*), with asterisks representing the significance level. This results in only one of three hypotheses being supported by our calculations: Both variables of general life conception show a highly significant negative correlation with depressive symptoms, therefore H1 is supported. Conceptions about other people showed no significant impact, therefore H2 is refuted. While satisfaction with economy still shows a highly significant negative correlation with depression in the multivariate test, satisfaction with democracy surprisingly shows a slightly significant positive correlation in this model - therefore H3 is (partly) refuted.

Discussion:

Against expectations, two of three hypotheses were not supported by the statistical calculations. This could be due to various factors. Depression is a multifaceted and complex issue that is difficult to determine and measure. Moreover, its determinants cannot be attributed to any single variable or perception, as they result from the complex interaction of various psychological and social factors that are challenging to fully capture. The support for H1 also seemed most likely, considering that the feeling of control over life and satisfaction with life are already very similar to the variables included in the depression scale, they might as well have been part of it. It is likely that for many people, the feeling of not being satisfied with life means the same as feeling depressed and therefore the answers to this question may have directly reflected the answers from the CES-D8 variables. In contrast, perceptions of the national economy might be more detached from personal well-being, as people may not directly relate economic conditions to their mental state in their daily lives.


http://rpubs.com/Pablo99/Final_Paper

References:

Fermin, Alan S. R.; Kiyonari, Toko; Matsumoto, Yoshie; Takagishi, Haruto; Li, Yang; Kanai, Ryota et al. (2022): The neuroanatomy of social trust predicts depression vulnerability. In: Scientific reports 12 (1), S. 16724. DOI: 10.1038/s41598-022-20443-w.

Frasquilho, Diana; Matos, Margarida Gaspar; Salonna, Ferdinand; Guerreiro, Diogo; Storti, Cláudia C.; Gaspar, Tânia; Caldas-de-Almeida, José M. (2016): Mental health outcomes in times of economic recession: a systematic literature review. In: BMC public health 16, S. 115. DOI: 10.1186/s12889-016-2720-y.

Gallagher, Stephen; Daynes-Kearney, Rosemary; Bowman-Grangel, Aoife; Dunne, Nikki; McMahon, Jennifer (2022): Life satisfaction, social participation and symptoms of depression in young adult carers: evidence from 21 European countries. In: International Journal of Adolescence and Youth 27 (1), S. 60–71. DOI: 10.1080/02673843.2021.2025115.

Gigantesco, Antonella; Fagnani, Corrado; Toccaceli, Virgilia; Stazi, Maria Antonietta; Lucidi, Fabio; Violani, Cristiano; Picardi, Angelo (2019): The Relationship Between Satisfaction With Life and Depression Symptoms by Gender. In: Frontiers in psychiatry 10, S. 419. DOI: 10.3389/fpsyt.2019.00419.

Lombardo, Patrick; Jones, Wayne; Wang, Liangliang; Shen, Xin; Goldner, Elliot M. (2018): The fundamental association between mental health and life satisfaction: results from successive waves of a Canadian national survey. In: BMC public health 18 (1), S. 342. DOI: 10.1186/s12889-018-5235-x.

Poses, Carlos; Revilla, Melanie (2022): Measuring satisfaction with democracy: how good are different scales across countries and languages? In: Eur. Pol. Sci. Rev. 14 (1), S. 18–35. DOI: 10.1017/S1755773921000266.