Socioeconomic inequalities and non-communicable diseases in Serbia : national health survey

Background/Aim. Non-communicable diseases (NCDs) are a major public health challenge worldwide. Although they are preventable, NCDs are the major global causes of morbidity and mortality, absenteeism, disability and premature death. The aim of this study was to examine socioeconomic inequalities in the prevalence of non-communicable diseases in Serbia. Methods. Data from the 2013 National Health Survey of the population of Serbia was used in this study. There were 13,765 adults interviewed, aged ≥ 20 years. We used multivariate logistic regression analyses with demographic and socioeconomic determinants of health as independent variables and prevalence of non-communicable diseases as a dependent variable. The minimum level of significance was p < 0.05. Results. Hypertension was the most prevalent NCDs (36.1%). The prevalence of multimorbidity was 47.1%. Multivariate logistic regression analysis showed that gender, age, place of residence, employment status and education were associated with the presence of NCDs. The odds ratio (OR) for age was 1.074 [95% confidence interval (CI) : 1.070–1.077). Women were at a higher risk of NCDs by 58.9% when compared to men (OR = 1.589; 95% 95% CI : 1.467–1.726). Respondents who lived in the rural areas were at a higher risk for NCDs by 14.1% compared to those who lived in urban areas (OR = 1.141; 95% CI : 1.047– 1.244). Odds ratio for unemployment was 1.227 (95% CI: 1.118–1.346). Respondents with primary education were at a higher risk for chronic diseases by 47.1% (OR = 1.471; 95% CI : 1.281–1.687) while those with secondary school were at a higher risk by 27.7% (OR = 1.277; 95% CI : 1.142–1.428) compared to respondents who had higher education. When it comes to Wealth Index, univariate logistic regression analysis showed that respondents who belonged to the poor and middle classes were at a higher risk for NCDs (OR = 2.031; 95% CI : 1.819–2.267; OR = 1.473; 95% CI : 1.343– 1.615) compared to respondents who belonged to the rich class. Multivariate logistic regression analysis did not show statistically significant correlations between the Wealth index and NCDs. Conclusion. Socioeconomic inequalities in health status are the major challenge and should be a target of national health policy in Serbia, not only because they represent social injustice but also because solving the health problems of underprivileged groups of the population can influence improvement of health status of the population as a whole.

who lived in urban areas (OR = 1.141; 95% CI : 1.047-1.244).Odds ratio for unemployment was 1.227 (95% CI: 1.118-1.346).Respondents with primary education were at a higher risk for chronic diseases by 47.1% (OR = 1.471; 95% CI : 1.281-1.687)while those with secondary school were at a higher risk by 27.7% (OR = 1.277; 95% CI : 1.142-1.428)compared to respondents who had higher education.When it comes to Wealth Index, univariate logistic regression analysis showed that respondents who belonged to the poor and middle classes were at a higher risk for NCDs (OR = 2.031; 95% CI : 1.819-2.267;OR = 1.473; 95% CI : 1.343-1.615)compared to respondents who belonged to the rich class.Multivariate logistic regression analysis did not show statistically significant correlations between the Wealth index and NCDs.Conclusion.Socioeconomic inequalities in health status are the major challenge and should be a target of national health policy in Serbia, not only because they represent social injustice but also because solving the health problems of underprivileged groups of the population can influence improvement of health status of the population as a whole.

Introduction
Non-communicable diseases (NCDs) are a major public health challenge worldwide 1 .Although they are preventable, NCDs (mainly cardiovascular diseases, neoplasms, chronic respiratory diseases and diabetes) are the major global causes of morbidity and mortality, absenteeism, disability and premature death 2 .According to the World Health Organization (WHO) report, more than 36 million people die from NCDs annually which is about 63% of the 57 million global deaths 3 .About 16 million people die prematurely, before the age of 70, during the most productive period of life.The prevalence and impact of NCDs continues to rise, especially in countries with low and middle income that already participate in premature deaths with 86% 4 .NCDs have serious socioeconomic consequences, ranging from increasing individual and household impoverishment to high cost of healthcare which hinder social and economic development 5 .The rapidly growing number of NCDs is under the influence of many factors including population aging, negative effects of globalization, rapid and unplanned urbanization and unhealthy behaviours 6 .
Individual characteristics and socioeconomic status are important determinants of health inequalities 7 .The impact of these determinants on the morbidity has been studied in many countries, and the results showed clear correlations between socioeconomic determinants and health status of respondents 1,8,9 .In all countries, there are significant differences in health between socioeconomic groups.People with lower socioeconomic status (SES) are associated with higher prevalence of chronic diseases and injuries, unhealthy behaviors such as smoking, inadequate diet, alcohol use, and lack of physical exercise [10][11][12] .People with lower level of education, lower occupational class, or lower income live shorter in good health, have higher rates of mortality and die at younger age 10,13 .There is substantial evidence that those with lower socioeconomic status are at increased risk of adverse health outcomes, including cardiovascular disease, neoplasms and mental health problems [14][15][16][17] .
The presence of diseases and their symptoms are indicators of health status of the population.In health research, self-reporting of diseases is widely used.Information on the prevalence of NCDs in the population is most commonly obtained through questionnaires 3 .
The aim of this study was to examine socioeconomic inequalities in the prevalence of NCDs in Serbia, using the data from the 2013 National Health Survey.

Study population and sample
This study used the data from the 2013 National Health Surveys for Serbia.The study was designed as a cross-sectional study on a representative probability sample of the population aged 15 years and above.The survey was conducted in accordance with the methodology and instruments of the European Health Interview Survey wave 2 (EHISwave 2).It was implemented by the Ministry of Health of the Republic of Serbia.
The sample consisted of all households listed by all enumeration areas of Census 2013.The mechanism used to generate a random sample of households and respondents is a combination of two sampling techniques: stratification and multi-stage sampling.Two-stage stratified sample of the population of the Republic of Serbia was selected in order to obtain a statistically reliable estimation of indicators which affect the health of the population at the national level and at the levels of four geographical regions of Serbia (Vojvodina, Belgrade, Šumadija, Western Serbia and South-Eastern Serbia) 18 .
Of the total number of 10,089 households contacted, 6,500 of them agreed to participate in the survey, so that the response rate of households was 64.4%.Of the total number of 16,474 registered household members aged 15 years and more, 14,623 of them agreed to be interviewed, giving a response rate of 88.9%.For the purposes of this study, we analyzed data related to respondents aged 20 years and older (13,765 interviewed respondents).

Instruments
Data on demographic and socioeconomic characteristics of the respondents and their own health assessment was ob- tained through a face-to-face interview carried out at home, while information about the wealth level of the household was obtained by means of a household questionnaire.The questions were validated instruments and based on the standard questionnaires from similar types of surveys.Ethical Standards in this study are in compliance with the international ethical standards (the World Medical Association Declaration of Helsinki) and the specific legislation of our country.
As independent variables, we used demographic characteristics (age, gender, type of settlement and marital status) and socioeconomic status (education, employment status and Wealth index).The age of participants was categorized into eight age groups (≤ 24 years, 25-34 years, 35-44 years, 45-54 years, 55-64 years, 65-74 years, 75-84 years and 85 years or more).Gender is coded as male and female, place of residence as urban and rural while the marital status was categorized as married/living with a partner and not married, divorced/widowed.Education was defined as high level (university degree), medium level (three of four years of secondary school), or low level (no education, incomplete primary school, or primary school).Employment status was categorized as employed and unemployed (including economically inactive people: pensioners, people attending some form of education, housewives who are inactive due to family reasons, people who are ill, unable to work or elderly and other inactive categories).The Wealth Index was based on household assets and housing characteristics, such as (the number of bedrooms per household member, materials used in the construction of a floor, roof, and walls, type of drinking water source and sanitation facilities, fuel used for heating, color TV set, mobile phone, refrigerator, washing machine, dish washer, personal computer, air conditioning, central heating, car and access to the Internet).According to the Wealth Index, households were divided into five equal-sized groups (quintiles): the poorest (Q1), poorer (Q2), middle (Q3), richer (Q4) and the richest (Q5).For the purposes of this paper, the respondents were classified into three socioeconomic categories: poor class, middle class and rich class.
Self-reported diagnosis of NCDs was taken as a dependent variable.The respondents were asked: "Did you have some of the following diseases or conditions in the last twelve months: hypertension, deformity of lower spine or other chronic back problems, hyperlipidemia, neck deformity or another chronic problem with cervical spine, coronary heart disease or angina pectoris, arthrosis, allergies (excluding allergic asthma), diabetes and depression?" Of the total seventeen NCDs reported in the National Health Survey, we considered nine to be major NCDs.All diseases were coded as dummy variables (yes/no).Multimorbidity was used to refer to those who had two or more morbidities.

Statistical analysis
All data of interest were presented and analyzed by adequate mathematical-statistical methods appropriate for the data type.The χ 2 test was used to compare proportions between the groups.The t-test was used to compare continuous variables between the groups.The relations between the presence of NCDs, as a dependent variable and a set of independent variables was examined by univariate and multivariate logistic regression.Univariate logistic regression models were used to examine the associations between potential factors and the presence of NCDs.The variables that were statistically significant (p < 0.05) were further examined in multivariable logistic regression.The unadjusted odds ratios (ORs) with their corresponding 95% confidence intervals (95% CI) were also obtained.All statistical calculations were performed using the commercial, standard software package SPSS, version 18.0.[The Statistical Package for Social Sciences software (SPSS Inc, version 18.0, Chicago, IL)].

Results
The study included 13,765 respondents, where 46% were men and 54% were women.The demographic and socioeconomic characteristics of the sample of respondents are presented in Table 1.The mean age of the respondents was 51.78 (SD = 17.467); there were 20.9% in the 55-64 age group.The majority of respondents resided in urban areas (56.4%), 65.4% of them were married or lived with a partner, 54.2% had a middle level of education, 67.2% were unemployed (and belonged to the middle class (60.1%) according to the Wealth Index.
In the present study, 60% reported at least one of nine analyzed chronic diseases.Table 2 shows the prevalence of NCDs by demographic and socioeconomic variables.The most prevalent self-reported NCDs across the sample were hypertension (36.1%) followed by deformity of the lower spine or other chronic problems with the back (21.5%) and hyperlipidemia (14.7%).Allergy (8.9%) and depression (7.2%) were the least reported NCDs in the study sample.All diseases were more prevalent among women than among men.Significant differences by age were found for all diseases.The difference in mean values of years of life between respondents with chronic diseases and respondents without them is statistically significant (t = -69,635, p < 0.0005).The respondents with a chronic disease were 59.45, (SD = 14.94) years on average and those without chronic diseases were on average at the age of 41.26 (SD = 15.01)years.
Two fifths (40%) of the respondents reported that they did not have any NCDs, while 12.9% had one NCDs.More than two NCDs was reported by 47.1% of the respondents (Table 3).The median number of diseases among persons with multimorbidity was 3, ranging from 2 to 10.The number of NCDs differed significantly by gender, age groups, marital status, education, employment status and Wealth index (p < 0.005).The number of NCDs did not differ significantly by type of settlement (p = 0.412).Women, elderly people, those with low educational level, unemployed, inactive and respondents who belonged to the poor class were sensitive to multimorbidity.
Results of the univariate and multivariate logistic regression analysis are presented in Table 4. Univariate logistic regression analysis showed that the prevalence of NCDs differed significantly across gender, age, marital status, education, place of residence, employment status, education and the Wealth Index.Multivariate logistic regression analysis showed that gender, age, place of residence, employment status and education can be associated with the presence of NCDs.The prevalence of NCDs increased with age.Number of chronic diseases was positively correlated with age (r = 0.557; p < 0.0005).The OR for age was 1.074 (95% CI : 1.070-1.077),which means that each year the risk of a chronic disease was increasing by 7.4%.Women were under the increased risk of a chronic disease by 58.9% compared to men (OR = 1.589; 95% CI : 1.467 to 1.726).Respondents who lived in rural areas were under the higher risk of chronic non-communicable diseases by 14.1% compared to those who lived in urban areas (OR = 1.141; 95% CI : 1.047-1.244).Odds ratio for unemployment was 1.227 (95% CI : 1.118-1.346).Unemployed and inactive people were under the increased risk of a chronic disease by 22.7% in relation to the employed.(OR = 1.227; 95% CI: 1.118-1.346).The prevalence of a chronic diseasewas inversely proportional to the level of education.Respondents with primary education were under the higher risk for NCDs by 47.1% (OR = 1.471; 95% CI : 1.281-1.687)while those with secondary school were under the higher risk by 27.7% of (OR = 1.277, 95% CI : 1.142-1.428)compared to the respondents who had higher education.When it comes to the Wealth index, univariate logistic regression analysis showed that the respondents who belonged to the poor and middle class were in a higher risk for NCDs (OR = 2.031; 95% CI : 1.819-2.267;OR = 1.473; 95% CI : 1.343-1.615)compared to the respondents who belonged to the rich class.Multivariate logistic regression analysis did not show a statistically significant impact of the Wealth Index on the prevalence of NCDs.

Discussion
The results showed that there were significant differences in the prevalence of NCDs, depending on the demographic and socioeconomic variables.
This study presented a high prevalence of NCDs.Similar results were found in a population survey conducted in eight countries in Europe where 55.1% of the adult population had at least one chronic disease 19 .National Population Health Survey in Canada found that more than a half of adults aged 25 years and over had a chronic condition 20 .
The prevalence of multimorbidity was 47.1% which is similar to that found in the other studies (Banglades -53.7%, Germany -58.6%, Sweden -55%) [21][22][23] .Opposite to our data some studies found a smaller proportion of multimorbidity (Canada -24%, South Africa -22.5%) 24,25 .Many studies reported variable prevalence of multimorbidity, which is likely due to a sample type, source of data, data collection method, observed age groups, diagnoses that were considered and study population 26 .

Men
< 0.0005 Emplyment status employed 2,583 ( Low SES means greater prevalence of almost all diseases.Systematic reviews by Sommer et al. 27 showed that low SES increased the risk of cardiovascular diseases, lung and gastric cancer, type 2 diabetes and chronic obstructive pulmonary disease, etc.The famous Whitehall study 28 conducted among state officials in the UK showed that socioeconomic inequalities were reflected in various diseases such as heart diseases, some forms of malignant diseases, chronic lung disease, gastrointestinal disease, depression, suicide and back pain.Socioeconomic inequalities in NCDs were associated with the unequal distribution of behavioural risk factors, particularly smoking, alcohol use, an unhealthy diet and a sedentary lifestyle 29 .
Similar to other studies, increasing age, being female, having low educational level, being unemployed, inactive and belonging to the poor class were characteristics associated with presence of chronic diseases 21,25,30 .
Demographic characteristics of the respondents such as gender and age are important predictors of health.There is a positive correlation between them and the morbidity.Our analysis revealed that the prevalence of NCDs was increasing with age.In other words, morbidity prevalence increased with age 31 .Most NCDs were found to be prevalent among women compared to men 32,33 .These findings repeated in many studies focusing on poorer health of women and the elderly [34][35][36][37] , indicating that special attention should be focused on the care of these vulnerable groups, as they were more likely to develop morbidity.
Education is a very important socioeconomic determinant of health.Men and women with lower levels of education had higher morbidity rates and premature mortality from all causes than their higher-educated counterparts 33 .People with higher levels of education were more likely to be employed, to have a higher social status and more stable income and they have more skills to cope with and overcome everyday life difficulties which could negatively affect their health 37 .
In this study we found that unemployed people had poorer health status compared to those who were employed.This could be explained by the fact that a large part of the unemployed were economically inactive, the largest number of them being pensioners who were more frequently physically inactive and more often suffered from hypertension.The second large group of economically inactive persons were housewives.It is generally assumed that female workers had better health than full-time housewives 35 .
The studies in several countries showed that the unemployed and their families had poorer health and were under the significantly higher risk of premature death than the employed 36 .The health consequences of unemployment were associated with both psychological and financial consequences of unemployment, such as the inability to satisfy everyday needs, insecurity and lack of self-esteem.The unemployed used health services to a lesser extent, and rarely visited a doctor of general medicine/occupational medicine and got hospital treatment 37 .
In the current study, univariate logistic regression analysis showed that the respondents who belonged to the poor and middle class were in a higher risk for NCDs compared to the respondents who belonged to the rich class.
Multivariate logistic regression analysis did not show a statistically significant impact of the Wealth Index on the prevalence of NCDs.We found that many NCDs were mostly concentrated among the poor and middle class, and were inversely associated with decreasing wealth level.Our findings were in accordance with previous reports about the presence of angina, arthritis, asthma, depression, gastritis and migraine.However, cancer, allergy and diabetes mellitus were slightly more concentrated among wealthy individuals 30 .This results could be explained by the fact that people who lived in poverty may experience material deprivation and high stress levels, which may lead to constrained choices and a higher likelihood of engaging in risky health behaviours, increasing the risk of disease, following disease onset and reduced access to healthcare hindered opportunities to prevent complications 38 .
Unmarried, divorced or widowed people had higher values of the morbidity index and often assessed their health worse than those who were married 39 .

Strengths and limitations
The main strength of this study is the large sample which is representative of the adult population of the Republic of Serbia (aged ≥ 20 years).However, this study has several limitations.First, the cross-sectional design of the study makes it difficult to judge causal relations.This limitation can be overcome with the use of longitudinal data, which might better explain the changes in socioeconomic status and their impact on NCDs.Another limitation of the study is the problem of the accuracy of self-reports for chronic diseases, which may be subject to bias.Differences in prevalence rates between self-reported diagnoses and standardized measure may vary.The accuracy of self-reports for NCDs depends on different factors such as knowledge of the health problem, consequences on everyday life, willingness to report the problem and frequency of visits to healthcare services.According to the current literature, individuals with lower socioeconomic status tend to under-report symptoms which might result in an underestimation of the presence of NCDs.
Regardless of the above mentioned shortcomings of the methodical approach and a lack of a better, but much more expensive study design, this study, if repeated with the same methodical approach in a period of several years, could make us able to estimate the trends of diseases and their relationships to health determinants.That is why it is important to promote the repetition of similar studies on the health of the population in Serbia.

Conclusion
The elderly, females, those with lower levels of education and unemployed people have a greater prevalence of chronic diseases.Socioeconomic inequalities in health are the major challenge for health policy, not only because they represent social injustice but also because solving the health problems of underprivileged groups of the population can influence the improvement of the health status of the population as a whole.