- Email: [email protected]
Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis
ARTICLE IN PRESS
PCD-554; No. of Pages 8
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
Contents lists available at ScienceDirect
Primary Care Diabetes journal homepage: http://www.elsevier.com/locate/pcd
Original research
Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis Milena Ilic a,∗ , Irena Ilic b a b
Department of Epidemiology, Faculty of Medical Sciences, University of Kragujevac, Kragujevac 34000, Serbia Faculty of Medical Sciences, University of Kragujevac, Serbia
a r t i c l e
i n f o
a b s t r a c t
Article history:
Purpose: The aim of this study was to analyze the mortality trends of diabetes mellitus in
Received 2 May 2016
Serbia (excluding the Autonomous Province of Kosovo and Metohia).
Received in revised form
Methods: A population-based cross sectional study analyzing diabetes mortality in Serbia in
11 August 2016
the period 1991–2015 was carried out based on official data. The age-standardized mortality
Accepted 16 August 2016
rates (per 100,000) were calculated by direct standardization, using the European Stan-
Available online xxx
dard Population. Average annual percentage of change (AAPC) and the corresponding 95% confidence interval (CI) were computed using the joinpoint regression analysis.
Keywords:
Results: More than 63,000 (about 27,000 of men and 36,000 of women) diabetes deaths
Diabetes
occurred in Serbia from 1991 to 2015. Death rates from diabetes were almost equal in men
Mortality
and in women (about 24.0 per 100,000) and places Serbia among the countries with the
Trend
highest diabetes mortality rates in Europe. Since 1991, mortality from diabetes in men sig-
Joinpoint regression analysis
nificantly increased by +1.2% per year (95% CI 0.7–1.7), but non-significantly increased in women by +0.2% per year (95% CI −0.4 to 0.7). Increased trends in diabetes type 1 mortality rates were significant in both genders in Serbia. Trends in mortality for diabetes type 2 showed a significant decrease in both genders since 2010. Conclusion: Given that diabetes mortality trends showed different patterns during the studied period, our results imply that further observation of trend is needed. © 2016 Primary Care Diabetes Europe. Published by Elsevier Ltd. All rights reserved.
1.
Introduction
Diabetes is an increasing global health problem in the world during the last decades [1–3]. Based on the World Health Organization (WHO) 2012 estimates, diabetes directly caused 1.5 million of deaths in the world in 2012 (this represents about 2.7% of all deaths globally), up from 1.0 million (2.0%) deaths
∗
in 2000 [1]. More than 80% of diabetes deaths occur in low- and middle-income countries [1]. Diabetes alone is the third most common cause of death in low- and middle-income countries in WHO Region of the Americas, accounting for nearly 0.2 million deaths each year: approximately 6.5% of all deaths in both genders are from diabetes [1]. In Europe, diabetes caused nearly 35,000 deaths in 2012 (accounts for 1.3% of all deaths). Although the number of deaths from diabetes is almost the same in both developed and developing regions, the death rates from diabetes vary by more than 50 times around the world [4]. The death rates from diabetes were high in the predominantly islands population (Trinidad and Tobago—128.2
Corresponding author. E-mail address: [email protected] (M. Ilic). http://dx.doi.org/10.1016/j.pcd.2016.08.019 1751-9918/© 2016 Primary Care Diabetes Europe. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS
2
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
per 100 000 and Saint Vincent and the Grenadines—125.8) [4]. In 2012, the mortality of diabetes varied greatly between European countries, from the highest rates in Armenia (41.5 per 100 000) and Cyprus (32.8), to the lowest rate in United Kingdom (5.4) [5]. The majority of developed countries have shown a decline in the diabetes mortality in the last two decades—the United Kingdom, Germany, Nordic countries, France, Malta [5]. In contrast, a rise in diabetes mortality was noticed in most of the developing countries—Armenia, Bosnia and Herzegovina, Georgia, and Estonia [5]. The reduction in diabetes mortality in Western countries may be attributable to the availability and use of improved treatment and management (such as the advent of blood glucose self-monitoring, hemoglobin A1c testing, etc.) and the reductions in major risk factors, while the increased mortality in developing countries probably reflects changes in lifestyle that accompany industrialization, including increased consumption of animal fat, obesity, and physical inactivity [6–9]. Serbia (officially the Republic of Serbia) is a small country in southeast Europe; in addition to civil wars (fought from 1991 to 1999), disintegration of the country, the United Nations economic sanctions imposed in 1992 and complete isolation of the country, economic collapse in 1993 with disastrous hyperinflation of national currency and severe consequences on economy and social life in general, the inability to purchase needed medications, the deterioration of public health, as well as NATO’s 79-day bombing (1999) and almost 1 million refugees, democratic changes and financial crisis marked the beginning of the century. In addition, these dramatic changes have devastating effects on the health system in Serbia. Still, despite certain improvements in medical services, Serbia suffers from significant health problems [10,11]. In this study, we assessed mortality trends from diabetes mellitus in Serbia over a 25-year period (time of civil wars and global crisis), and provided the first comprehensive estimates of the nation-wide Serbian diabetes mortality pattern.
2.
Methods
2.1.
Study design
For this population-based cross sectional study, we used annual underlying cause of death data from Serbia to describe trends in mortality from diabetes mellitus for the period 1991–2015.
2.2.
Data sources
This study comprised the whole population of the Republic of Serbia, during the period 1991–2015, excluding the Autonomous Province of Kosovo and Metohia, for which the data have been unavailable since 1998. Data on people who died of diabetes mellitus (site code 250 revision 9 and codes E10-E14 revision 10 of the International Classification of Diseases (ICD) to classify desease, injury and cause of death) were obtained from the Statistical Office of the Republic of Serbia (unpublished data). In Serbia, from 1990 to 1996 data about the main cause of death are classified by 9th Revision of ICD.
Since 1997 the data processing of mortality statistics is based on 10th Revision of ICD. All deaths occurring in the territory of Serbia are registered with a certificate of death. The underlying cause of death is certified by an authorized medical doctor in the health care organization, or a coroner. The procedure comprises several levels of control and verification: the local registrar controls and forwards death files to the referral public health institute, where another trained medical doctor controls and possibly corrects the files. The WHO has judged the overall quality of death registration data in Serbia as medium [12]. The completeness of all cause-of-death registration estimated as 100% and coverage of death registration assessed as 97% [12]. Percentage of deaths coded to ill-defined conditions (ICD-9 codes 780–799 and ICD-10 codes R00–R99) was about 6.4% in observed period [13]. Consequently, changes in certification and coding of cause of death (introduction of the new ICD 10 revision) could not affect the comparability of the data overtime. From the same source, the Statistical Office of the Republic of Serbia, data on the number and composition of the Serbian population by gender and age were obtained. We used estimated Serbian population figures, which were produced on the basis of census results (in the years 1991, 2002 and 2011) and annual results of processing of population movement for each year. This methodology is consistent with the methodology provided in the United Nations Principles and Recommendations for a Vital Statistics System. Starting from 1998 the Statistical Office of the Republic of Serbia has not at disposal and cannot make available the vital statistics for the Autonomous Province Kosovo and Metohia (Kosovo declared independence in 2008), and therefore these data are not included in the coverage of Serbia.
2.3.
Statistical analysis
Three types of mortality rates (per 100,000) were calculated: crude, age- and sex-specific and age-standardized. The age-standardized rates (ASRs) were calculated by direct standardization, using the European Standard Population (stratified by 10-year age strata). The analysis was conducted on the whole Serbian population (approximately 7.2 million inhabitants). Mortality rates of type 1 diabetes (ICD-10: E10), type 2 diabetes (ICD-10: E11), and all types of the disease (ICD10: E10–E14) were calculated. Mortality trend from diabetes was assessed using the joinpoint regression analysis (Joinpoint Regression Software, Version 4.0.4—May 2013; Statistical Methodology and Applications Branch, Surveillance Research Program of the US National Cancer Institute), according to the method proposed by Kim et al. [14]. This method uses a statistical algorithm to define a best-fitting regression line through mortality data across time, determining how many, if any, joinpoints should be used to determine where significant changes take place. The Annual Percentage Change (APC) with the corresponding 95% confidence interval (CI) was estimated for each identified trend, by fitting a regression line to the natural logarithm of the rates, using calendar year as a regression variable. We selected the log-linear model type and heteroscedastic errors option, which requires that the by-group with a zero to drop from the analysis. Therefore, joinpoint results are
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS 3
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
not shown for age groups <30 for diabetes type 1 and for age groups <40 for diabetes type 2, because count of diabetes deaths that occurred in each of the decennium in any year was zero. The analysis began with the minimum number of joinpoints (e.g., 0 joinpoint, representing a straight line), and tested whether one or more joinpoints were significant. The joinpoint analysis provided average annual percentage change (AAPC)—a summary measure over a fixed interval, with the corresponding 95% confidence interval (95%CI) [15]. For zero joinpoints, APC and AAPC are identical. In describing trends, the term “increase” was used when the slope (positive APC or AAPC) of the trend was statistically significant. Significant differences by genders were detected using a specific procedure—comparability test [16]. The test is applied to compare diabetes mortality rates between two age groups using tests of parallelism and coincidence of time trends. Test determines whether the two regression mean functions are parallel allowing different intercepts (test of parallelism) or whether two joinpoint regression functions are identical (test of coincidence). Two-sided P values were considered to indicate statistical significance when they were less than 0.05.
2.4.
Ethical considerations
This study is approved by the Ethics Committee of the Faculty of Medical Sciences, University of Kragujevac (Ref. No.: 01-1176).
3.
Results
More than 63,000 (about 27,000 of men and 36,000 of women) diabetes deaths occurred in Serbia (excluding the Autonomous Province of Kosovo and Metohia) during the observed period, with the average annual ASRs being 24.2 per 100,000 (Table 1). Mortality rates from diabetes were almost equal in men and in women (23.9 per 100,000 men and 24.0 per 100,000 women). Mortality from diabetes significantly increased since 1991 by +1.2% (95% CI 0.7–1.7) yearly in men, but non-significantly increased by +0.2% (95% CI −0.4 to 0.7) yearly in women (Fig. 1). However, for mortality from diabetes in women, there was one joinpoint at 2010: ASRs significantly increased by +0.9% (95% CI 0.3–1.5) yearly from 1991 to 2010, and then rapidly decreased by −4.8% (95% CI −9.3 to −0.2) yearly from 2010 to 2015. According to the comparability test, mortality trends from diabetes in men and women differed significantly (final selected model rejected parallelism, P = 0.0002). Overall mortality rates from diabetes significantly increased by 0.6% (95% CI 0.1–1.1) in Serbia in the whole observed period (data not shown). Changes in diabetes type 1 mortality rates were large and significant in both genders in Serbia (Fig. 2). Trends in ASRs for diabetes type 1 mortality showed a similar pattern in men and women: two joinpoints were detected for men (a significant decrease by 12.1% from 1997 to 2002, and then a significant increase by 22.8% from 2002 to 2006, followed by a significant increase by 3.8% from 2006 to onwards) and in women (a significant decrease by 11.5% from 1997 to 2002 and a significant increase by 11.4% from 2002 to 2011, and then a non-significant
Table 1 – Diabetes mellitus mortality in Serbia, excluding the Autonomous Province of Kosovo and Metohia, in 1991–2015 period. Year
No.
Crude rates
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
1904 1952 1769 1759 2104 2181 2209 2232 2514 2506 2336 2553 2660 2571 2575 2541 3005 3113 3066 3195 3135 2998 2802 2513 3027
25.1 25.7 23.2 23.1 27.6 28.6 29.1 29.5 33.3 33.3 31.1 34.0 35.6 34.5 34.6 34.3 40.7 42.4 41.9 43.8 43.2 41.6 39.1 35.2 42.5
22.9 23.0 20.3 19.4 23.0 23.6 23.2 23.1 25.8 25.4 23.0 24.8 25.3 24.0 23.8 23.2 27.3 28.0 27.0 28.1 27.2 25.5 23.8 20.7 24.8
63,220
34.1
24.2
Overall
ASRs
ASR = age standardized rate (per 100,000 persons, using European standard population).
decline by −6.7% per year from 2011 to onwards). According to the comparability test, mortality trends from diabetes type 1 in men and women were parallel (final selected model failed to reject parallelism, P > 0.05). Trends in ASRs for diabetes type 2 mortality showed a significant decrease in both genders since 2010 (Fig. 3). One joinpoint was detected for men (a non-significant increased trend was observed from 1997 to 2010, but then a significantly decreased trend by 8.3% was recorded from 2010 to onwards) and in women (a non-significant change was observed from 1997 to 2010, but from 2010 to onwards a significantly decreasing trend by 9.8% was recorded). According to the comparability test, mortality trends from diabetes type 2 in men and women were not parallel (final selected model rejected parallelism, P = 0.0129). Since the 2000s, a significant increase in mortality rates from diabetes type 1 was identified in almost all age groups in both genders, except in one age group (50–59 years) in women where a significant decrease was present (Table 2). Nonsignificant trends in both genders were observed in the age groups of 30–49 years (increased trends in men and decreased trends in women). Trends in age-specific mortality rates from diabetes type 2 showed a similar pattern in men and women (Table 3). Declines in mortality rates from diabetes type 2 were identified in almost all age groups, except in the oldest age group in both genders where a significant increase was present during the whole period. Significant declining trends in type 2
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS
4
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
Fig. 1 – Diabetes mellitus mortality trends, men and women in Serbia, in 1991–2015 period: a joinpoint regression analysis. *Statistically significant trend; APC—annual percent change. diabetes mortality were observed for men and women aged 40–69 and for men aged 70–79 for the entire period. For men, a significant decrease in type 2 diabetes mortality started in 2010 for those aged 60–69 and 70–79 years respectively. This was also the case for women: those aged 50–59 years from 2004, those aged 60–69 from 2011, and those aged 70–79 from 2010.
4.
Discussion
Our results show an overall unfavorable trend for diabetes mortality in Serbia over the observed 25-year period. Diabetes kills about 2500 people annually in Serbia. The mortality rate from diabetes places Serbia among the countries with the highest diabetes mortality rates in Europe. Particularly worrisome is the increased trend in mortality due to diabetes type 1
Fig. 2 – Diabetes mellitus type 1 mortality trends, men and women in Serbia, in 1997–2015 period: a joinpoint regression analysis. *Statistically significant trend; APC—annual percent change. Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
ARTICLE IN PRESS
PCD-554; No. of Pages 8
5
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
Fig. 3 – Diabetes mellitus type 2 mortality trends, men and women in Serbia, in 1997–2015 period: a joinpoint regression analysis. *Statistically significant trend; APC—annual percent change.
in men. The exception was a decreased trend in diabetes type 2 mortality rates both in men and women. The ASR of diabetes (25.5 per 100,000) in Serbia was two times higher than the average ASR in countries of European region (12.9 per 100,000) in 2012 [5]. In 2012, in Europe, high rates have been observed in Armenia (41.5 per 100.000) and the Mediterranean countries (Cyprus—32.8, Turkey—30.8, Israel—26.5, Portugal—23.0); the lowest rate (approximately—5.5) in United Kingdom, Finland and Greece [5]. Finally, the above mentioned mortality rate puts Serbia
on the fifth place among the countries with the highest diabetes mortality rates in Europe. Some unfavorable changes (economic crisis, poverty, civil wars, refugees, unemployment, disintegration of the country, NATO bombing) occurred in Serbia in the last decades, that have had a considerable impact on lifestyle as well as on the functioning of the health services in Serbia during the studied period (lack of drugs, public health weakness) [17,18]. The National Health Survey revealed that 18.3% of adult male population in Serbia were obese (body mass index ≥30) in 2006, significantly more
Table 2 – Trendsa in age-specific diabetes type 1 mortality rates, men and women in Serbia, in 1997–2015 period: a joinpoint regression analysis. Ageb
AAPC (95% CI)
Men 30–40 40–49 50–59 60–69 70–79 80+
0.9 (−3.7 to 5.6) 0.7 (−1.1 to 2.5) 3.1a (0.7–5.6) 5.3a (2.5–8.3) 6.8a (4.4–9.2) 11.9a (9.4–14.5)
Women 30–40 40–49 50–59 60–69 70–79 80+
−3.0 (−7.7 to 1.9) −1.8 (−5.5 to 2.1) −2.6a (−5.1 to −0.2) 1.1 (−1.5 to 3.7) 5.5a (3.4–7.6) 13.3a (9.4–17.3)
Trend 1
APC (95% CI)
Trend 2
APC (95% CI)
1997–2001 1997–2001 1997–2001
−13.4 (−28.4 to 4.8) −13.5 (−30.3 to 7.4) −10.5 (−25.2 to 7.1)
2001–2015 2001–2015 2001–2015
6.1a (3.2–9.2) 8.8a (5.4–12.3) 9.9a (7.0–12.8)
1997–2000 1997–2001
−23.8 (−45.6 to 6.6) −11.3 (−23.4 to 2.6)
2000–2015 2001–2011
3.9a (1.0–6.9) 11.4a (6.8–16.2)
Trend 3
2011–2015
APC (95% CI)
−4.8 (−17.7 to 10.1)
AAPC = average annual percent change; APC = annual percent change; CI = confidence interval. Statistically significant trend. b Joinpoint results are not shown for age subgroups <30, because count of diabetes type 1 deaths occurred in each of the decennium in any year was zero. a
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS
6
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
Table 3 – Trendsa in age-specific diabetes type 2 mortality rates, men and women in Serbia, in 1997–2015 period: a joinpoint regression analysis. Ageb
AAPC (95% CI)
Men 40–49 50–59 60–69 70–79 80+
−4.3a (−6.7 to −1.9) −2.9a (−4.5 to −1.3) −1.9a (−3.4 to −0.3) −1.1 (−2.3 to 0.1) 2.8a (1.5–4.1)
Women 40–49 50–59 60–69 70–79 80+
−5.6a (−8.2 to −3.0) −6.0a (−7.9 to −4.0) −4.4a (−5.7 to −3.0) −2.2a (−3.4 to −0.9) 2.3a (0.9–3.7)
Trend 1
APC (95% CI)
Trend 2
APC (95% CI)
1997–2010 1997–2010
0.7 (−1.1 to 2.6) 1.0 (−0.2 to 2.2)
2010–2015 2010–2015
−11.4a (−18.0 to −4.3) −9.0a (−13.5 to −4.3)
1997–2004 1997–2011 1997–1999 1997–2010
2.5 (−4.2 to 9.7) −2.3a (−3.6 to −1.0) 12.3 (−5.9 to 33.9) 4.3a (2.3–6.3)
2004–2015 2011–2015 1999–2010 2010–2015
−10.2a (−13.2 to −7.0) −15.9a (−23.0 to −8.2) −0.7 (−2.1 to 0.7) −5.2 (−12.7 to 2.9)
Trend 3
2010–2015
APC (95% CI)
−10.2a (−13.7 to −6.6)
AAPC = average annual percent change; APC = annual percent change; CI = confidence interval. Statistically significant trend. b Joinpoint results are not shown for age subgroups <40, because count of diabetes type 1 deaths occurred in each of the decennium in any year was zero. a
than in 2000 (17.4%), while the pre-obese category (body mass index = 25–29.9) was more often found among men (41.4%) than among women (29.1%) in 2013 [19]. In Serbia, more than one half of the population smoked in some part of their lives (51.8%) [19]. In Serbia, Sipetic´ et al. [20] found that infections during the 6 months preceding the onset of the disease and stressful events were significantly related to diabetes and supported the hypothesis that environmental factors play a role in the development of type 1 diabetes. Certainly, differences in the quality of data on cause of death across countries would have led to some of the effects that were found, but bias due to secular improvements in diabetes registration is unlikely to explain most of the observed changes [21]. Previously published studies on diabetes temporal trends have shown inconclusive results [22–24]. Large geographical variations and an increasing temporal trend in diabetes mortality are evident among most countries. Globally, total deaths from diabetes are projected to rise by more than 50% in the next 10 years [5]. In contrast to the reduction in diabetes mortality in majority of developed countries [25], the significant increase in diabetes mortality was observed in Serbia in last decades. Similar, an increased mortality of diabetes was revealed in many Central and Eastern European countries [26,27]. Differences in diabetes mortality rates between countries could in part be due to variations in incidence of diabetes, as well as in the variations in the availability and use of prevention and treatment. Evidence shows that type 1 diabetes incidence is increasing more steeply in some Central and Eastern European countries. Some authors noted the rising incidence of type 1 diabetes in Belgrade children aged 0–14 years in the period from 1982 to 2005 [28]. The increased frequency from diabetes is likely attributable to rapid economic development, rapid changes in lifestyle (dietary changes, reduced physical activity and unhealthy behaviors), an aging population, and a Westernized lifestyle in developing countries [29]. Some studies have proposed that the increasing early life exposure to infectious diseases, which typically occurred over the past several decades in developing countries, could be involved in this trend [30]. Besides, greater
responsibility and stress in times of wars and economic crisis placed more on men than on women could be an explanation for the more unfavorable mortality trend from diabetes in men in Serbia [31]. At the end of 1999, Serbia had almost 1 million refugees and internally displaced people, which made it the country with the highest number of refugees in Europe. Refugees came to Serbia in waves: the first big wave was in 1992, the second was in 1995 and the third was in 1999. In 1995, 400,000 people from Croatia were expelled and moved to Serbia in a very short period of about two weeks. In 1999, about 300,000 people from Kosovo moved to Serbia after NATO’s bombing [32]. Over the past decades, around 200,000 refugees have returned to their countries of origin or migrated to third countries, and majority of refugees decided to integrate in Serbia [32,33]. Although the number of refugees in Serbia is decreasing, their social status has been bad for years and the qualitative changes are slow, further was deteriorated by unemployment, long lasting life in collective centers with very poor sanitary conditions, lack of financial sources for food, medicines, hygiene, education, clothes etc. The United Nations High Commissioner for Refugees classified Serbia among five countries of ‘long-term refugees’ crisis. Unfortunately, in Serbia there are no precise indicators of the health status of refugees, which would enable comparison of health status before the war, during the war and after the war [34]. Numerous studies evidenced that diabetes is a problem in refugees populations of South and East Asian, Latin American, Middle East (Iraqi, Syrian) and African, Eastern Europe (countries of the former Soviet Union and Yugoslavia) origins, etc. [35–37]. But, there is no available mortality data for Serbian refugees for the previous period, even for those who still live in Serbia, nor for those who have left Serbia. Unfavorable trends in mortality from diabetes among men in Serbia could mainly be attributed to mortality from diabetes type 1. Also, sex differences in mortality of both diabetes type 1 and type 2, with higher risk in males than in females, have been previously reported in many studies [38]. This finding may suggest a role for sex-linked differences in function
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
or in insulin sensitivity, but involves complex social and cultural issues of gender [23]. Type 1 diabetes includes those cases currently ascribable to an autoimmune process and those for which etiology is unknown, while type 2 diabetes is caused by both environmental (such as physical inactivity, drugs and toxic agents, obesity, viral infection) and genetic factors [39]. The increased rates of deaths from diabetes type 1 might have been a consequence of unfavorable circumstances in Serbia in the observed period: the general standard was poor, the quality of health services was unsatisfactory (including the lack of vaccines, drugs, medical equipment, a large number of wounded as well as refugees, decreasing hospitalization rates, particularly for people aged 60 years and older), social disintegration, aging of the population, etc. [9]. Relatively favorable trend in mortality from diabetes type 2 may be attributed to the implementation of mass preventive programs (tobacco control, etc.) and improvements in diabetes treatment in Serbia [19]. Paradoxically, the decline trend in diabetes mortality in women in Serbia coincided with the global economic crisis. Therefore, these results should encourage the launching of epidemiological studies in order to identify future trends in diabetes mortality and to plan for health care delivery, especially in developing countries.
4.1.
Strengths and limitations of this study
This study provides the first nationwide estimates of diabetes mortality in Serbia over 25 years. The strength of our study lies in the fact that it is a population-based study, using quality death data with comprehensive coverage, completeness and comparability, with temporal trends assessed by joinpoint analysis. Thanks to the WHO assessment of the quality of data on the cause of death in Serbia as intermediate [12], we could give a good assessment of national mortality trend for diabetes and carry out a comparison with the pattern of diabetes mortality in other countries. Also, our study implies the need for effective measures of diabetes prevention and treatment in Serbia. However, there were several sources of limitations in this paper. We acknowledge that a longer study period may be better to more accurately assess mortality time-trends, but in Serbia there was no available data for this. Also, a limitation is that there are no separate data on diabetes deaths among refugees that can possibly confound the diabetes mortality pattern in Serbia: for internally displaced persons and refugees data were included in the Serbian population and could not be set aside as a special contingent. Certainly, a better quality deaths registration system, which is available in high-income countries, is needed in Serbia. Namely, mortality rates from diabetes may have been underestimated, because of the errors in coding the causes of death and recording the diabetes as antecedent, instead of underlying main cause of death, particularly in those who died from infarction, stroke, and renal diseases [40]. Also, there are no data of diabetes incidence for the entire study period to be used for explanation of temporal mortality trends (Diabetes Registry was implemented since 2006). The available literature does not have enough relevant information about the presence of risk factors for diabetes in the Serbian population. In the applied log-linear model in this study, dropping a particular year observations
7
with count of diabetes deaths being zero from the analysis may shift or affect the detection or locations of joinpoints affecting the analysis. Despite these limitations, this study helped to elucidate diabetes death trends in Serbia, which still need to be clarified in analytical epidemiological studies in the future.
Conflict of Interest The authors state that they have no conflict of interest.
references
[1] World Health Organization, WHO methods and data sources for global causes of death 2000–2012, in: Global Health Estimates, Technical Paper WHO/HIS/HSI/GHE/2014.7, World Health Organization, 2014. [2] L. Chen, D.J. Magliano, P.Z. Zimmet, The worldwide epidemiology of type 2 diabetes mellitus—present and future perspectives, Nat. Rev. Endocrinol. 8 (4) (2011) 228–236. [3] C.D. Mathers, D. Loncar, Projections of global mortality and burden of disease from 2002 to 2030, PLoS Med. 3 (11) (2006) e442, http://dx.doi.org/10.1371/journal.pmed.0030442. [4] WHO Global Infobase Team, The SuRF Report 2. Surveillance of Chronic Disease Risk Factors: Country-level Data and Comparable Estimates, WHO, Geneva, 2005, Available at: https://www.who.int/ncd surveillance/infobase/web/surf2/ (accessed 27.02.16). [5] World Health Organization, European Health for all Database, WHO Regional Office for Europe, Copenhagen, Denmark, 2016. [6] C.C. Patterson, E. Gyürüs, J. Rosenbauer, et al., Trends in childhood type 1 diabetes incidence in Europe during 1989–2008: evidence of non-uniformity over time in rates of increase, Diabetologia 55 (8) (2012) 2142–2147. [7] P.J. Liao, Z.Y. Lin, J.C. Huang, et al., The relationship between type 2 diabetic patients’ early medical care-seeking consistency to the same clinician and health care system and their clinical outcomes, Medicine (Baltimore) 94 (7) (2015) e554, http://dx.doi.org/10.1097/MD.0000000000000554. [8] Writing Group for the DCCT/EDIC Research Group, T.J. Orchard, D.M. Nathan, B. Zinman, et al., Association between 7 years of intensive treatment of type 1 diabetes and long-term mortality, JAMA 313 (1) (2015) 45–53. [9] C.C. Patterson, G. Dahlquist, G. Soltész, A. Green, EURODIAB ACE Study Group. Europe and Diabetes, Is childhood-onset type I diabetes a wealth-related disease? An ecological analysis of European incidence rates, Diabetologia 44 (Suppl. 3) (2001) B9–B16. [10] M. Ilic, I. Ilic, Gender disparities in mortality from infectious diseases in Serbia, 1991–2014: a time of civil wars and global crisis, Epidemiol. Infect. 144 (12) (2016) 2473–2484. [11] M. Ilic, I. Ilic, Suicide in Serbia, J. Affect. Disord. 193 (2016) 187–193. [12] C.D. Mathers, D. Ma Fat, M. Inoue, et al., Counting the dead and what they died from: an assessment of the global status of cause death data, Bull. World Health Organ. 83 (2005) 171–177. [13] Statistical Office of the Republic of Serbia, Statistical Yearbook of the Republic of Serbia for 1991–2014, Statistical Office of the Republic of Serbia, Belgrade, 2015. [14] H.J. Kim, M.P. Fay, E.J. Feuer, et al., Permutation tests for joinpoint regression with applications to cancer rates, Stat. Med. 19 (2000) 335–351.
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS
8
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
[15] L.X. Clegg, B.F. Hankey, R. Tiwari, et al., Estimating average annual per cent change in trend analysis, Stat. Med. 28 (2009) 3670–3682. [16] H.J. Kim, M.P. Fay, B. Yu, et al., Comparability of segmented line regression models, Biometrics 60 (2004) 1005–1014. [17] The Impact of UN Security Council Sanctions on the Health of the Population of FR Yugoslavia, Federal Institute for Public Health, Belgrade, Serbia, 1993. [18] S.J. Kunitz, The making and breaking of Yugoslavia and its impact on health, Am. J. Public Health 94 (2004) 1894–1904. [19] Ministry of Health, Republic of Serbia, National Health Survey, Serbia 2006, Ministry of Health, Republic of Serbia, Belgrade, 2007. ´ H.D. Vlajinac, N.I. Kocev, et al., The Belgrade [20] S.B. Sipetic, childhood diabetes study: a multivariate analysis of risk determinants for diabetes, Eur. J. Public Health 15 (2005) 117–122. [21] L.N. McEwen, A.J. Karter, J.D. Curb, et al., Temporal trends in recording of diabetes on death certificates: results from Translating Research Into Action for Diabetes (TRIAD), Diabetes Care 34 (7) (2011) 1529–1533. [22] G. Roglic, N. Unwin, Mortality attributable to diabetes: estimates for the year 2010, Diabetes Res. Clin. Pract. 87 (1) (2010) 15–19. [23] M.Ø. Krag, L. Hasselbalch, V. Siersma, et al., The impact of gender on the long-term morbidity and mortality of patients with type 2 diabetes receiving structured personal care: a 13 year follow-up study, Diabetologia 59 (2) (2016) 275–285. [24] W.H. Lin, M.C. Wang, W.M. Wang, et al., Incidence of and mortality from type I diabetes in Taiwan from 1999 through 2010: a nationwide cohort study, PLoS One 9 (1) (2014) e86172, http://dx.doi.org/10.1371/journal.pone.0086172.eCollection. [25] S. Wild, G. Roglic, A. Green, et al., Global prevalence of diabetes: estimates for the year 2000 and projections for 2030, Diabetes Care 27 (2004) 1047–1053. [26] P.E. Wändell, A.C. Carlsson, Time trends and gender differences in incidence and prevalence of type 1 diabetes in Sweden, Curr. Diabetes Rev. 9 (4) (2013) 342–349. [27] J.M. Lawrence, G. Imperatore, D. Dabelea, et al., Trends in incidence of type 1 diabetes among non-Hispanic white youth in the U.S., 2002–2009, Diabetes 63 (11) (2014) 3938–3945. [28] S. Sipetic, J. Maksimovic, H. Vlajinac, et al., Rising incidence of type 1 diabetes in Belgrade children aged 0–14 years in the period from 1982 to 2005, J. Endocrinol. Investig. 36 (5) (2013) 307–312.
[29] W.P.T. James, R. Jackson-Leach, C. Ni Mhurchu, et al., Overweight and obesity (high body mass index), in: M. Ezzati, A. Lopez, A. Rodgers, C.J.L. Murray (Eds.), Comparative Quantification of Health Risks: Global and Regional Burden of Disease Attributable to Selected Major Risk Factors, WHO, Geneva, 2004. [30] J.F. Bach, The effect of infections on susceptibility to autoimmune and allergic diseases, N. Engl. J. Med. 347 (2002) 911–920. [31] S. Thurner, P. Klimek, M. Szell, et al., Quantification of excess risk for diabetes for those born in times of hunger, in an entire population of a nation, across a century, Proc. Natl. Acad. Sci. U. S. A. 110 (12) (2013) 4703–4707. [32] Commissariat for Refugees and Migration of the Republic of Serbia: Process of Voluntary Repatriation of Refugees to the Republic of Croatia and Bosnia and Herzegovina, United Nations High Commissioner for Refugees (UNHCR), Belgrade, 2010, Available at: http://www.kirs.gov.rs/ (accessed 07.08.16). [33] V. Lukic, Two Decades of Refuge in Serbia: Census of Population, Households and Dwellings 2011 in the Republic of Serbia, Statistical Office of the Republic of Serbia, Belgrade, 2015. [34] S. Litvinjenko, Migration and health, Srp. Arh. Celok. Lek. 125 (1997) 191–196 (Article in Serbian). [35] M. DesMeules, J. Gold, S. McDermott, et al., Disparities in mortality patterns among Canadian immigrants and refugees, 1980–1998: results of a national cohort study, J. Immigr. Health 7 (4) (2005) 221–232. [36] J. Wagner, S.M. Berthold, T. Buckley, et al., Diabetes among refugee populations: what newly arriving refugees can learn from resettled Cambodians, Curr. Diabetes Rep. 15 (8) (2015) 56, http://dx.doi.org/10.1007/s11892-015-0618-1. [37] F.J. Mateen, M. Carone, H. Al-Saedy, et al., Medical conditions among Iraqi refugees in Jordan: data from the United Nations Refugee Assistance Information System, Bull. World Health Organ. 90 (6) (2012) 444–451. [38] G. Bruno, M. Maule, F. Merletti, et al., Age-period-cohort analysis of 1990–2003 incidence time trends of childhood diabetes in Italy: the RIDI study, Diabetes 59 (9) (2010) 2281–2287. [39] E. Adeghate, P. Schattner, E. Dunn, An update on the etiology and epidemiology of diabetes mellitus, Ann. N. Y. Acad. Sci. 1084 (2006) 1–29. [40] N.J. Morrish, S.L. Wang, L.K. Stevens, et al., Mortality and causes of death in the WHO Multinational Study of vascular disease in diabetes, Diabetologia 44 (2001) S14–S21.
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
Contents lists available at ScienceDirect
Primary Care Diabetes journal homepage: http://www.elsevier.com/locate/pcd
Original research
Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis Milena Ilic a,∗ , Irena Ilic b a b
Department of Epidemiology, Faculty of Medical Sciences, University of Kragujevac, Kragujevac 34000, Serbia Faculty of Medical Sciences, University of Kragujevac, Serbia
a r t i c l e
i n f o
a b s t r a c t
Article history:
Purpose: The aim of this study was to analyze the mortality trends of diabetes mellitus in
Received 2 May 2016
Serbia (excluding the Autonomous Province of Kosovo and Metohia).
Received in revised form
Methods: A population-based cross sectional study analyzing diabetes mortality in Serbia in
11 August 2016
the period 1991–2015 was carried out based on official data. The age-standardized mortality
Accepted 16 August 2016
rates (per 100,000) were calculated by direct standardization, using the European Stan-
Available online xxx
dard Population. Average annual percentage of change (AAPC) and the corresponding 95% confidence interval (CI) were computed using the joinpoint regression analysis.
Keywords:
Results: More than 63,000 (about 27,000 of men and 36,000 of women) diabetes deaths
Diabetes
occurred in Serbia from 1991 to 2015. Death rates from diabetes were almost equal in men
Mortality
and in women (about 24.0 per 100,000) and places Serbia among the countries with the
Trend
highest diabetes mortality rates in Europe. Since 1991, mortality from diabetes in men sig-
Joinpoint regression analysis
nificantly increased by +1.2% per year (95% CI 0.7–1.7), but non-significantly increased in women by +0.2% per year (95% CI −0.4 to 0.7). Increased trends in diabetes type 1 mortality rates were significant in both genders in Serbia. Trends in mortality for diabetes type 2 showed a significant decrease in both genders since 2010. Conclusion: Given that diabetes mortality trends showed different patterns during the studied period, our results imply that further observation of trend is needed. © 2016 Primary Care Diabetes Europe. Published by Elsevier Ltd. All rights reserved.
1.
Introduction
Diabetes is an increasing global health problem in the world during the last decades [1–3]. Based on the World Health Organization (WHO) 2012 estimates, diabetes directly caused 1.5 million of deaths in the world in 2012 (this represents about 2.7% of all deaths globally), up from 1.0 million (2.0%) deaths
∗
in 2000 [1]. More than 80% of diabetes deaths occur in low- and middle-income countries [1]. Diabetes alone is the third most common cause of death in low- and middle-income countries in WHO Region of the Americas, accounting for nearly 0.2 million deaths each year: approximately 6.5% of all deaths in both genders are from diabetes [1]. In Europe, diabetes caused nearly 35,000 deaths in 2012 (accounts for 1.3% of all deaths). Although the number of deaths from diabetes is almost the same in both developed and developing regions, the death rates from diabetes vary by more than 50 times around the world [4]. The death rates from diabetes were high in the predominantly islands population (Trinidad and Tobago—128.2
Corresponding author. E-mail address: [email protected] (M. Ilic). http://dx.doi.org/10.1016/j.pcd.2016.08.019 1751-9918/© 2016 Primary Care Diabetes Europe. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS
2
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
per 100 000 and Saint Vincent and the Grenadines—125.8) [4]. In 2012, the mortality of diabetes varied greatly between European countries, from the highest rates in Armenia (41.5 per 100 000) and Cyprus (32.8), to the lowest rate in United Kingdom (5.4) [5]. The majority of developed countries have shown a decline in the diabetes mortality in the last two decades—the United Kingdom, Germany, Nordic countries, France, Malta [5]. In contrast, a rise in diabetes mortality was noticed in most of the developing countries—Armenia, Bosnia and Herzegovina, Georgia, and Estonia [5]. The reduction in diabetes mortality in Western countries may be attributable to the availability and use of improved treatment and management (such as the advent of blood glucose self-monitoring, hemoglobin A1c testing, etc.) and the reductions in major risk factors, while the increased mortality in developing countries probably reflects changes in lifestyle that accompany industrialization, including increased consumption of animal fat, obesity, and physical inactivity [6–9]. Serbia (officially the Republic of Serbia) is a small country in southeast Europe; in addition to civil wars (fought from 1991 to 1999), disintegration of the country, the United Nations economic sanctions imposed in 1992 and complete isolation of the country, economic collapse in 1993 with disastrous hyperinflation of national currency and severe consequences on economy and social life in general, the inability to purchase needed medications, the deterioration of public health, as well as NATO’s 79-day bombing (1999) and almost 1 million refugees, democratic changes and financial crisis marked the beginning of the century. In addition, these dramatic changes have devastating effects on the health system in Serbia. Still, despite certain improvements in medical services, Serbia suffers from significant health problems [10,11]. In this study, we assessed mortality trends from diabetes mellitus in Serbia over a 25-year period (time of civil wars and global crisis), and provided the first comprehensive estimates of the nation-wide Serbian diabetes mortality pattern.
2.
Methods
2.1.
Study design
For this population-based cross sectional study, we used annual underlying cause of death data from Serbia to describe trends in mortality from diabetes mellitus for the period 1991–2015.
2.2.
Data sources
This study comprised the whole population of the Republic of Serbia, during the period 1991–2015, excluding the Autonomous Province of Kosovo and Metohia, for which the data have been unavailable since 1998. Data on people who died of diabetes mellitus (site code 250 revision 9 and codes E10-E14 revision 10 of the International Classification of Diseases (ICD) to classify desease, injury and cause of death) were obtained from the Statistical Office of the Republic of Serbia (unpublished data). In Serbia, from 1990 to 1996 data about the main cause of death are classified by 9th Revision of ICD.
Since 1997 the data processing of mortality statistics is based on 10th Revision of ICD. All deaths occurring in the territory of Serbia are registered with a certificate of death. The underlying cause of death is certified by an authorized medical doctor in the health care organization, or a coroner. The procedure comprises several levels of control and verification: the local registrar controls and forwards death files to the referral public health institute, where another trained medical doctor controls and possibly corrects the files. The WHO has judged the overall quality of death registration data in Serbia as medium [12]. The completeness of all cause-of-death registration estimated as 100% and coverage of death registration assessed as 97% [12]. Percentage of deaths coded to ill-defined conditions (ICD-9 codes 780–799 and ICD-10 codes R00–R99) was about 6.4% in observed period [13]. Consequently, changes in certification and coding of cause of death (introduction of the new ICD 10 revision) could not affect the comparability of the data overtime. From the same source, the Statistical Office of the Republic of Serbia, data on the number and composition of the Serbian population by gender and age were obtained. We used estimated Serbian population figures, which were produced on the basis of census results (in the years 1991, 2002 and 2011) and annual results of processing of population movement for each year. This methodology is consistent with the methodology provided in the United Nations Principles and Recommendations for a Vital Statistics System. Starting from 1998 the Statistical Office of the Republic of Serbia has not at disposal and cannot make available the vital statistics for the Autonomous Province Kosovo and Metohia (Kosovo declared independence in 2008), and therefore these data are not included in the coverage of Serbia.
2.3.
Statistical analysis
Three types of mortality rates (per 100,000) were calculated: crude, age- and sex-specific and age-standardized. The age-standardized rates (ASRs) were calculated by direct standardization, using the European Standard Population (stratified by 10-year age strata). The analysis was conducted on the whole Serbian population (approximately 7.2 million inhabitants). Mortality rates of type 1 diabetes (ICD-10: E10), type 2 diabetes (ICD-10: E11), and all types of the disease (ICD10: E10–E14) were calculated. Mortality trend from diabetes was assessed using the joinpoint regression analysis (Joinpoint Regression Software, Version 4.0.4—May 2013; Statistical Methodology and Applications Branch, Surveillance Research Program of the US National Cancer Institute), according to the method proposed by Kim et al. [14]. This method uses a statistical algorithm to define a best-fitting regression line through mortality data across time, determining how many, if any, joinpoints should be used to determine where significant changes take place. The Annual Percentage Change (APC) with the corresponding 95% confidence interval (CI) was estimated for each identified trend, by fitting a regression line to the natural logarithm of the rates, using calendar year as a regression variable. We selected the log-linear model type and heteroscedastic errors option, which requires that the by-group with a zero to drop from the analysis. Therefore, joinpoint results are
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS 3
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
not shown for age groups <30 for diabetes type 1 and for age groups <40 for diabetes type 2, because count of diabetes deaths that occurred in each of the decennium in any year was zero. The analysis began with the minimum number of joinpoints (e.g., 0 joinpoint, representing a straight line), and tested whether one or more joinpoints were significant. The joinpoint analysis provided average annual percentage change (AAPC)—a summary measure over a fixed interval, with the corresponding 95% confidence interval (95%CI) [15]. For zero joinpoints, APC and AAPC are identical. In describing trends, the term “increase” was used when the slope (positive APC or AAPC) of the trend was statistically significant. Significant differences by genders were detected using a specific procedure—comparability test [16]. The test is applied to compare diabetes mortality rates between two age groups using tests of parallelism and coincidence of time trends. Test determines whether the two regression mean functions are parallel allowing different intercepts (test of parallelism) or whether two joinpoint regression functions are identical (test of coincidence). Two-sided P values were considered to indicate statistical significance when they were less than 0.05.
2.4.
Ethical considerations
This study is approved by the Ethics Committee of the Faculty of Medical Sciences, University of Kragujevac (Ref. No.: 01-1176).
3.
Results
More than 63,000 (about 27,000 of men and 36,000 of women) diabetes deaths occurred in Serbia (excluding the Autonomous Province of Kosovo and Metohia) during the observed period, with the average annual ASRs being 24.2 per 100,000 (Table 1). Mortality rates from diabetes were almost equal in men and in women (23.9 per 100,000 men and 24.0 per 100,000 women). Mortality from diabetes significantly increased since 1991 by +1.2% (95% CI 0.7–1.7) yearly in men, but non-significantly increased by +0.2% (95% CI −0.4 to 0.7) yearly in women (Fig. 1). However, for mortality from diabetes in women, there was one joinpoint at 2010: ASRs significantly increased by +0.9% (95% CI 0.3–1.5) yearly from 1991 to 2010, and then rapidly decreased by −4.8% (95% CI −9.3 to −0.2) yearly from 2010 to 2015. According to the comparability test, mortality trends from diabetes in men and women differed significantly (final selected model rejected parallelism, P = 0.0002). Overall mortality rates from diabetes significantly increased by 0.6% (95% CI 0.1–1.1) in Serbia in the whole observed period (data not shown). Changes in diabetes type 1 mortality rates were large and significant in both genders in Serbia (Fig. 2). Trends in ASRs for diabetes type 1 mortality showed a similar pattern in men and women: two joinpoints were detected for men (a significant decrease by 12.1% from 1997 to 2002, and then a significant increase by 22.8% from 2002 to 2006, followed by a significant increase by 3.8% from 2006 to onwards) and in women (a significant decrease by 11.5% from 1997 to 2002 and a significant increase by 11.4% from 2002 to 2011, and then a non-significant
Table 1 – Diabetes mellitus mortality in Serbia, excluding the Autonomous Province of Kosovo and Metohia, in 1991–2015 period. Year
No.
Crude rates
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
1904 1952 1769 1759 2104 2181 2209 2232 2514 2506 2336 2553 2660 2571 2575 2541 3005 3113 3066 3195 3135 2998 2802 2513 3027
25.1 25.7 23.2 23.1 27.6 28.6 29.1 29.5 33.3 33.3 31.1 34.0 35.6 34.5 34.6 34.3 40.7 42.4 41.9 43.8 43.2 41.6 39.1 35.2 42.5
22.9 23.0 20.3 19.4 23.0 23.6 23.2 23.1 25.8 25.4 23.0 24.8 25.3 24.0 23.8 23.2 27.3 28.0 27.0 28.1 27.2 25.5 23.8 20.7 24.8
63,220
34.1
24.2
Overall
ASRs
ASR = age standardized rate (per 100,000 persons, using European standard population).
decline by −6.7% per year from 2011 to onwards). According to the comparability test, mortality trends from diabetes type 1 in men and women were parallel (final selected model failed to reject parallelism, P > 0.05). Trends in ASRs for diabetes type 2 mortality showed a significant decrease in both genders since 2010 (Fig. 3). One joinpoint was detected for men (a non-significant increased trend was observed from 1997 to 2010, but then a significantly decreased trend by 8.3% was recorded from 2010 to onwards) and in women (a non-significant change was observed from 1997 to 2010, but from 2010 to onwards a significantly decreasing trend by 9.8% was recorded). According to the comparability test, mortality trends from diabetes type 2 in men and women were not parallel (final selected model rejected parallelism, P = 0.0129). Since the 2000s, a significant increase in mortality rates from diabetes type 1 was identified in almost all age groups in both genders, except in one age group (50–59 years) in women where a significant decrease was present (Table 2). Nonsignificant trends in both genders were observed in the age groups of 30–49 years (increased trends in men and decreased trends in women). Trends in age-specific mortality rates from diabetes type 2 showed a similar pattern in men and women (Table 3). Declines in mortality rates from diabetes type 2 were identified in almost all age groups, except in the oldest age group in both genders where a significant increase was present during the whole period. Significant declining trends in type 2
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS
4
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
Fig. 1 – Diabetes mellitus mortality trends, men and women in Serbia, in 1991–2015 period: a joinpoint regression analysis. *Statistically significant trend; APC—annual percent change. diabetes mortality were observed for men and women aged 40–69 and for men aged 70–79 for the entire period. For men, a significant decrease in type 2 diabetes mortality started in 2010 for those aged 60–69 and 70–79 years respectively. This was also the case for women: those aged 50–59 years from 2004, those aged 60–69 from 2011, and those aged 70–79 from 2010.
4.
Discussion
Our results show an overall unfavorable trend for diabetes mortality in Serbia over the observed 25-year period. Diabetes kills about 2500 people annually in Serbia. The mortality rate from diabetes places Serbia among the countries with the highest diabetes mortality rates in Europe. Particularly worrisome is the increased trend in mortality due to diabetes type 1
Fig. 2 – Diabetes mellitus type 1 mortality trends, men and women in Serbia, in 1997–2015 period: a joinpoint regression analysis. *Statistically significant trend; APC—annual percent change. Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
ARTICLE IN PRESS
PCD-554; No. of Pages 8
5
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
Fig. 3 – Diabetes mellitus type 2 mortality trends, men and women in Serbia, in 1997–2015 period: a joinpoint regression analysis. *Statistically significant trend; APC—annual percent change.
in men. The exception was a decreased trend in diabetes type 2 mortality rates both in men and women. The ASR of diabetes (25.5 per 100,000) in Serbia was two times higher than the average ASR in countries of European region (12.9 per 100,000) in 2012 [5]. In 2012, in Europe, high rates have been observed in Armenia (41.5 per 100.000) and the Mediterranean countries (Cyprus—32.8, Turkey—30.8, Israel—26.5, Portugal—23.0); the lowest rate (approximately—5.5) in United Kingdom, Finland and Greece [5]. Finally, the above mentioned mortality rate puts Serbia
on the fifth place among the countries with the highest diabetes mortality rates in Europe. Some unfavorable changes (economic crisis, poverty, civil wars, refugees, unemployment, disintegration of the country, NATO bombing) occurred in Serbia in the last decades, that have had a considerable impact on lifestyle as well as on the functioning of the health services in Serbia during the studied period (lack of drugs, public health weakness) [17,18]. The National Health Survey revealed that 18.3% of adult male population in Serbia were obese (body mass index ≥30) in 2006, significantly more
Table 2 – Trendsa in age-specific diabetes type 1 mortality rates, men and women in Serbia, in 1997–2015 period: a joinpoint regression analysis. Ageb
AAPC (95% CI)
Men 30–40 40–49 50–59 60–69 70–79 80+
0.9 (−3.7 to 5.6) 0.7 (−1.1 to 2.5) 3.1a (0.7–5.6) 5.3a (2.5–8.3) 6.8a (4.4–9.2) 11.9a (9.4–14.5)
Women 30–40 40–49 50–59 60–69 70–79 80+
−3.0 (−7.7 to 1.9) −1.8 (−5.5 to 2.1) −2.6a (−5.1 to −0.2) 1.1 (−1.5 to 3.7) 5.5a (3.4–7.6) 13.3a (9.4–17.3)
Trend 1
APC (95% CI)
Trend 2
APC (95% CI)
1997–2001 1997–2001 1997–2001
−13.4 (−28.4 to 4.8) −13.5 (−30.3 to 7.4) −10.5 (−25.2 to 7.1)
2001–2015 2001–2015 2001–2015
6.1a (3.2–9.2) 8.8a (5.4–12.3) 9.9a (7.0–12.8)
1997–2000 1997–2001
−23.8 (−45.6 to 6.6) −11.3 (−23.4 to 2.6)
2000–2015 2001–2011
3.9a (1.0–6.9) 11.4a (6.8–16.2)
Trend 3
2011–2015
APC (95% CI)
−4.8 (−17.7 to 10.1)
AAPC = average annual percent change; APC = annual percent change; CI = confidence interval. Statistically significant trend. b Joinpoint results are not shown for age subgroups <30, because count of diabetes type 1 deaths occurred in each of the decennium in any year was zero. a
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS
6
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
Table 3 – Trendsa in age-specific diabetes type 2 mortality rates, men and women in Serbia, in 1997–2015 period: a joinpoint regression analysis. Ageb
AAPC (95% CI)
Men 40–49 50–59 60–69 70–79 80+
−4.3a (−6.7 to −1.9) −2.9a (−4.5 to −1.3) −1.9a (−3.4 to −0.3) −1.1 (−2.3 to 0.1) 2.8a (1.5–4.1)
Women 40–49 50–59 60–69 70–79 80+
−5.6a (−8.2 to −3.0) −6.0a (−7.9 to −4.0) −4.4a (−5.7 to −3.0) −2.2a (−3.4 to −0.9) 2.3a (0.9–3.7)
Trend 1
APC (95% CI)
Trend 2
APC (95% CI)
1997–2010 1997–2010
0.7 (−1.1 to 2.6) 1.0 (−0.2 to 2.2)
2010–2015 2010–2015
−11.4a (−18.0 to −4.3) −9.0a (−13.5 to −4.3)
1997–2004 1997–2011 1997–1999 1997–2010
2.5 (−4.2 to 9.7) −2.3a (−3.6 to −1.0) 12.3 (−5.9 to 33.9) 4.3a (2.3–6.3)
2004–2015 2011–2015 1999–2010 2010–2015
−10.2a (−13.2 to −7.0) −15.9a (−23.0 to −8.2) −0.7 (−2.1 to 0.7) −5.2 (−12.7 to 2.9)
Trend 3
2010–2015
APC (95% CI)
−10.2a (−13.7 to −6.6)
AAPC = average annual percent change; APC = annual percent change; CI = confidence interval. Statistically significant trend. b Joinpoint results are not shown for age subgroups <40, because count of diabetes type 1 deaths occurred in each of the decennium in any year was zero. a
than in 2000 (17.4%), while the pre-obese category (body mass index = 25–29.9) was more often found among men (41.4%) than among women (29.1%) in 2013 [19]. In Serbia, more than one half of the population smoked in some part of their lives (51.8%) [19]. In Serbia, Sipetic´ et al. [20] found that infections during the 6 months preceding the onset of the disease and stressful events were significantly related to diabetes and supported the hypothesis that environmental factors play a role in the development of type 1 diabetes. Certainly, differences in the quality of data on cause of death across countries would have led to some of the effects that were found, but bias due to secular improvements in diabetes registration is unlikely to explain most of the observed changes [21]. Previously published studies on diabetes temporal trends have shown inconclusive results [22–24]. Large geographical variations and an increasing temporal trend in diabetes mortality are evident among most countries. Globally, total deaths from diabetes are projected to rise by more than 50% in the next 10 years [5]. In contrast to the reduction in diabetes mortality in majority of developed countries [25], the significant increase in diabetes mortality was observed in Serbia in last decades. Similar, an increased mortality of diabetes was revealed in many Central and Eastern European countries [26,27]. Differences in diabetes mortality rates between countries could in part be due to variations in incidence of diabetes, as well as in the variations in the availability and use of prevention and treatment. Evidence shows that type 1 diabetes incidence is increasing more steeply in some Central and Eastern European countries. Some authors noted the rising incidence of type 1 diabetes in Belgrade children aged 0–14 years in the period from 1982 to 2005 [28]. The increased frequency from diabetes is likely attributable to rapid economic development, rapid changes in lifestyle (dietary changes, reduced physical activity and unhealthy behaviors), an aging population, and a Westernized lifestyle in developing countries [29]. Some studies have proposed that the increasing early life exposure to infectious diseases, which typically occurred over the past several decades in developing countries, could be involved in this trend [30]. Besides, greater
responsibility and stress in times of wars and economic crisis placed more on men than on women could be an explanation for the more unfavorable mortality trend from diabetes in men in Serbia [31]. At the end of 1999, Serbia had almost 1 million refugees and internally displaced people, which made it the country with the highest number of refugees in Europe. Refugees came to Serbia in waves: the first big wave was in 1992, the second was in 1995 and the third was in 1999. In 1995, 400,000 people from Croatia were expelled and moved to Serbia in a very short period of about two weeks. In 1999, about 300,000 people from Kosovo moved to Serbia after NATO’s bombing [32]. Over the past decades, around 200,000 refugees have returned to their countries of origin or migrated to third countries, and majority of refugees decided to integrate in Serbia [32,33]. Although the number of refugees in Serbia is decreasing, their social status has been bad for years and the qualitative changes are slow, further was deteriorated by unemployment, long lasting life in collective centers with very poor sanitary conditions, lack of financial sources for food, medicines, hygiene, education, clothes etc. The United Nations High Commissioner for Refugees classified Serbia among five countries of ‘long-term refugees’ crisis. Unfortunately, in Serbia there are no precise indicators of the health status of refugees, which would enable comparison of health status before the war, during the war and after the war [34]. Numerous studies evidenced that diabetes is a problem in refugees populations of South and East Asian, Latin American, Middle East (Iraqi, Syrian) and African, Eastern Europe (countries of the former Soviet Union and Yugoslavia) origins, etc. [35–37]. But, there is no available mortality data for Serbian refugees for the previous period, even for those who still live in Serbia, nor for those who have left Serbia. Unfavorable trends in mortality from diabetes among men in Serbia could mainly be attributed to mortality from diabetes type 1. Also, sex differences in mortality of both diabetes type 1 and type 2, with higher risk in males than in females, have been previously reported in many studies [38]. This finding may suggest a role for sex-linked differences in function
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
or in insulin sensitivity, but involves complex social and cultural issues of gender [23]. Type 1 diabetes includes those cases currently ascribable to an autoimmune process and those for which etiology is unknown, while type 2 diabetes is caused by both environmental (such as physical inactivity, drugs and toxic agents, obesity, viral infection) and genetic factors [39]. The increased rates of deaths from diabetes type 1 might have been a consequence of unfavorable circumstances in Serbia in the observed period: the general standard was poor, the quality of health services was unsatisfactory (including the lack of vaccines, drugs, medical equipment, a large number of wounded as well as refugees, decreasing hospitalization rates, particularly for people aged 60 years and older), social disintegration, aging of the population, etc. [9]. Relatively favorable trend in mortality from diabetes type 2 may be attributed to the implementation of mass preventive programs (tobacco control, etc.) and improvements in diabetes treatment in Serbia [19]. Paradoxically, the decline trend in diabetes mortality in women in Serbia coincided with the global economic crisis. Therefore, these results should encourage the launching of epidemiological studies in order to identify future trends in diabetes mortality and to plan for health care delivery, especially in developing countries.
4.1.
Strengths and limitations of this study
This study provides the first nationwide estimates of diabetes mortality in Serbia over 25 years. The strength of our study lies in the fact that it is a population-based study, using quality death data with comprehensive coverage, completeness and comparability, with temporal trends assessed by joinpoint analysis. Thanks to the WHO assessment of the quality of data on the cause of death in Serbia as intermediate [12], we could give a good assessment of national mortality trend for diabetes and carry out a comparison with the pattern of diabetes mortality in other countries. Also, our study implies the need for effective measures of diabetes prevention and treatment in Serbia. However, there were several sources of limitations in this paper. We acknowledge that a longer study period may be better to more accurately assess mortality time-trends, but in Serbia there was no available data for this. Also, a limitation is that there are no separate data on diabetes deaths among refugees that can possibly confound the diabetes mortality pattern in Serbia: for internally displaced persons and refugees data were included in the Serbian population and could not be set aside as a special contingent. Certainly, a better quality deaths registration system, which is available in high-income countries, is needed in Serbia. Namely, mortality rates from diabetes may have been underestimated, because of the errors in coding the causes of death and recording the diabetes as antecedent, instead of underlying main cause of death, particularly in those who died from infarction, stroke, and renal diseases [40]. Also, there are no data of diabetes incidence for the entire study period to be used for explanation of temporal mortality trends (Diabetes Registry was implemented since 2006). The available literature does not have enough relevant information about the presence of risk factors for diabetes in the Serbian population. In the applied log-linear model in this study, dropping a particular year observations
7
with count of diabetes deaths being zero from the analysis may shift or affect the detection or locations of joinpoints affecting the analysis. Despite these limitations, this study helped to elucidate diabetes death trends in Serbia, which still need to be clarified in analytical epidemiological studies in the future.
Conflict of Interest The authors state that they have no conflict of interest.
references
[1] World Health Organization, WHO methods and data sources for global causes of death 2000–2012, in: Global Health Estimates, Technical Paper WHO/HIS/HSI/GHE/2014.7, World Health Organization, 2014. [2] L. Chen, D.J. Magliano, P.Z. Zimmet, The worldwide epidemiology of type 2 diabetes mellitus—present and future perspectives, Nat. Rev. Endocrinol. 8 (4) (2011) 228–236. [3] C.D. Mathers, D. Loncar, Projections of global mortality and burden of disease from 2002 to 2030, PLoS Med. 3 (11) (2006) e442, http://dx.doi.org/10.1371/journal.pmed.0030442. [4] WHO Global Infobase Team, The SuRF Report 2. Surveillance of Chronic Disease Risk Factors: Country-level Data and Comparable Estimates, WHO, Geneva, 2005, Available at: https://www.who.int/ncd surveillance/infobase/web/surf2/ (accessed 27.02.16). [5] World Health Organization, European Health for all Database, WHO Regional Office for Europe, Copenhagen, Denmark, 2016. [6] C.C. Patterson, E. Gyürüs, J. Rosenbauer, et al., Trends in childhood type 1 diabetes incidence in Europe during 1989–2008: evidence of non-uniformity over time in rates of increase, Diabetologia 55 (8) (2012) 2142–2147. [7] P.J. Liao, Z.Y. Lin, J.C. Huang, et al., The relationship between type 2 diabetic patients’ early medical care-seeking consistency to the same clinician and health care system and their clinical outcomes, Medicine (Baltimore) 94 (7) (2015) e554, http://dx.doi.org/10.1097/MD.0000000000000554. [8] Writing Group for the DCCT/EDIC Research Group, T.J. Orchard, D.M. Nathan, B. Zinman, et al., Association between 7 years of intensive treatment of type 1 diabetes and long-term mortality, JAMA 313 (1) (2015) 45–53. [9] C.C. Patterson, G. Dahlquist, G. Soltész, A. Green, EURODIAB ACE Study Group. Europe and Diabetes, Is childhood-onset type I diabetes a wealth-related disease? An ecological analysis of European incidence rates, Diabetologia 44 (Suppl. 3) (2001) B9–B16. [10] M. Ilic, I. Ilic, Gender disparities in mortality from infectious diseases in Serbia, 1991–2014: a time of civil wars and global crisis, Epidemiol. Infect. 144 (12) (2016) 2473–2484. [11] M. Ilic, I. Ilic, Suicide in Serbia, J. Affect. Disord. 193 (2016) 187–193. [12] C.D. Mathers, D. Ma Fat, M. Inoue, et al., Counting the dead and what they died from: an assessment of the global status of cause death data, Bull. World Health Organ. 83 (2005) 171–177. [13] Statistical Office of the Republic of Serbia, Statistical Yearbook of the Republic of Serbia for 1991–2014, Statistical Office of the Republic of Serbia, Belgrade, 2015. [14] H.J. Kim, M.P. Fay, E.J. Feuer, et al., Permutation tests for joinpoint regression with applications to cancer rates, Stat. Med. 19 (2000) 335–351.
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019
PCD-554; No. of Pages 8
ARTICLE IN PRESS
8
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 6 ) xxx–xxx
[15] L.X. Clegg, B.F. Hankey, R. Tiwari, et al., Estimating average annual per cent change in trend analysis, Stat. Med. 28 (2009) 3670–3682. [16] H.J. Kim, M.P. Fay, B. Yu, et al., Comparability of segmented line regression models, Biometrics 60 (2004) 1005–1014. [17] The Impact of UN Security Council Sanctions on the Health of the Population of FR Yugoslavia, Federal Institute for Public Health, Belgrade, Serbia, 1993. [18] S.J. Kunitz, The making and breaking of Yugoslavia and its impact on health, Am. J. Public Health 94 (2004) 1894–1904. [19] Ministry of Health, Republic of Serbia, National Health Survey, Serbia 2006, Ministry of Health, Republic of Serbia, Belgrade, 2007. ´ H.D. Vlajinac, N.I. Kocev, et al., The Belgrade [20] S.B. Sipetic, childhood diabetes study: a multivariate analysis of risk determinants for diabetes, Eur. J. Public Health 15 (2005) 117–122. [21] L.N. McEwen, A.J. Karter, J.D. Curb, et al., Temporal trends in recording of diabetes on death certificates: results from Translating Research Into Action for Diabetes (TRIAD), Diabetes Care 34 (7) (2011) 1529–1533. [22] G. Roglic, N. Unwin, Mortality attributable to diabetes: estimates for the year 2010, Diabetes Res. Clin. Pract. 87 (1) (2010) 15–19. [23] M.Ø. Krag, L. Hasselbalch, V. Siersma, et al., The impact of gender on the long-term morbidity and mortality of patients with type 2 diabetes receiving structured personal care: a 13 year follow-up study, Diabetologia 59 (2) (2016) 275–285. [24] W.H. Lin, M.C. Wang, W.M. Wang, et al., Incidence of and mortality from type I diabetes in Taiwan from 1999 through 2010: a nationwide cohort study, PLoS One 9 (1) (2014) e86172, http://dx.doi.org/10.1371/journal.pone.0086172.eCollection. [25] S. Wild, G. Roglic, A. Green, et al., Global prevalence of diabetes: estimates for the year 2000 and projections for 2030, Diabetes Care 27 (2004) 1047–1053. [26] P.E. Wändell, A.C. Carlsson, Time trends and gender differences in incidence and prevalence of type 1 diabetes in Sweden, Curr. Diabetes Rev. 9 (4) (2013) 342–349. [27] J.M. Lawrence, G. Imperatore, D. Dabelea, et al., Trends in incidence of type 1 diabetes among non-Hispanic white youth in the U.S., 2002–2009, Diabetes 63 (11) (2014) 3938–3945. [28] S. Sipetic, J. Maksimovic, H. Vlajinac, et al., Rising incidence of type 1 diabetes in Belgrade children aged 0–14 years in the period from 1982 to 2005, J. Endocrinol. Investig. 36 (5) (2013) 307–312.
[29] W.P.T. James, R. Jackson-Leach, C. Ni Mhurchu, et al., Overweight and obesity (high body mass index), in: M. Ezzati, A. Lopez, A. Rodgers, C.J.L. Murray (Eds.), Comparative Quantification of Health Risks: Global and Regional Burden of Disease Attributable to Selected Major Risk Factors, WHO, Geneva, 2004. [30] J.F. Bach, The effect of infections on susceptibility to autoimmune and allergic diseases, N. Engl. J. Med. 347 (2002) 911–920. [31] S. Thurner, P. Klimek, M. Szell, et al., Quantification of excess risk for diabetes for those born in times of hunger, in an entire population of a nation, across a century, Proc. Natl. Acad. Sci. U. S. A. 110 (12) (2013) 4703–4707. [32] Commissariat for Refugees and Migration of the Republic of Serbia: Process of Voluntary Repatriation of Refugees to the Republic of Croatia and Bosnia and Herzegovina, United Nations High Commissioner for Refugees (UNHCR), Belgrade, 2010, Available at: http://www.kirs.gov.rs/ (accessed 07.08.16). [33] V. Lukic, Two Decades of Refuge in Serbia: Census of Population, Households and Dwellings 2011 in the Republic of Serbia, Statistical Office of the Republic of Serbia, Belgrade, 2015. [34] S. Litvinjenko, Migration and health, Srp. Arh. Celok. Lek. 125 (1997) 191–196 (Article in Serbian). [35] M. DesMeules, J. Gold, S. McDermott, et al., Disparities in mortality patterns among Canadian immigrants and refugees, 1980–1998: results of a national cohort study, J. Immigr. Health 7 (4) (2005) 221–232. [36] J. Wagner, S.M. Berthold, T. Buckley, et al., Diabetes among refugee populations: what newly arriving refugees can learn from resettled Cambodians, Curr. Diabetes Rep. 15 (8) (2015) 56, http://dx.doi.org/10.1007/s11892-015-0618-1. [37] F.J. Mateen, M. Carone, H. Al-Saedy, et al., Medical conditions among Iraqi refugees in Jordan: data from the United Nations Refugee Assistance Information System, Bull. World Health Organ. 90 (6) (2012) 444–451. [38] G. Bruno, M. Maule, F. Merletti, et al., Age-period-cohort analysis of 1990–2003 incidence time trends of childhood diabetes in Italy: the RIDI study, Diabetes 59 (9) (2010) 2281–2287. [39] E. Adeghate, P. Schattner, E. Dunn, An update on the etiology and epidemiology of diabetes mellitus, Ann. N. Y. Acad. Sci. 1084 (2006) 1–29. [40] N.J. Morrish, S.L. Wang, L.K. Stevens, et al., Mortality and causes of death in the WHO Multinational Study of vascular disease in diabetes, Diabetologia 44 (2001) S14–S21.
Please cite this article in press as: M. Ilic, I. Ilic, Diabetes mortality in Serbia, 1991–2015 (a nationwide study): A joinpoint regression analysis, Prim. Care Diab. (2016), http://dx.doi.org/10.1016/j.pcd.2016.08.019