Regional patterns in cause-specific mortality in Montenegro, 1991–2019

2.1 Study area

The Republic of Montenegro is located in the southern part of Europe. Its land area extends between 41°50′26″ and 43°33′23″ northern latitude and 18°26′00″ and 20°21′42″ eastern longitude. According to the Census 2023, Montenegro had a population of 623,633 [16].

Analysis in this article is based on mortality trends in the three main regions in the country: northern, central, and coastal (Figure 1). These regions are defined by the National Law on regional development and are adopted in the Spatial Plan of Montenegro.^[1] According to this document, the northern region comprises the following municipalities: Andrijevica, Berane, Belo Polje, Kolašin, Mojkovac, Plav, Pljevlja, Plužine, Rozaje, Savnik, Zabalj, Petnjica i Gusinje. The central region includes: Podgorica, Cetinje, Danilovgrad, Nikšic, Tuzi i Zeta. The coastal includes: Bar, Budva, Herceg-Novi, Kotor, Tivat, and Ulcinj. Each of these regions has specific demographic, socioeconomic, social, and environmental features. The central region is the most settled area with around 300,000 inhabitants, while two other regions count significantly lower numbers of population (Table 1).

Figure 1

The location of Montenegro and the position of the regions (source: base map taken from https://www.arcgis.com/home/webmap/viewer.html?layers=10df2279f9684e4a9f6a7f08febac2a9, edited by the author in ArcGis).

2.2 Data and methods

Mortality data were obtained from the Statistical Office of Montenegro (MONSTAT) and the Institute of Public Health of Montenegro. Data were provided by cause of death, for the period 1991–2019, and the specific causes of death were coded as per the International Classification of Diseases (ICD-10). Analyses were focused on three leading causes of death: circulatory diseases (I00–I99), neoplasms (C00–D48), and ill-defined causes (R00–R99). Ill-defined diseases include causes from symptoms, signs, and abnormal clinical and laboratory findings not elsewhere classified. Analysis was performed using crude death rates by cause of death for three regions in Montenegro. The total population in each region was derived from the censuses 1991 and 2003 and an estimation for 2019 of the Statistical Office of Montenegro.

Data related to the health (doctors and hospital beds per 10,000 inhabitants) and socio-economic indicators (GINI coefficient and HDI) were obtained from the database of the World Health Organization.^[2],^[3],^[4],^[5] The GINI coefficient is an indicator used to measure income inequality within a country. A lower value of the GINI coefficient indicates a more equitable distribution of income, while a higher value suggests greater inequality. Human Development Index (HDI) is an indicator created using multiple data (life expectancy, education, and per capita income) for assessing the overall level of development in a country. Data about doctors and hospital beds were available for the period 2000–2019, while the GINI coefficient and HDI were observed for the period 2003–2015.

All data were organized in a specially designed database, and statistical analysis was conducted using STATISTICA software, version 14.1.0.8. This study employed a structured three-stage analytical approach to investigate regional patterns and determinants of cause-specific mortality in Montenegro. In the first phase, crude death rates were calculated, and descriptive statistics with trend analysis using linear regression were conducted for each region and cause of death to identify general mortality trends across the country. The second phase involved statistical testing, using the t-test to assess whether differences in mortality rates between regions were statistically significant for each specific cause of death. In the third phase, multiple linear regression analysis was performed to examine determinants of mortality patterns. The regression analysis explored the relationship between mortality rate and selected health and socio-economic indicators.

3.1 Regional trends in cause-specific mortality

Figure 2 shows the general outlook of regional mortality trends by leading causes of death in Montenegro over the last three decades. The three main causes of death are circulatory diseases, neoplasms, and ill-defined causes. Circulatory diseases are responsible for about 42% of all deaths, neoplasms for 17%, and ill-defined causes have a share of about 15%. All other causes contributed to mortality, with approximately 25%. During the analyzed period, from 1991 to 2019, the average mortality rate from circulatory diseases was 451.6‰ in the coastal region, 387.6‰ in the northern region, and 344.1‰ in the central region. Average mortality rate from neoplasms was 168.2‰ in the central region, followed by the coastal (158.6‰) and northern (127‰) areas. And average mortality rate caused by ill-defined diseases was the highest in the northern (181.4‰) part of the country, subsequently 134.7‰ in the central region and 68.4‰ in the coastal region (Table S1).

Figure 2

Mortality from leading causes of death by region in Montenegro, 1991–2019. (a) Circulatory diseases (I00–I99), (b) neoplasms (C00–D48), and (c) ill-defined diseases (R00–R99).

Results of trend analysis for each region and cause of death confirmed a statistically significant trend (p < 0.05), particularly for mortality caused by neoplasms. Mortality from neoplasms is rising significantly in the northern (5.3‰ per year, p = 0.0000) and central (4.4‰ per year, p = 0.0000) regions, and in the coastal area, a slight increase of 0.5‰ was confirmed, but it was not statistically significant (p = 0.156294159). Mortality from circulatory diseases is increasing in the northern region (10.2‰ per year, p = 0.000142) but decreasing in the coastal region (6.8‰, p = 0.00095), while the central region experienced a slight decline (0.26‰ per year, p = 0.902247) not statistically significant. Causes from ill-defined diseases are declining across all regions, principally in the northern area (9.6 per year, ‰, p = 0.00000), followed by the coastal (2.9‰, p = 0.00000) and central regions (1.0‰, p = 0.207549) (Table S2).

Furthermore, analysis of t-test confirmed statistically significant differences (p < 0.05) in death rates between almost all regions for each cause of death. In terms of mortality caused by circulatory diseases, the coastal region is significantly different from both northern (t = 2.07147, p = 0.04293) and central (t = 4.24479, p = 0.00008) areas, while between the central and northern regions, anaysis did not confirmed statistically significant differences (t = −1.43508, p = 0.156825). Results for mortality from neoplasms confirmed that the northern part differs significantly from both central (t = 3.49128, p = 0.00095) and coastal (t = 3.26214, p = 0.00189), while central and coastal areas are statistically similar (t = −1.17954, p = 0.243171). All regions differ significantly from each other in ill-defined causes, respectively, coastal are differs from northern (t = −5.85857, p = 0.000000) and central (t = −7.3276, p = 0.000000) region, but central and northern are also confirmed statistical differences (t = −2.4078, p = 0.019370), (Table S3).

3.2 Analysis of health and socio-economic factors associated with mortality

Mortality is known to be a highly complex demographic and public health outcome, and indicators like cause-specific mortality rates could be determined by various factors. Using regression analysis, this article explored two health (number of doctors and hospital beds per 10,000 inhabitants) and two socio-economic (GINI coefficient and HDI) risk factors. In model analysis, the mortality rate was set as the dependent variable, while doctors and hospital beds per 10,000 inhabitants, GINI coefficient, and HDI are employed as independent variables. Data for health and socio-economic indicators were limited at the country level, which may mitigate the association. Results for health indicators show that higher numbers of doctors are significantly associated with higher mortality rates from neoplasms in the central region (b = 7.21542, p = 0.000879) and lower mortality rates from circulatory diseases in the northern (b = −14.365, p = 0.020782) and coastal (b = −20.397, p = 0.008536) areas. Changes in the mortality rate caused by ill-defined diseases were not associated with the number of doctors. Data about hospital beds did not show a statistically significant relationship for any cause of death or region (Table S4). The analysis of socio-economic variables reveals a statistically significant relationship between the GINI coefficient and mortality rates only in the coastal region, where a higher GINI is associated with lower mortality rates from circulatory diseases (b = −9.93, p = 0.030344), while HDI shows varying associations across regions. In the northern region, higher HDI is significantly related to higher mortality from neoplasms (b = 1319.140, p = 0.000296), but lower mortality rates from circulatory diseases in the central (b = −2948.99, p = 0.005718) and coastal (b = −3375.11, p = 0.000558) regions, as well as lower mortality rate for ill-defined causes in the northern area (b = −2587.07, p = 0.007782) (Table S5).

4.1 Leading findings

This study analyzed regional trends and patterns of specific-cause mortality in Montenegro, focusing on three leading causes of death: circulatory diseases, neoplasms, and ill-defined conditions.

Circulatory diseases remain the leading cause of death, and regarding the regional pattern, a south–north gradient was found with significant differences in mortality rate due to circulatory diseases in the coastal region compared to the central and northern areas. Similar results were confirmed for mortality caused by neoplasms with south–north pattern, while for ill-defined diseases, a reversed slope north–south has been observed.

Related regional patterns have also been identified in other countries, such as Belarus, Germany, and Ukraine, as well as in the German-speaking area of Europe (Germany, Austria, and Switzerland), where north–south, west–east, and east–west gradients have been documented [9,17,18,19,20,21]. In Belarus and Ukraine, regional analyses have particularly highlighted inequalities in mortality from CVDs and external causes, including alcohol-related and injury-related deaths [17,21]. A study focused on Ukraine reveals that differences between the Western regions compared to the eastern and southestern parts of the country are likely influenced by local socioeconomic conditions, healthcare infrastructure, and cultural factors [21].

The study by Grigoriev et al. [19] discloses that smoking is a major factor behind regional mortality differences in Germany, while avoidable mortality has primarily contributed to the north–south pattern across the German-speaking area [20].

Wider regional disparities are evident across Europe, with east–west divides. Countries in western Europe generally have lower rates due to earlier adoption of preventive strategies, better control of risk factors (like hypertension and cholesterol), and more effective healthcare systems [22].

The increasing mortality rate due to neoplasms, particularly in the northern and central regions of Montenegro, aligns with persistent disparities across Europe, too. Countries in the Eastern European area continue to experience higher cancer mortality rates compared to Western Europe, primarily due to postponed implementation of screening programs, limited access to advanced treatments, and higher prevalence of risk factors such as tobacco and alcohol consumption [23]. Although this study did not analyze neoplasms subgroups, recent findings confirm that lung cancer remains the leading cause of cancer deaths, with Eastern Europe exhibiting higher mortality rates [24]. In Montenegro, according to the Global Burden of Disease (GBD) data, lung cancer is the leading cause of neoplasm-related deaths, with tobacco use being a major contributing factor.

This study’s analysis of factors associated with mortality trends and changes in mortality rate in Montenegro included health and socio-economic variables: number of doctors and hospital beds, GINI coefficient, and HDI. A positive relationship between neoplasms mortality and the number of doctors in the central region could possibly be explained by the fact that a higher number of doctors leads to higher quality of medical services and better diagnosis and reporting [25], whereas a negative relationship between the mortality rates of circulatory diseases and doctors indicates effective medical treatment and prevention. Mortality caused by ill-defined diseases was found not to be significantly associated with the number of doctors or hospital beds. With regard to socio-economic indicators, the GINI coefficient was found to be negatively associated with mortality rates from circulatory diseases in the coastal region. Similarly, HDI showed a negative association with mortality from circulatory diseases in both the central and coastal regions. In the northern region, HDI was positively associated with mortality from neoplasms and negatively associated with mortality from ill-defined causes. This suggests that higher HDI may be related to improved diagnostic capacity, leading to better detection and classification of neoplasms. At the same time, it may contribute to a reduction in deaths attributed to ill-defined causes, reflecting progress in overall diagnostic practices. These results are driven by various demographic and economic inequalities, considering that the coastal and central areas are more urbanized with higher economic development.

Risk factors for mortality are related to any characteristics or exposures of an individual, and they can be classified into various groups. Principally, contemporary research commonly identifies behavioral (e.g., tobacco use, alcohol consumption, physical inactivity, an unhealthy diet), metabolic (e.g., high systolic blood pressure, hypertension, overweight and obesity, high blood glucose levels and high levels of fat in the blood), and environmental risk factors (e.g., air poluttion, heat, safe chemical use, water sources) as primary contributors [26,27]. Behavioral risk factors, such as tobacco use, physical inactivity, harmful alcohol consumption, and unhealthy diets, play a significant role in the burden of NCDs in Montenegro. In particular, dietary risks, including high sodium intake and low fruit consumption, were responsible for 14.2% of total mortality. Montenegro faces a particularly high burden from tobacco use too and data from 2023 reported that about 32% of people aged 15 years and older are estimated to be daily smokers, which exceeds regional and EU rates. Notably, smoking prevalence was higher among females (33.2%) than males (30.7%) [27]. Alcohol consumption is one of the most analyzed behavioral risk factors, and in 2019, alcohol usage in Montenegro has increased above the average among countries in the South-eastern European Health Network. Opposite to tobacco use, alcohol consumption was higher among males. In addition to the above, dietary behavior such as excessive sugar and salt intake was also identified as an important factor contributing to NCD burden in Montenegro [28].

The top three metabolic risk factors in the population of Montenegro are high systolic blood pressure, high body mass index, and high fasting plasma glucose. In 2021, high systolic blood pressure alone accounted for approximately 25% of all deaths [27,29]. Similar trends have been observed in Hungary (in Central Europe) and other Southeastern and Eastern European countries, including Poland, Slovakia, the Republic of Srpska (Bosnia and Herzegovina) and Litvania, where the prevalence of metabolic risk factors have been raising, reinforcing the need for targeted public health interventions [30,31,32,33,34].

Among environmental risk factors, air pollution is one of the most important issues that affect human health and the mortality of the population in Montenegro. Air quality in Montenegro is estimated as “moderately unsafe,” with high concentrations of PM2.5 particles contributing to premature deaths [35]. Central and Eastern Europe record the highest PM2.5 emissions from residential heating using solid fuels. In South-East Europe, the biomass, especially wood, is commonly used for domestic heating due to its affordability and availability, particularly in rural areas, and the reliance on outdated and inefficient heating appliances leads to substantial emissions of particulate matter (PM2.5), black carbon, and polycyclic aromatic hydrocarbons, all of which significantly degrade air quality and increase health risks [36,37].

The regional patterns of cause-specific mortality in Montenegro reflect the cumulative impact of the behavioral, metabolic, and environmental risk factors. These factors not only explain the south–north mortality gradient for circulatory diseases and neoplasms but also align Montenegro with broader East-West mortality divides observed across Europe.

Considering mortality changes in Montenegro, it is important to emphasize that, from the end of the 20th through the beginning of the 21st century, the country has undergone profound political and socio-economic transformations, which significantly influenced its demographic trends, including mortality rates. The disintegration of Yugoslavia, regional conflicts, and the transition from a socialist to a market-based economy in the 1990s led to structural shifts in the country’s economic landscape. These changes, combined with political transformations, triggered high unemployment, poverty, and reduced public services, which in turn had an impact on NCDs and mortality as well.

4.2 Study limitations

This study has some limitations that should be acknowledged. The major issue is the quality of data about causes of death, particularly related to the ill-defined conditions. Analyzing causes of death is an important diagnostic tool that can help in understanding the mechanisms of mortality change, monitoring epidemiological trends in a population, and formulating effective public health policies [38]. However, the comparability of cause-specific mortality data is often limited by the proportion of deaths attributed to ill-defined causes [39]. According to the ICD-10 codes, ill-defined causes are the symptoms, signs, and abnormal clinical and laboratory findings, not elsewhere classified. These codes refer to conditions that either cannot or should not be classified as underlying causes of death [40]. During the observed period (1991–2019) in Montenegro, ill-defined causes accounted for approximately 15% of all deaths. However, this proportion has shown substantial variation over time and across regions. The highest share of ill-defined causes was observed in the northern (22.4%) region, followed by the central (16.4%) and coastal (7.4%) areas. During the 1990s, a particular unfavorable situation was in the northern region, where ill-defined conditions accounted for more than 40% of all recorded deaths. In 1998, this share has decreased below 40%, and in 2010, has decreased below 20%. This phenomenon occurs in all classifications of death statistics globally, and the share of ill-defined causes varies largely across European countries, but without a clear regional pattern [41,42]. Depending on the proportion of ill-defined or incomplete data, cause of death statistics may lead to an underestimation of the mortality from the well-defined causes of death and cannot accurately reflect a country’s true mortality patterns, potentially hampering cross-country comparisons and leading to biased public health priorities [42]. A high share of ill-defined conditions can result from several factors, such as lack of information about the death cause or when the physician has insufficient knowledge of diseases causing death, or has not completed the death certificate adequately [41,43,44]. Existing literature suggests that ischemic heart disease or suicide, in particular, may be misclassified as ill-defined causes of death [41,45,46]. Misclassification of ischemic heart disease can occur when myocardial infarctions lead to sudden death, and when caused by cardiac arrhythmia, the underlying pathology may remain undetectable, even at autopsy [41,47].

This study’s analysis of regional mortality patterns did not account for differences by gender or other demographic characteristics (e.g., age, education level, marital status), which may have influenced the observed regional gradient. Additionally, the health (number of doctors, hospital beds) and socio-economic indicators (GINI coefficient, HDI) were only available at the national level, not for individual regions. As a result, the regression model’s findings may be limited or underestimated due to the absence of data at the subregional level.