THE INFLUENCE OF EDUCATION ON DIFFERENCES IN DEPRESSIVE SYMPTOMS BETWEEN MEN AND WOMEN IN
SLOVENIA
VPLIV IZOBRAZBE NA RAZLIKE MED SPOLOMA V PRISOTNOSTI SIMPTOMOV DEPRESIJE V SLOVENIJI
Tanja Kamin1, Nejc Berzelak1, Mirjana Ule1
Prispelo: 14. 7. 2011 - Sprejeto: 21. 10. 2011
Original scientific article UDC 616.89(497.4):331.21
Abstract
Background: This p^ap^erdiscusses de^pi^essive symptoms among me^n and women in Slovenia e^nd theiri^elationship to v^ar^ious socioeconomic factors, and education in particular^.
Methods: The analysis is based on the Eur^op^an Social Surrey Round 3 (ESS-3) fr^om 2006, for the Slovene population (n = 1,282). Depressive symptoms, as a dependent variable, are measured using an 8-item version of the Centre for Epidemiological Studies - Dep^ression (CES-D 8) scale. Independent v^ar^iables included in the r^esear^ch mod^l a^re: gender, age, education, income, m^r^ital and employment status and the pr^es^nce of a child under the age of 12 in the household. Using mean compar-isons of depr^ession symptoms and re^res^ion a^nal^s^s, the papier pr^esents gender diff^i^ences in d^pr^ession levels and factors that influence it.
Results: Education contr-ibutes to low^er d^pr^ession levels in both genders; h^owever, its influence is substantially higher among wom^n. Dep^r^ession symptoms ar^e closely r^elated to education. Lower educated women show a significantly higher scor^e in d^pr^ession symptoms th^n lower educated men. however-, higher educated women show t^ett^r mental health than higher educated men. Differ^ent socioc^emogra^phic factors influence the levels of d^pr^ession symptoms differently between genders. The imp^act of housework as a^n employment status thus significantly influences higher levels of dep^r^ession only a^mong men. Simllar indications for age, v^ido^hood and the absence of partnership. In contr^ast, the influence of work disability on depr^ession is only significant for women. Conclusion: The influence of socioeconomic and cultur^^l factors on d^pr^ession symptoms is gr^^ater for women th^n men.
Key words: gender, inequalities, depression symptoms, education, women, mental health
Izvirni znanstveni članek UDK 616.89(497.4):331.21
Izvleček
Uvod: Prispevek razpravlja o spolnih r^azlikah vpr^isotnosti simptomov depr^esivnosti v Sloveniji in njihovi pov^z^nosti s temeljnimi socialnodemogr^afskimi dejavniki, predvsem z izobr^azbo.
Metode: Analiza temelji na anketnih podatkih Evropske družboslovne r^aziskave iz leta 2006. Vzorec sestavlja 1.282 anketirancev, starih od 18 do 75 let. Ocena stopnje depr^esivnosti temelji na skr^ajšani lestvici CES-D 8. Neodvisne spr^emenljivke so: spol, starost, dosežena izobrazba, dohodek, zaposlitveni in zakonski status ter prisotnost otroka do 12 let v gospodinjstvu. S pr^imerjavo povprečij stopnje simptomov depresivnosti in regresijskimi analizami prispevek pr^ikaže razlike v stopnji depresivnosti in dejavnikih vpliva nanjo glede na spol.
Rezultati: Izobr^azba prispeva k nižji stopnji depr^esivnosti pri obeh spolih, vendar je njen vpliv izr^azito močnejši med ženskami. Čim nižja je izobr^azba, tem večje so r^azlike v stopnji depr^esivnosti med spoloma v škodo žensk. Ta trend se obrane pri najvišji izobr^azbi, kjerje stopnja depr^esivnosti pr^i moških višja kot pri ženskah. Pr^eostali socialnodemogr^afski dejavniki različno vplivajo na stopnjo depr^esivnosti glede na spol. Vplivi starosti, dela v gospodinjstvu, ovdovelosti
in nevključenosti v partnersko zvezo na stopnjo depresivnosti so značilni le za moške. Nasprotno je le pri ženskah mogoče zaznati statistično značilen vpliv nezmožnosti za delo na stopinjo d^pr^esivnosti.
Zaključek: Socialno-e^onomski in l^ultur^ni deja^ni^i izr^aziteje vplivajo na oceno d^pr^esi^nosti pr^i ^ensk^a^h kot f^ri moških.
Ključne besede: depresivnost, neenakosti, spol, izobrazba, ženske, duševno zdravje
1 Introduction
Poor mental health is becoming a major burden in the European Union: it is anticipated that problems related to mental health will become one of the most prominent health issues by 2020 (1). Like other health problems, we find that difficulties in mental health are not distributed evenly; they are concentrated in socially recognized groups of population. The results of the national research on the mental health of Slovenian citizens, conducted in 2006 by the Institute of Public Health, also demonstrate this (2). Among those who reported problems with mental health, there were significantly more women. This is not specific to Slovenia, since this gender gap in depression is one of the most consistent findings in the social epidemiology of mental health. Depression is approximately twice as common among women as among men (3, 4, 5). The question that arises is why?
Usually the differences in depression across the genders are substantiated by biological factors (neurohormonal or genetic) (4, 5) and psychological factors (the role of stressors and the individual reaction to them...) (6). The stress models and the vulnerability models mostly describe the connection between the stressors a person is subjected to and the individual reaction to them as the main reason for depressive reactions (6). In the past decades, social studies and social epidemiology are trying to divert attention away from de-contextualized individual origins of mental problems to analysing their social contexts; thus to inequality in health in relation to gender, social position, economic status, ethnicity and so forth (1, 7). The differences between the genders in the prevalence of depressive symptoms differ from country to country. This has encouraged us to think about the interfering factors that crucially moderate the strength of the influence of gender on mental health. The analysis of socially and culturally burdensome factors or the group of gender-specific social risk factors is, in our opinion, pivotal to understanding the differences in depression across the genders. Recently, this has also been pointed out by Van de Velde et al (5). Using information from the third wave of the European Social Survey (ESS-3), they illustrated cross-national variation in gender differences in
depression in 23 countries. This shows the presence of various social origins of stress: in different countries, the daily lives of men and women are imbedded in different social circumstances and their acknowledgement is of crucial importance for understanding the gender gap in depression. Their explanatory models show that the socioeconomic factors of stress (education, household income and employment status) as well as family-related factors of stress (marital status, type of household, children in the household) moderate the relationship between gender and depression. They ascertain that a low risk of depression is typical of those with a higher socioeconomic position and those who are in a stable partnership. Gender differences in depression are higher in some of the Southern and Eastern European countries. Among the social factors of stress, the role of education has been emphasized; higher education is associated with better mental health. This association was found to be more important among women than among men (5). This data suggests that institutional conditions, such as the accessibility of education, employment and economic independence, substantially influence the position of women in society and their self-image, and consequently also their mental health and well-being. In another study based on the same data, Von dem Knesebeck and his colleagues observe the relationship between education and depressive symptoms and conclude that people with low education have elevated risks of experiencing a high score of depressive symptoms. They present the crosscountry variability in depressive symptoms across 22 European countries and point out that education is very strongly connected to depression across both genders, especially in Hungary and Slovenia (8). Both Von dem Knesebeck et al and Van de Velde et al highlighted the gender gap in the prevalence of depressive symptoms and the role of some social factors that operate either protectively or as risk factors in depression development (8, 5). At the same time, they emphasise that their research is only the beginning, which encourages more thorough research of the association between the chosen factors and the differences in the prevalence of depression between the genders, in order to better understand these relationships. They have not explained the country-specific factors that moderate the
association between gender, education and depression, and the related deviation of individual countries, including Slovenia, from the average in the prevalence of depressive symptoms, but they have emphasized the need to do so.
In this article, we take the above appeal as our starting point. We will focus on the influence of education, which has, according to numerous studies, proved to be the most important indicator of socioeconomic position, which influences the differences in the assessment of depression (5). Muntaner and his colleagues tried to prove that education in itself does not influence health directly, but influences access to more or less dominant social positions (9). The competition for less qualified jobs is higher and so is the insecurity of employment in these positions; the exchangeability and dispensability of people are much higher than in more qualified workplaces. Generally we observe that the higher people's education, the better their self-assessed health. In Slovenia, people with a basic education or less have the worst self-assessed health (10). Furthermore, mortality among less educated people due to preventable causes is twice as high as among better-educated people (11). Common sense tells us that the connection between education and health is as follows: the more educated people are healthier because they are better informed about the influence of lifestyle on their health. Nevertheless, being informed is only one of the variables we can use to explain the education-related differences in health. Education, as Bourdieu said, is just one form of cultural capital, so-called institutionalized cultural capital (12). Up to a point, this form of capital can be directly expressed in economic capital, e.g. the income related to the work position or in special money or material bonuses that come with the work position (e.g. profit-sharing, using a company cell phone or car). It can also be expressed as a special skill or as symbolic capital and the authority that comes with it. Better education on the one hand enables a better job, more autonomy and consequently a better reputation, higher income and a better social position. From this point of view, cultural capital converts to economic and social capital, which indirectly enables better access to health. But education and acquired skills give us, at the same time, a wider view of the world, easier access to new knowledge, the skill to search for information, faster learning and a better understanding of new data, openness to novelties, the ability of critical judgement and the flexible use of new knowledge to improve life in general (13).
Education undeniably influences the prevalence of depression significantly. But it is interesting that its
influence is particularly strong in some countries. This is especially true for Slovenia (8). In this article, with the help of the same database that was used by the previously mentioned authors, i.e. the European Social Survey (14), we will inspect the influence of education on the level of depressive symptoms in Slovenia and how much education influences this assessment in both genres compared to some other socio-demographic factors. We aim to expose and discuss potential reasons for the significant importance of education as a differing factor in gender related inequality in the level of depression in Slovenia.
2 Methods
We obtained the data for analysis from the European Social Survey (14). We used the data for Slovenia that was collected in the third round of data collection in 2006/2007 (15). 1,476 individuals were surveyed using face-to-face interviews with an overall response rate of 65%. The sample was obtained using stratified random sampling from the Central Population Register (14). The whole sample for Slovenia is composed of 45% men and 55% women. The mean age of the respondents was 46.8 years.
We limited the analysis to respondents between 18 and 75 years old. This ensures higher variability in the socio-demographic variables related to education, employment status and marital status. The sample is thus also directly comparable to the sample of the study conducted by Van de Velde et al (5). This restricted sample consists of 46% men and 54% women with a mean age of 46.1 years across both genders.
2.1 Depression Level
In order to estimate the level of an individual's depression, we used a similar approach to Van de Velde et al (5). Depression was measured using the short-form depression scale CES-D (16), which consists of eight indicators of the current frequency of depression symptoms as presented in Table 1. Each of them can take values from 0 (none or almost none of the time) to 3 (all or almost all the time). Only respondents with valid values for all the indicators were included in the analyses. We calculated the overall depression level as the sum of the values of the individual indicators. Depression level scores thus range from 0 to 24, where a higher value denotes a higher depression level.
Table 1. The text of CES-D items used in the ESS Round 3. All indicators are measured on a four-point scale: none or almost none of the time, some of the time, most of the time and all or almost all of the time.
Tabela 1. Besedilo indikatorjev CES-D, uporabljeno v tretjem krogu ankete ESS. Lestvica je pri vseh indikatorjih štiristopenjska: nikoli ali skoraj nikoli, le malo časa, večino časa in ves čas ali skoraj ves čas.
Question text: Please tell me how much of the time during the past week^ /
Besedilo vprašanja: Prosim, povejte, koliko časa ste se prejšnji teden
Indicators / Indikatorji_
^ you felt depressed? / . počutili depresivno?
you felt that everything you did was an effort? / . ste imeli občutek, da vse stvai^i delate s težavo? . your sleep was restless? / ^ ste imeli nemiren spanec? . you were happy? / . ste bili srečni? . you felt lonely? / . ste se počutili osamljeno? . you enjoyed life? / . ste uživali v življenju? . you felt sad? / . ste bili žalostni?
. you could not get started? / ^ se nekako niste mogli spraviti v pogon?
2.2 Independent Variables
In the first step, we observed the differences in the levels of depression between genders according to education achieved. Education was measured with three levels: basic education or less, upper secondary education and higher education. For further analysis of the socio-demographic characteristics that influence levels of depression, we used the following independent variables:
•	age,
•	household income,
•	number of completed years of education,
•	employment status,
•	marital status and
•	parenthood of a child under 12 years old.
We included the socio-demographic characteristics in a linear regression as the predictors (independent variables) of the level of depression symptoms measured with CES-D 8 (dependent variable). We dichotomized the values of the nominal variables. In contrast to the analysis by Van de Velde et al (5), we did not include the variable of the respondent currently living with a partner. The regression diagnostics highlighted a high value of variance inflation factor for this variable (VIF= 12.14), indicating the presence of multicolinearity. Since this presents a violation of one of the regression assumptions, we decided to exclude the variable from the analysis.
3 Results
A comparison of the mean level of depression in men and women is presented in Table 2. The mean level of depression for all respondents is 5.53 on the scale from 0 to 24. Among men, this level equals 5.18 and it is somewhat higher in women at 5.82. The mean difference is thus 0.64 and is statistically significant at 1% level of significance. As a group, women exhibit only slightly higher depression levels than men. However, it is necessary to consider the heterogeneity of women regarding their varying social, cultural and economic contexts and the influence of such contexts on mental health. Several studies in social epidemiology confirmed the highly burdensome effects of a low socio-economic status for women. We selected education as a basic indicator of differences in the social and cultural position of women. The comparative cross-country studies in the ESS show that education is the most important socioeconomic indicator that influences variability in depression (5). According to Von dem Knesebeck et al (8), this is even more evident in Slovenia than in other European countries. Table 3 shows the differences in the levels of depression according to education and between genders.
Table 2. A comparison of the mean depression level measured with CES-D in men and women. Tabela 2. Pr^imerja^a povpi^ečne stopanje d^pr^esi^nosti po CES-D med mošk^imi in ženskami.
Gender / Spol
Mean depression symptom scores / Povprečna stopnja depresivne simptomatike
Mean / Povprečje
Std. deviation / Std. odklon
Difference (F - M) / Rc^z^Hk^a (Ž - M)
Men / Moški	5.18	0.13	
(n = 569)			0.64***
Women / Ženski	5.82	0.15	
(n = 673)			
Total / Skupaj (n=1230)
5.53
0.10
|f| = 3.13, p < 0.01
Table 3. A comp^ar^ison of the mea^n depr^ession le^el measur^ed with CES-D in men and women accor^ding to achieved education.
Tabula 3. Pr^im^r^java povp^r^ečne stopnje d^p^r^esi^nos^ipo CES-D medmošk^imi in ž^nsk^ami glede na dos^ž^no izobr^azbo.
Mean depression symptom scores by gender (standard deviation) /
Education /	Povprečna stopnja depresivne simptomatike po spolu
(standardni odklon)
	Men /	Women /	Total /	Difference (F	- M) /
	Moški	Ženske	Skupaj	Razlikca (Ž -	M)
Basic or less /	6.05	7.91	7.29		
Osnovna šola ali manj	(3.83)	(4.62)	(4.45)	1.86	
	n = 86	n = 173	n = 259		
Upper secondary /	5.18	5.62	5.39		
Srednja šola	(3.15)	(3.59)	(3.37)	0.44	
	n = 371	n = 342	n = 713		
Higher or more /	4.51	3.97	4.19		
Višja šola ali več	(2.56)	(2.91)	(2.78)	-0.54	
	n = 112	n = 158	n = 270		
Total /	5.16	5.79	5.52		
Skupaj	(3.19)	(4.00)	(3.66)		
	n = 569	n = 673	n = 1242		
ANOVA: F = 25.47, p < 0.01; Both the main effects and the interaction of gender and education are significant at p < 0.01 / Oba glavna učinka in interakcija spola ter izobrazbe so statistično značilni pri p < 0.01.
Depression level decreases with higher education for both genders. We also observe that the differences in the mean depression level between genders depend on the level of education. Among the respondents with the lowest level of education (basic or less), the mean depression score is higher for women, with a difference of 1.86. The lowest difference is between respondents with upper secondary education (0.44) and only slightly higher among those with higher education (0.54). An important observation is that women exhibit a higher level of depression in the two groups of the lower educated. On the contrary, among those with higher education, the depression level is higher for men. This is confirmed by the statistically significant interaction between both variables, suggesting that the impact of gender on depression depends on the education achieved.
In summary, the key findings are:
1.	Education is an important differing factor for depression level between genders.
2.With	a lower education level, women exhibit higher depression than men.
Table 4. Multiple regression of socio-demographic variables on depression levels by gender. Tabela 4. Multipla regresijska analiza vpliva socio-demografskih spremenljivk na stopnjo depresivnosti po spolu.
3.This trend reverses in the group of the highly educated where the depression level is higher for men.
To study the degree of contribution (influence) of education and other socio-demographic independent variables on depression level, we conducted multiple linear regression analysis. In order to further examine the gender differences in the influence of these factors, we performed the analysis separately for men and women (Table 4). The standardized regression coefficients in Table 4 provide information on the strength of the independent influence of the predictor variables on depression level. It should be noted that education is here measured as the number of completed years of education in contrast to the three levels of education used above. Since the number of years is a continuous variable, it is more appropriate and informative to use this in the regression instead of the three-level ordinal variable.
Regression coefficient (standardized coefficient) / Regresijski koeficient (standardiziran koeficient)
			Total / Skupaj	Men / Moški	Women / Ženske
Gender (0 = women) /			-0.643***	-	-
Spol (0 = ženski)			(-0.088)		
Age1) /			0.027**	0.030*	0.026
Starost			(0.117)	(0.148)	(0.101)
Household income1) /			-0.103	-0.032	-0.151
Dohodek gospodinjstva			(-0.047)	(-0.017)	(-0.063)
Education (years)1) /			-0.236***	-0.107**	-0.335***
Izobrazba (leta šolanja)			(-0.235)	(-0.117)	(-0.316)
Employment status /					
Zaposlitveni status					
Employed /	n	= 612	Reference category / Referenčna kategor^ija		
Zaposlen					
Unemployed /	n	= 72	0.671	0.614	0.755
Nezaposlen			(0.041)	(0.048)	(0.039)
Schooling /	n	= 121	0.019	-0.444	0.464
Šolajoč se			(0.001)	(-0.040)	(0.032)
Disabled, perm. sick /	n	= 13	4.525**	0.649	6.079***
Nezmožen			(0.096)	(0.014)	(0.131)
Retired / n = 283	0.060	0.339	-0.223
Upokojen	(0.007)	(0.045)	(-0.023)
Housework / n = 148	-0.006	1.791 **	-0.578
Delo v gospodinjstvu	(-0.001)	(0.111)	(-0.058)
Other / n = 16	0.160	0.372	-0.097
Drug status	(0.004)	(0.011)	(-0.002)
Marital status /			
Zakonski stan			
In partnership / n = 863	Reference category / Referenčna kategorija		
V zvezi			
Divorced, n = 81	0.311	0.784	0.062
separated /	(0.022)	(0.050)	(0.005)
Ločen			
Widowed / n = 82	0.714	1.807**	0.332
Vdovel	(0.044)	(0.071)	(0.023)
Never in partnership / n = 239	1.199***	1.758***	0.742
Nikoli v zvezi	(0.126)	(0.235)	(0.063)
Child under 12 years old in the	-0.158	0.125	-0.219
household / Otrok pod 12 let	(-0.018)	(0.016)	(-0.023)
starosti v gospodinjstvu			
(Constant) /	5.502	4.577	7.390
(Konstanta)			
F	8.761***	3.320***	8.446***
Adjusted R2 /	0.118	0.065	0.155
Popravljeni R2			
n	985	448	537
* p < 0.10, ** p < 0.05, *** p < 0.01 1) Centred data / Centrirani podatki
The regression model of socio-demographic variables on all data (for both genders) explains 12% of the variance in depression level. In line with expectations, the depression level is significantly higher for women and decreases with higher education (completed years of education) and increases with age. Work disability significantly increases depression level compared to the employed respondents. It should, however, be noted that the sample size in this group is very small (only 13 respondents in total), therefore the reliability of the results might be questioned. The same kind of influence holds for those never in a partnership compared to the respondents currently living in a partnership. The contributions of the other variables are not statistically significant.
The last two columns of Table 4 present regression models calculated separately for men and women. In order to statistically test for differences in coefficients,
we included interactions between gender and other independent variables in the initial model. The results (not shown in the table) showed significant interaction (p < 0.05) between gender and years of education, work disability and housework. This suggests that gender has a direct impact on depression level, but also moderates the influence of various other socio-demographic factors on depression level, as discussed below. This justifies fitting separate models by gender. As shown in Table 4, the model explains the substantially increased variance in depression for women than for men (16% compared to 7%), leading to the conclusion that socioeconomic and cultural factors more explicitly influence the depression level among women. A comparison of the models by gender indicates some significant and important differences. Although the interaction between gender and age was not significant, age has a marginally significant influence on depression
among men but not among women. While education contributes significantly to lower depression level for both genders, its influence is substantially more prominent for women. The only remaining significant predictor among women is unemployment due to illness or other work disability, which is not so significant for men. As stated above, the small sample size in this group does not allow us to draw reliable conclusions and this finding should be regarded as only an indication. Housework as a primary work activity significantly contributes to higher depression level among men only. Again, these results need to be considered with care, since only 20 men listed housework as their primary work activity. Specific for men are also the significant influences of widowhood and the absence of any partnership in the respondent's life. These, especially the former, have a relatively strong influence on depression level among men, while neither is significant for women. None of these two, however, showed a significant interaction with gender.
4 Discussion
We have acquired a considerable amount of significant data through our research. The analysis has confirmed the thesis, which was also highlighted by Van de Velde et al, that in Slovenia, in contrast to all the other countries included in the ESS research, socioeconomic position is a far more significant factor for the assessment of depression than family position. (5) We will start with the finding that is in contrast to the classical, mostly biological and psychological explanations of depression. If nothing else, this finding confirms the need to apply social models in explaining the gender gap in depression and probably also gender differences in other mental problems. The data shows that the level of depression increases with age when it comes to men, but not women. The connection between age and a poorer mental health with men could be explained by social power and position, which concentrate in the group of older men and can influence their greatest burdens, and consequently their poorer mental condition.
A surprising finding of our analysis is that the inability to work is a much larger risk factor for depression among women than among men. This demonstrates the transference of the role of breadwinner from exclusively men in traditional modernity to both genders, and the fact that women have strongly internalized this new role and form of responsibility. A higher level of depression has been noted among male home-makers (i.e. those,
who deal exclusively with unpaid housework) than among female home-makers. This is in accordance with the findings of others, who detect a higher level of depression among men who take over care for the household (5). Women are still socialized to connect housework with the ethics of care (17), to experience the housework and family tasks in a relational and personal sense, and that they "unquestionably" subordinate their own concerns to care for others. Men are more focused on and tuned to occupation and work issues, which brings them more satisfaction, and they spend less time in the household and are less burdened with concern for children.
A significant finding, and one that we will pay the most attention to in this discussion, is that education is the most important differing factor for the difference in depression assessment between the genders in Slovenia. In general, women in Slovenia do not assess their mental health as poorer than men do. At the same time though, we have noticed the highest and the lowest levels of depression among women. The highest level of depression was recorded among women with the lowest education, and the lowest level of depression was recorded among women with the highest education. We will try to understand this information in the context of emancipation and the role that education plays in the greater independence and autonomy for women in relation to gender structures. Stratification studies and especially women's studies warn that cultural factors play an important role in the mental and physical health of women, especially "gender culture" and "gender order^', "gender regime" in a society (18). Gender culture includes the ruling ideas about the desired and "correct" interrelations in work distribution and role distribution between the genders in a given society. Those ideas are institutionalized as norms, which is why they are relatively stable and present the main point of reference for the everyday behaviour of individuals (19). Gender culture in modern society often exposes women to various discriminatory factors, which have a negative effect on their well-being and consequently on their mental health. Among the primary discriminatory factors are: lack of educational, work and career opportunities (3); the underestimation of women's role and activities (feminized studies, professions and activities are underestimated, poorly evaluated in society and poorly paid (1)); they are overburdened by various social roles (20), they lack social power and have a lower social status (21). All these contribute to the emergence of specific problems with mental health, various expressions of crises among both women and men, and also of various medical
treatments (22). Due to the economic inequalities between genders and the more frequent role of carer for children, the elderly or sick members of the family, women are more often inclined to poverty. Two thirds of adults living in the poorest households are women; a similar percentage of adults that are dependent on social welfare are also women (23). Gender inequality often leads to devaluing the mental health of women and neglecting specific areas of their mental health. The stereotypical stance towards gender also leads to insufficiently diagnosed problems in the mental health of men. Even when men and women share the same symptoms of depression, it is less commonly identified among men than among women (23). In the late modernity, new discriminatory factors that were influenced by the arising consumer culture joined the classical list of discriminatory factors. Several new demands are aimed explicitly at women and additionally burden their self-image and state of being. The first of these demands is related to appearance according to strict bodily regimes, which are harsher towards women, and especially burden those who cannot follow them, e.g. women belonging to a lower socioeconomic class (24). The second demand relates to housekeeping regimes, which are still in the women's domain but reshaped according to media-promoted consumer ideals and spread to the areas of new dietary regimes, upbringing regimes, health regimes and ideals. All these demands trigger a feeling of guilt in those who cannot fulfil or follow these regimes. Further demands encompass the emotional engagement in a family. When a family regime changes from an upbringing community that is based on authority to an emotional community based on protection and support, a woman's emotional burden grows. A special type of demand includes taking care of parents, children and partners: these are demands inscribed in the women's role as a counsellor - therapist - supplier. Women carry the expenses of nursing and taking care of others (25), or are overburdened by the social and emotional problems of others. Those demands can lead to physical, psychological and emotional exhaustion; depression is the result of the combined feeling of being overburdened and feeling that their work is underrated (26). Apart from that, women are often at risk from the repeating experience of violence and abuse, both in childhood and in adulthood. Violence against women, caused by their intimate partners or strangers, is the most prevalent and gender related origin of depression among women (27). From Bourdieu and his perception of cultural capital (12), we assume that higher education enables women to deconstruct and question the gender-
based structures and to personally negotiate the acceptance of prescriptive social roles in the personal and public sphere. A higher education also offers better dispositions for resisting the incomprehensible demands of consumer ideology, in favour of more independent career decisions, more autonomy, better self-image and higher social and economic status. All this results in a more efficient navigation through the traditional and late-modern social roles, a better sense of controlling one's own life and consequently also better mental health.
Lower education results in poorer mental health in women. This is related to many issues, but also to the possibilities that women with a lower education have on the workforce market, in comparison with men with the same education. Women are exposed to precarious jobs more than men (20). A precarious job means all types of uncertain and impermanent jobs where the employees do not have the option to control their work conditions and working time. In the last decades especially, unemployment and poverty among women has been rising steadily (28). Many women compete for poorly paid jobs because of the need for combining housework and family obligations on the one hand, and work in the public sphere on the other. Their so-called dual career results in a lower income and a sustaining dependent position despite working for pay. Therefore, for most women, employment does not bring economic independence. This is why women are more often poor and therefore have a greater risk of experiencing mental distress. The experience of poverty is an important risk factor for mental health problems. Marmot and Wilkinson (29) prove how structural inequality influences the community level, changes relationships such as trust and solidarity, and increases the level of intolerance, which is especially visible in various forms of sexism and racism.
All this leads to the deliberation that the education of women is an important social factor that moderates the economic dependency and poverty of women connected to the gender-based work distribution, the lower position of women on the job market and their position inside the family, which is based on the idea of a woman's work being done because of and for love.
5 Conclusions
We have confirmed the initial assumption that a considerable gender difference in the assessment of depressive symptoms exists in Slovenia; and also that there are considerable differences across both genders,
according to various socio-demographic indicators. The most important factor that differentiates the level of depression between the genders is education. Both the highest and the lowest level of depression has been detected in women. For future research, we suggest empirically testing the complex model that ascertains precisely what influence education has on causing lower levels of depression. We think that education affects the well-being of women in more ways: it improves their social and economic position in the public and private sphere; it equips them with knowledge for deconstructing naturalized demands and has a favourable psychological effect on self-image and self-respect. This effect is also present in men, though their right to choose is culturally more self-evident and therefore does not influence the psychological condition and self-confidence to the same extent. We would like to emphasize that education can only be a temporary protective factor for the first generation of women who entered higher education in the 1980s and gained employment that suits their education. From then on, we can see a rapid increase in higher education among women (19). The role of education has also changed. Research shows that prolonged education, especially among women, is often not the result of personal preferences but a consequence of a lack of options on the workforce market. We presuppose that this influence education has on depression, as we have recorded in our analysis, is a transitional phenomenon that can lose its protective power if the problems of entry into a workforce market increase. It could even have a frustrating effect on well-educated women, who will not get suitable employment.
Literature
1.	Lubotsky Levin B, Becker MA, editors. A public health perspective of women's mental health. New York, London: Springer, 2010.
2.	Kamin T, Jeriček Klanšček H, Zorko M, Roškar S, Dernovšek, MD. Duševno zdravje prebivalcev Slovenije. Ljubljana: Inštitut za varovanje zdravja Republike Slovenije, 2009.
3.	Piccinelli M, Wilkinson G. Gender differences in depression-critical review. Br J Psychiatry 2000; 177: 486-492.
4.	Hopcroft, RL, Bradley DB. The sex difference in depression across 29 countries. Social Forces 2007; 85: 1483-1507.
5.	Van de Velde S, Bracke P, Levecque K. Gender differences in 23 European countries: cross-national variation in the gender gap in depression. Soc Sci Med 2010; 71: 305-313.
6.	Kuechner, C. Gender differences in unipolar depression: an update of epidemiological findings and possible explanations. Acta Psychiatr Scand 2003; 108: 163-174.
7.	Annandale E, Hunt K. Gender inequalities in health. Philadelphia: Open University Press, 2000.
8.	Von dem Knesebeck O, Pattyn E, Bracke P. Education and depressive symptoms in 22 European countries. Int J Public Health 2010; 56: 107-110.
9.	Muntaner C, Borrell C, Benach J, Pasarfn I, Fernandez E. The associations of social class and social stratification with patterns of general and mental health in a Spanish population. Int J Epidemiol 2003; 32: 950-958.
10.	Slovensko javno mnenje. 2009/1.
11.	Stirbu I, Kunst AE, Bopp M, Leinsalu M, Regidor E, Esnaola S. et al. Educational inequalities in avoidable mortality in Europe. J Epidemiol Community Health 2010; 4: 913-920.
12.	Bourdieu P. The forms of capital. In: Halsey AH, Lauder H, Brown P, Stuart Wells A, editors. Education, culture, economy, and society. Oxford: Oxford University Press, 1986/1997: 46-58.
13.	Kamin T, Tivadar B. Kapital(i) in zdravje. Teor Praksa 2011 (forthcoming).
14.	ESS - European social survey: ESS3 2006 documentation report. Bergen: Norwegian Social Science Data Services, 2011.
15.	Jowell R. European social survey 2006/2007: technical report. London: Centre for Comparative Social Surveys, 2007.
16.	Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas 1977; 4: 385-401.
17.	Gilligan C. In a different voice: psychological theory and women's development. Cambridge Ma: Harvard University Press, 1983.
18.	Pfau-Ef^inger B. Gender cultures and the gender arrangement: a theoretical framework for cross-national gender research. Innovation 1998; 11: 147-162.
19.	ULE M. Prikrite spolne neenakosti v izobraževalnih potekih in prehodih. Sodob Pedagog 2010; 61: 30-44.
20.	Leskošek V. et al. Vzroki in obseg pojava zaposlenih revnih: CRP V5-0292 - končno poročilo. Ljubljana: UL, Fakulteta za socialno delo, 2009.
21.	Collins R, Chafetz JS, Blumberg RL, Coltrane S, Turner JH. Toward an integrated theory of gender stratification. Sociol Perspect 1993; 36: 185-216.
22.	Recktenwald J, editor. The life of the women and men in Europe: a statistical portrait. Luxemburg: European Commission Office for Official Publications of the European Communities, 2008.
23.	Shulamit R, Ryan P, Urek M. Attempting to mainstream ethnicity in a multicountry: EU mental health and social inclusion project: lessons for social work'. E J Social Work 2010; 13: 163-182.
24.	Ule M, Šribar R. Položaj in družbene možnosti deklic in deklet v EU : s poudarkom na izobraževanju. Teor praksa 2008; 45: 301-323.
25.	Bracke P, Christiaens W. The pivotal role of woman in informal care. J Fam Issues 2008; 29: 1348-1378.
26.	McLeod E, Bywaters P. Social work, health and equality. London: Routledge, 2000.
27.	Women's mental health: into the mainstream: strategic development of mental health care for women. Department of Health, 2002.
28.	Bloom S. Organizational stress and trauma-informed services. In: Lubotsky B, Levin B, Becker MA, editors. A public health perspective of women's mental health. New York, London: Springer, 2010: 295-312.
29.	Marmot M, Wilkinson R. Psychosocial and material pathways in the relation between income and health. BMJ 2001; 322: 12331236.