SOCIOECONOMIC FACTORS, SEVERITY OF DEPRESSIVE SYMPTOMATOLOGY AND SICKNESS ABSENCE RATE IN THE HUNGARIAN POPULATION

SOCIECONOMIC FACTORS, DEPRESSION AND ABSENCE RATE

Abstract
On the basis of a study of 15.700 persons representative of the active (working and studying) Hungarian population over the age of 16 by age, sex, place of residence and occupation, an analysis was made on the relationships between socioeconomic factors, severity of depressive symptomatology, and sickness absence rate. Sickness absence rate was regarded as a measure of general morbidity rate. Measured by the Shortened Beck Depression Inventory, the severity of depression was found to be very closely connected to working absenteeism. Hierarchical log linear analysis was performed to investigate the interactive effects of socioeconomic factors, severity of depressive symptomatology on absent rate. The material socioeconomic factors such as housing situation, accessibility to a car, own properties have had no direct impact on general morbidity, only mediated by the effect of depression in the more disadvantageous groups of population. All of the measured socioeconomic factors with the exception of place of residence were in close direct connection with depressive symptomatology and according to the hierarchical log linear analysis depression mediates between socioeconomic factors and higher morbidity rate reflected in higher working absent rate in more disadvantageous groups. The persisting socioeconomic differences in health status of the population might be mediated by chronic hopelessness, manifested first of all in depressive symptoms.

Keywords: socioeconomic factors, depressive symptomatology, sickness absence rate, national representative study

Introduction
     An unresolved paradox of the public health studies today is, that in spite of the substantial improvement in the health status of the population, there is a considerable health inequality between the social strata. The socioeconomic situation seems to be the most important risk factor, more important than the traditional factors (1,2). A highly significant inverse association can be found between socioeconomic situation and morbidity and mortality rate in the western countries. This inverse relationship persists even when controlled for health practises such as smoking, obesity and lack of exercise (3-5). The well-developed societies grant adequate food, housing and medical care. Explanations other than material deprivation must be sought to understand the persisting socioeconomic differences in health status of the population.
     These considerations are particularly important in Hungary today, because in Hungary the life expectancy became the lowest and the mortality rate figures the highest among the European countries (6). Since the sixties there was a considerable worsening tendency in the mortality and morbidity statistics in Hungary. While in 1970 the mortality rate of Austria or Great Britain was higher than in Hungary, until the nineties Hungary has reached the worst situation, with 14,0 mortality per 1000 persons (6).
     On the other hand among 1960 and 1988 there was a constant increase in the gross domestic product in Hungary (with a 208% increase compared to the 1960 level) and Hungary was among the best economically situated compared to the neighbouring countries (6), thus the worsening health status of the Hungarian population cannot be explained by worsening material situation.
     The question arises which are the possible important background factors of the catastrophic mortality and morbidity rates in the eighties in Hungary? With the aid of two national representative studies conducted in 1983 and in 1988 we analysed the complex relationships between the health status of the population and the psychological and sociological background factors (7-11).
     Our initial hypothesis was that in modern societies the temporary or lasting failure to cope with the changing psychological, social, and economic situation and the consequent emotional disturbances are more important background factors of working disability than the physical handicaps and causes. According to Murphy et al (12), depression carried a substantial risk for poor health status of the population as a result of longitudinal research in Stirling County, Atlantic Canada.
     We found from the results of our pilot survey in 1983, that the population is not homogeneous as regards neurosis and chronic disorders. This survey showed considerable differences between social strata, with the highest rate of both neurotics at risk and chronic disorders among unskilled workers (7).
     The purpose of the survey conducted in 1988 was to study the interactions of the psychological, social and economic factors in connection with the health status and the reduced capacity for work in the Hungarian population.

Subjects and methods
     Sampling method and the organisation of the national representative survey
     The investigation is based on an analysis of the national representative study of the Hungarian population over the age of 16 according to sex, age and place of residence. The survey was conducted in the form of home interviews in 1988. We interviewed a total of 20,902 persons, among them 15,700 were active workers or students. In the present study we evaluated only the data of the active population.
     Combined selection method was used to form the sample. We used a selection method combining stratified sampling with multi-step sampling. The strata-forming criteria were:
- distribution of the population by sex
- distribution of the population by age
- distribution of the population by region (counties)
- distribution by settlement size (town-village ratio)
- distribution of the active population by occupation.
     On the basis of the determination of stratification, we selected the sample items in a number of steps. In the first step we selected the settlements to be observed from the official register of settlements of Hungary, in such a way that all settlement with a population over 5000 were included in the sample and the method of random selection was used among those with a population of less than 5000. In the second step we selected, in order, the district, street and house of the place of residence within the settlements included in the sample. The interviewer recorded the data of one person in every second household within the residential area indicated. Since the investigation concerned not the household, but an individual within the household, a further sampling was carried out according to sex, age and occupation, on the basis of criteria given in advance.
     The refusal rate was 12% for the full sample, although there were significant differences depending on the settlements. In big cities the refusal rate was much higher than in villages. In the case of refusal the interviewers selected an other person with similar sampling characteristics in the given area.

The questionnaires
     In the interviews, the following groups of questions were asked:
- demographic characteristics (sex, age)
- socioeconomic characteristics :
-father's employment (manager to unskilled worker),
-level of education (university degree to less than eight year of primary school),
-employment status (manager to unskilled worker),
-personal income(in Forint),
-place of residence(town to farm),
-housing situation(without comfort to luxury),
-access to a car,
- own property- summer residence
- Shortened Beck Depression Inventory (13,14,8)
- sickness absence rate in the past year (number of days on sick leave) by disease groups. The sickness absent rate was regarded as a general measure of the morbidity rate of the active population.
     The group of questions concerning the socio-economic situation contained 107 questions, the group concerning the functional emotional disturbances, that is depression, neurosis, health status, social maladjustment and background factors contained 209 questions.

Shortened version of the Beck Depression
Inventory (BDI)
     We analysed all the variables of the BDI (13) by stepwise regression analysis, with the aid of the data of our earlier surveys (8,15). The nine variables showing the closest correlation with the overall depression score were included in the shortened version. The questions of the shortened questionnaire are related to the following characteristics of depression: social withdrawal, inability to make decisions, sleep disorder, fatigability, exaggerated anxiety over somatic complaints, inability to work, pessimism, lack of satisfaction, and the inability to feel pleasure, and self-blame. Our shortened BDI is a modified version of the original shortened BDI (14). The Beck Depression Inventory and its shortened versions are reliable methods of measuring the seriousness of the depressive syndrome (16). It correlates very well with estimates of the seriousness of depression made by psychiatrists.
     On the basis of comparison of BDI and the shortened version we could transform relaibly the sum of the shortened questionnaire into the original score (8). The risk of seriousness of depressive syndrome can be evaluated according to Eaves and Rush (17) as follows:

Questionnaire on health status and sickness absence level
In elaborating the questionnaire the disease groups of the Yearbook of the Ministry for Health and Welfare were applied. The subjects answered two questions for 26 disease groups on whether they had suffered the given illness at some time in their life, and how many days they had spent on sick leave due to the given illness in the past year. Adding up the sick leave for the different disease groups, we calculated the total number of days on sick leave (sickness absence rate) for the whole past year for each subject. This item figures in the multivariate analyses as a general parameter of the mortality rate among the active population.
Statistical methods
     The SPSS statistical program system was used for data analysis. Hierarchical log linear analysis (18) was performed to investigate the interactive effect of social factors, depression, neurosis on working absenteeism. The parameters of the analysis are L= Likelihood ratio, df = degree of freedom, p = probability.

RESULTS
The distribution of depressive symptoms in the Hungarian population
Table 1. shows the prevalence of depressive symptoms in the Hungarian population in 1988.
Figure 1. shows the distribution of depressive symptoms by age and sex.
     24.3 % of the population over 16 complains of symptoms of depression, that is, one quarter of adults and adolescents. 7.5 % of those questioned suffer from moderately severe or severe depressive symptoms, 6.8 % of men and 8.1 % of women.
     According to social strata, symptoms of depression were least characteristic of managers (3.8 % suffer serious or moderately serious depression);among white collar workers not in manager position it was 5.5%, the corresponding proportion was 5.8 % for skilled workers and 10.2 % for unskilled labourers. Among the Beck Depression Inventory items the future was seen as hopeless by 6.4 % of managers, 8.6 % of skilled workers, 14.3 % of unskilled workers.(Table 2)

Sick leave - working absenteeism
     The statistical publications do not give the number of sick persons, but the incidence of diseases, that is, the number of cases. In our data survey we asked each subject about sick leave by groups of diseases.
     The proportion of occurrence of sick leave (working absenteeism) in the last year was 42.9 % for the active population. The sum of the percentages given in the table is higher than that, 73.6%, since individuals were on sick pay due to several different diseases.
     The figures of Table 3, based on the 1988 survey, show the percentages of the absent rates because of the different causes in the past year, by disease groups.
     The occurrence of sickness absenteeism in the last year was most frequent among unskilled workers (48%). It was 45% among the skilled workers, 43% among the non-manager white collar workers and 42% among managers. Exceptions are metabolic diseases other than hepatic diseases, diseases of the blood and blood forming organs, asthma bronchiale and diseases related to pregnancy which occurred most frequently among non-graduate white collar workers, and gastrointestinal diseases, chronic skin diseases and accidents that are most frequent among the skilled workers.
     There was no difference between the occurrence of absenteeism between the male and female workers (43% and 42% correspondingly).

Relationships between socioeconomic characteristics, depression and absent rate.
     Because of the considerable sex differences in depression we analysed the relationships separately for men and women.

Characteristics of men
     Figure 2. shows the relationships between socioeconomic characteristics, severity of depression and sickness absent rate for men by hierarchical log linear analysis in the cases where there was no direct relationship between the socioeconomic factors and morbidity, only mediated by depression. There was a very close direct relationship between depression and sickness absenteeism (L= 149.6, df= 49, p<0.000). According to Figure 2. neither housing condition (a) (convenience of flat),
(b) place of residence,
(c) accessibility to a car, nor own property
(d) showed direct association with working absenteeism.
     On the other hand, there was a very significant direct relationship between the housing condition and depression (L=234.5, df=28, p<0.0000), accessibility to a car and depression (L= 259.6, df= 21, p<0.0000) and own property (summer residence) and depression (L= 41.1, df= 21, p<0.005).
     The place of residence showed no connection with either depression or sickness absent rate among men.
     Figure 3. shows those socioeconomic characteristics among men that were in direct connection with working absent rate, but in each case depression mediated more effectively between the given factors and the working absent rate than these direct connections. The relationship between the given socioeconomic factor and depression was in each case significantly more close, than the direct association between the given socioeconomic factor and the working absenteeism. These socioeconomic factors are:

Characteristics of women
Among women the picture was less clear. The connection between depression and absent rate was also very close among women (L= 179.2, df= 49, p<0.00000). In the case of the overall socioeconomic characteristics of the family, depression mediates between these factors and the absent rate among women too.
Figure 4. There was no direct connection between the housing conditions, accessibility to a car and absent rate, but a highly significant connection can be seen between the housing conditions and depression (L= 216.1, df= 28, p<0.0000), and the accessibility to a car and depression (L= 178.4, df=21, p<0.0000).
     It was a significant direct connection between
own property (summer residence) and absent rate (L= 37.7, df= 21, p<0.01) but the effect mediated by depression was more significant:
summer residence and depression ( L= 43.5, df= 21, p<0.002) (Figure 5)
     In the cases of the father's employment and the level of education there was a significant direct connection with absent rate, but including depression into the analysis, its effect overweighed the direct effects of these socioeconomic factors:

     In the case of women on the contrary to men, the place of residence was significantly connected with absent rate. Including depression into this analysis too, its effect proved to be also important.
     Place of residence and absent rate (L= 49.6, df= 28, p<0.007) place of residence and depression (L= 52,6, df= 28, p<0.003).
     On the other hand, in the group of women in the cases of personal income and the employment status, there were more close direct connections with absent rate than mediated by depression. (Figure 6.)

Discussion
     In the last years depression or vital exhaustion has been found to be risk indicators for coronary heart disease (19-22). Learned helplessness or hopelessness, which can be regarded the best model of depression result in decreased immunological activity that influences the tumour growth (23) and an inverse relationship can be seen between severity of depression and susceptibility to different kinds of infections (24) Self-destructive behaviours such as suicidal behaviour, alcohol abuse, smoking are in very close association with depression (7). Hippius (25) stated, that the considerable proportion of the population suffers from depressive symptoms and complaints that cannot be sufficiently included into the present diagnostic categories. The severity of depressive symptomatology can be regarded as an important measure of mental state of the population. According to our study about one fourth of the active population can be regarded as at risk of mild or more severe depressive smptomatology and 7.5 % as suffering from moderately severe to severe depressive symptomatology.
     According to our national representative study in the active population of Hungary a very close positive correlation was found between working absent rate and severity of depressive symptomatology both among men and women. The absent rate can be regarded as the direct measure of general morbidity of the different origin. It is especially true for men, because the lasting working absenteeism has serious financial consequences. In 1988, the year of our study, there was no unemployment in Hungary, this fact underlines the significance of the absence rate among men. In spite of the fact that most of the women had had a working place in 1988, in the family economy the salary of the women was in many cases of minor importance. The salaries of the women in the similar situations were much lower than the corresponding salaries of the men. Therefore for a woman the working absenteeism could be an intermediate possibility to arrange family problems, not speaking of the child care, which is also part of the absent rate in the case of women. Because of these considerations, the working absent rate of men can be regarded as a direct measure of morbidity rate among men, and only with precautions among women.
     Therefore the interrelationships of socioeconomic factors, depression and general morbidity rate (working absenteeism) among men have special importance. The severity of depressive symptomatology has had a direct negative connection with social class as measured by working activity. These results are in good agreement with the literature (1,3,26). The item analysis of depressive symptomatology has special significance, showing higher level of hopelessness in the lower social strata that can be an important background factor of higher susceptibility of illnesses. All of the measured socioeconomic factors with the exception of place of residence were in close direct connection with higher depressive symptomatology and according to the hierarchical log linear analysis depression mediates between socioeconomic factors and higher morbidity rate reflected in higher working absent rate in more disadvantageous groups. The material socioeconomic characteristics first of all accessibility to car, housing conditions, and own property were not at all connected directly to morbidity, that is absent rate, only mediated by higher depression symptomatology in persons with no car, living in houses without comfort and having no own summer residence. This connection might be two sided. It might be, that persons with lasting mood disorders are both more susceptible to different kinds of disorders, and parallely they are less able to improve their social conditions. An other explanation might be that relatively deprived persons living among undesirable social conditions (such as without car, without comfort) might suffer from constant relative deprivation (12,26,27), might sense helplessness and constant loss of control over their own situation which in long run might result in depressive symptoms. They might regard the future as hopeless - as we found according to the item analysis, they might blame themselves for their low achievements.
     Further socioeconomic situations were partly in direct connection with morbidity but in each case the mediating effect of depression overweighed this direct effect. The father's employment and the level of education have special importance, because especially the first one cannot have two-sided explanations. The father's employment shows very significant connection with depression, that is the family origin determines a considerable part of the later risk for depression and thus higher morbidity rate. In the case of education it is possible to hypothesise that persons with depressive personality have a lower chance for higher education. In this case also more probable, that depressive symptoms can arise as a result of frustration caused by inability to learn. This hypothesis can be supported by the results of the same study evaluating by multiple regression analysis the background factors of depression. According to this analysis an important background factor of depressive symptoms is, that the persons wanted to learn further but were not able to do so (9). Among men the low level of education shows among the not material socioeconomic factors the most significant connection with depression.
     Both the personal income and the employment status are more closely connected to depression than to morbidity rate directly.
     The connection between morbidity (absent) rate and depression seems to be even more important among women than among men. Among women the low level of education shows the closest connection with depression among the socioeconomic factors, the second most important factor was the housing condition, that is the lack of comfort of the lodging.
     The socioeconomic situation of women is more closely connected to the overall situation of the family (Such as own property, accessibility to car) which are related to depression and through depression to higher morbidity rate among women too. In contrast to men, among women, their own income or employment situation were in direct connection with absent rate without being in connection with depression. Among women with low income and low employment situation the absence rate might be a result of rational considerations to remain at home for a while with children or to solve other family problems such as the care of an elderly relative. In these cases the absent rate might be an index of the cost-benefit analysis of the family, because the financial consequences of the absenteeism have no serious implications in the majority of women living in family.
     These results indicate that severity of depressive state and symptoms can be regarded serious risk factors of general morbidity and the majority of risk consequences of low socioeconomic situation are mediated by depressive symptomatology, that is by hopelessness, self-blame, pessimism, working disability, lack of satisfaction. It is an important task of future preventive interventions to take into consideration depression as a fundamental psychosocial risk factor.

Table 1.
Prevalence of depressive symptoms in the Hungarian population in 1988

Table 3. Percentage of absent rates because of different causes in the last year

References
1. Syme,S.L., Berkman,L.F.: Social class, susceptibility and sickness. Am. J. Epidemiology, 1976, 104,1-8.
2. Marmot,M.G., Kogevinas, M., Elston, M.A.: Social/economic status and disease. Annual Rev., Public Health, 1987, 8,111- 135.
3. Scherwitz,L., Perkins,L., Chesney,M., Hughes,G.: Cook-Medley Hostility Scale and Subsets: Relationship to Demographic and Psychosocial Characteristics in young adults in the CARDIA study, Psychosomatic Medicine, 1991, 53,36-49.;
4. Slater,C.H., Lorimer,R.J., Larison,D.R.: The independent contribution of socoeconomic status and health practices to health status. Prev.Med., 1985, 14,372-373.
5. Wiley,J.A., Comacho,T.C.: Life-style and future health: Evidence from the Alameda county study. Prev. Med., 1986, 9,1-21.
6. Statistical Yearbook of Hungary, Budapest, 1993.
7. Kopp,M.S., Skrabski,Á., Magyar I.: Neurotics at risk and suicidal behaviour in the Hungarian population, Acta Psychiatrica Scandinavica, 1987, 76,406-413.
8. Kopp,M.S., Skrabski,Á.: Összehasonlító mentálhigiénés vizsgálatokhoz ajánlott módszertan. (Methodology of comparative mental health studies) Végeken, 1990, 2,2,4-24 (in Hungarian)
9. Kopp,M.S., Skrabski, Á.: Magyar lelkiállapot (Hungarian state of mind) Végeken Publ.Budapest (in Hungarian), 1992.
10. Skrabski,Á., Kopp,M.S.: Needs, decrease of ability to work and disorders of social adaptation. Res., Rev. Hungarian Soc. Sci., 1989, 71-88.
11. Skrabski,Á., Kopp,M.S.: Health behaviour, psychiatric symptoms and psychosocial background factors. In Dauwalder (ed.), Swiss Monographs in Psychology, 1994, Vol. 2: 21-27.
12. Murphy,J.M., Olivier,D.C., Monson,P.R., Sobol,A.M., Federman,E.B., Leighton,A.H. Depression and anxiety in relation to social status. Arch. Gen. Psychiatry, 1991, 48: 223-229.
13. Beck,A.T., Ward,C.H., Mendelsohn,M., Mosk,J., Erbaugh,J.: An inventory for measuring depression, Archives of General Psychiatry, 1961, 4,561-571
14. Beck,A.T., Beck,R.W.: Shortened version of BDI. Post. Grad.Med., 1972, 52,81-85.
15. Kopp, M.S.: Clinical psychophysiology, Psychophysomatic booklets,2,1-35.(in Hungarian), 1985.
17. Eaves,G., Rush,A.J.: Cognitive pattern in symptomatic and remitted unipolar major depression, Abnorm. Psychol., 1984, 93, 1, 31-40.
18. Szedmák,S., Kopp,M.S., Skrabski,Á.: Mathematical Models for Complex Non-linear Systems, Semmelweis Science Forum, Budapest, 1994.
19. Appels,A.: The year before myocardial infarction In: Biobehavioural bases of coronary heart disease (Eds Dembroski, T.M., Smidt,-t..H., Blumchen,G.) Karger, Basel, 1983.
20. Falger,P., Appels,A.: Psychological risk factors over the life course of myocardial infarction patients. Advances in Cardiology, 1982, 29, 132-139.
21. Freedland,K.E., Carney,R.M.: Depression as a risk factor for coronary heart disease, Proc.Third Internat. Congr. of Behavioural Medicine,Amsterdam, 1994
22. Williams,R.: Hostility, depression and CHD: a common biological mechanism? Proc. Third Internat. Congr. of Behavioural Medicine, Amsterdam, 1994.
23. Sklar,L. and Anisman,H.: Stress and coping factors in fluence tumour growth. Science, 1979, 205, 513-515.
24. Lázár,I.: Psychoneuroimmunology, Végeken Publ., Budapest, (in Hungarian), 1991.
25. Hippius: Medikamentöse Therapie der Panik-Reaction. Proc of 14th Danube Symposium of Psychiatry, Budapest, 1990
26. Marmot,M.G.: Relative deprivation as a psychosocial concept,Proc. Third Internat. congress of Behavioural Medicine, Amsterdam, 1994.
27. Johnson, J.V.: Work experience over the life course: the influence of occupational socialization, exposure intensity and career directionality on cardiovascular disease, Proc.Third Internat. congress of Behavioural Sciences,Amsterdam, 1994.