Time series modeling to predict mortality rates and estimate years of life lost due to injuries in West Azerbaijan province, Iran, 2013–2024

Abstract

Background

Injuries remain major causes of preventable mortality and premature loss of life. This study aimed to examine 12-year trends in mortality and years of life lost (YLL) due to injuries in West Azerbaijan Province, Iran, and to forecast future mortality patterns.

Methods

This registry-based retrospective observational study investigated the mortality patterns and YLL due to injuries. Data were extracted from the death registration and classification system of the Iranian Ministry of Health and Medical Education. Indicators including crude mortality rates, Age-Standardized Rates (ASR), and YLL were calculated. Future trends were predicted using ARIMA modeling (313).

Results

A total of 20,003 deaths due to injuries were recorded over the 12-year period. The majority of cases were men (76%) and occurred in urban areas (59.7%). The annual number of deaths increased by 37.94%, rising from 1,389 in 2013 to 1,916 in 2024. Traffic accidents remained the primary cause of death (increasing by 21.35%), while suicide rates showed a significant surge of 141.50%. The total YLL during the study period reached 473,700 years (17 per 1,000 in males and 6 per 1,000 in females). The highest burden of YLL was observed in the 15–29-year age group. Prediction models suggest that mortality will continue to fluctuate between 120 and 182 deaths per month over the next three years.

Conclusion

Injuries, particularly traffic accidents and the rising trend of suicide, impose a heavy and increasing burden of premature mortality in West Azerbaijan. The high YLL among adolescents and young adults underscores the critical need for urgent preventive strategies, improved road safety measures, and targeted mental health interventions to mitigate this public health crisis.

Introduction

Injuries remain a major public health concern and continue to account for a considerable share of preventable death and disability worldwide [1]. In settings with limited preventive infrastructure, the burden is often amplified by unsafe roads, seasonal hazards, and unequal access to timely trauma care [2]. In Iran, injury-related mortality has remained an important health problem for years [3].

With a death rate of 20 per 100,000 population due to traffic accidents, Iran is among the countries with the highest death rate, and over 84,000 people lost their lives in road accidents in Iran from 2013 to 2019, a level that is still higher than many countries and reflects the continuing public health importance of external causes of injury [4]. At the provincial level, injury patterns may differ substantially because geography, climate, transport conditions, and population structure are not uniform across the country [5].

West Azerbaijan Province, located in northwest Iran, is one of the country’s 31 provinces, sharing borders with Turkey, Armenia, Azerbaijan, and Iraq, as well as neighboring East Azerbaijan, Zanjan, and Kurdistan provinces. The provincial capital is Urmia, and the region comprises 16 additional counties. With a population of 3,265,219 as of the 2016 national census, the province provides a relevant setting for investigating injury-related health issues [6, 7]. Its diverse terrain, seasonal climatic variations, and combination of urban and rural communities may affect the incidence and severity of injuries. Local data further highlight the epidemiological significance of injury-related conditions, emphasizing the need for context-specific analyses [8].

Previous Iranian studies have mostly described injury mortality in a cross-sectional manner or focused on limited subgroups such as road traffic injuries [9]. Although useful, these studies do not fully explain how mortality and premature death change over time at the provincial level. In particular, evidence is still limited on long-term forecasting of both injury mortality and Years of Life Lost in one framework [10].

Years of Life Lost is especially important because it reflects not only how many people die, but also how early they die [11]. This is critical for injuries, since many deaths occur in adolescents and young adults who otherwise would have had many remaining years of life. For this reason, Years of Life Lost provides policy makers with information that mortality rates alone cannot capture, including loss of productive years, social consequences for families, and the broader economic impact on the community [12].

Including Years of Life Lost in a forecasting model adds another layer of interpretation. Two provinces may have similar death counts, yet the one with more deaths at younger ages will carry a much larger burden of premature mortality. This makes Years of Life Lost a more policy-relevant measure for identifying where prevention efforts are likely to have the greatest impact [13, 14].

Time-series methods are appropriate for injury mortality data because such outcomes often show trend, autocorrelation, and seasonal fluctuation [15]. Among these methods, Seasonal Autoregressive Integrated Moving Average models are useful because they can capture both long-term trend and seasonal patterns within a single model [16]. This makes SARIMA a reasonable approach for short-term forecasting when regular monthly data are available and recurring temporal variation is expected [17].

Despite the value of these methods, few studies in Iran have simultaneously examined injury mortality and Years of Life Lost at the provincial level using time-series forecasting [18]. This gap is particularly important in West Azerbaijan Province, where local injury patterns may differ from national averages. Therefore, the present study aimed to describe temporal trends in injury mortality and Years of Life Lost, and to forecast future mortality patterns using SARIMA models. The main research questions were: what were the time trends in injury mortality and Years of Life Lost in West Azerbaijan Province from 2013 to 2024, which age and sex groups carried the greatest burden, and how well can SARIMA models predict future mortality patterns?

Methods

Study design and setting

This registry-based retrospective observational study was conducted using death registry data from 2013 to 2024. The study focused on all registered deaths attributed to external causes of morbidity and mortality (Injuries) across urban and rural areas of West Azerbaijan province, Iran, during the study period.

West Azerbaijan Province is located in northwestern Iran, with Urmia serving as its capital city. Covering an area of approximately 37,059 km² (excluding Lake Urmia), the province ranks as the 13th largest in Iran, comprising 2.25% of the country’s total landmass. According to the 2006 census, the population of West Azerbaijan was 2,873,459, accounting for 4.08% of Iran’s total population, making it the 8th most populous province in the country. Regarding the urban-rural distribution, 68% of the population resides in urban areas, while 32% inhabit rural settlements. Geographically, the province is characterized by its mountainous terrain.

Data sources

Mortality data were extracted from the electronic population-based death registration system. For injuries, the Legal Medicine Organization (Forensic Medicine) served as the primary and most critical data source. To ensure maximum sensitivity and completeness, this data was integrated with records from hospitals, cemeteries, and both urban and rural health centers.

Data integration process

In West Azerbaijan province, a centralized mortality registration unit operates with trained personnel responsible for entering and managing death records within the electronic registry system. All deaths are reported separately to this unit by the Legal Medicine Organization, hospitals, cemeteries, and urban and rural health centers using a standardized national reporting form.

Upon receipt of records, data integration is performed in a structured, multi-step process. First, records are matched using the National Identification Number as a unique identifier to detect and remove duplicate entries. Subsequently, registry staff conduct a detailed review of the submitted forms to resolve inconsistencies and address missing or incomplete data.

The electronic registry system is directly linked to the national civil registration database, enabling automatic verification of identification numbers and retrieval of basic demographic information. The system does not permit the entry of duplicate identification numbers, thereby reducing the risk of repeated records and enhancing data accuracy.

Data completeness assessment and quality assurance

In Iran, mortality registration is conducted through an integrated national system coordinated by the Ministry of Health and Medical Education, in collaboration with the National Organization for Civil Registration and the Legal Medicine Organization. Given the mandatory forensic investigation of all deaths under Iranian law, the electronic registry system has high coverage. Also, an independent quality control officer conducts periodic random audits of death records, cross-validating electronic registry data against the corresponding paper-based source documents to assess accuracy and maintain data quality standards.

Inclusion and exclusion criteria

Inclusion criteria

All registered deaths attributed to external causes (Injuries) among residents of West Azerbaijan province during the study period (2013–2024), regardless of age or sex.

Exclusion criteria

Records identified as duplicates, deaths due to natural causes (non-external), and cases where documentation was insufficient to confirm the specific underlying cause of injury or accident.

Coding of causes of death

Causes of death were coded according to the International Classification of Diseases, 10th Revision (ICD-10). Specifically, codes V01–Y89 (External causes of mortality) were utilized. This range encompasses transport accidents, falls, drownings, burns, poisonings, and intentional self-harm or assault. A multi-layer verification process was employed to cross-reference registry data with forensic documentation to minimize potential misclassification.

In this study, place of death refers to the location where the death occurred and was officially recorded in the mortality registry. Also, traffic accidents include all fatal road traffic-related incidents involving car crashes, motorcycle accidents, bicycle accidents, and pedestrian injuries.

Statistical analysis

First, using graphs and descriptive statistics, the trend in the mortality ratio was illustrated in relation to demographic variables. The subsequent formula was employed to determine the YLL as a result of deaths caused by injuries [19]:

Where N is the number of deaths at a certain age and gender. L is the standard of living of the deceased at the same age and gender. r is the Discounting Rate, which is equal to 0.03. β is the contract rate in calculating the age value, equal to 0.04. C is a modified constant value equal to 0.1658. a is the age at which death occurred, and e denotes Euler’s number (≈ 2.718). The main idea behind the YLL calculation is to estimate the total number of years that are prematurely lost due to early death by considering the age at death, the expected remaining years based on life expectancy, and adjustments for factors like quality of life and standardization.

The formula begins with the number of deaths at each age, multiplies that by the expected remaining years of life for those individuals, incorporates adjustments for economic and demographic variables using standard constants, and then applies a discount rate to reflect the present value of future years lost. To ensure standardization, the standard population data from 2013 for low- and middle-income countries was used. Age-specific standard mortality rates for males and females in the Asian population were used to calculate the expected number of deaths in each age group by multiplying the relevant mortality rate by the corresponding population size [20]. Time series models with monthly time intervals, including 132 months, were used to analyze the data. The data was initially verified for accuracy, and any necessary adjustments were made. Moreover, statistical methods were utilized to exclude or modify outliers. Box-Jenkins and Autoregressive Integrated Moving Average (ARIMA) models were employed to achieve this objective. The Box–Jenkins approach was implemented in four sequential stages. First, during the identification stage, the time series was visually inspected and the autocorrelation function (ACF) and partial autocorrelation function (PACF) plots were evaluated to detect patterns of trend, seasonality, and stationarity; differencing was performed when required. Second, in the estimation stage, several candidate ARIMA models were fitted and their parameters were estimated. Third, for diagnostic checking, residual plots and autocorrelation analyses were used to assess model adequacy and to verify that residuals behaved as white noise, while alternative models were compared using standard goodness-of-fit measures. Finally, in the forecasting stage, the best-performing model was selected to predict mortality trends. The process of modeling with Box-Jenkins and ARIMA models involves preparing a series, identifying the primary model using parameter estimation, accurately determining the model parameters, and selecting the appropriate model. Minitab Statistics software version 17 was used for data analysis in this study.

Ethical statement

Ethical approval was obtained from the Research Ethics Committees of Urmia University of Medical Sciences (Approval ID: IR.UMSU.REC.1404.292). All procedures followed relevant ethical standards and institutional guidelines for retrospective data.

Results

injuries resulted in 20,003 deaths in West Azerbaijan Province over a 12-year study period from 2013 to 2024. Out of all cases, cities were where 59.7% (11,940 cases) took place, villages were where 37.3% (7,467 cases) took place, and there were no places noted in 3% (596 cases). The majority of deaths occurred in men, with 76% (15,211 cases), and the percentage of women was 24% (4,792 cases).

The number of deaths increased from 1,389 in 2013 to 1,916 in 2024, resulting in an overall increase of 37.94%. The mean age at death was 36.9 ± 21.2 years in 2013 and increased to 39.0 ± 19.3years in 2024. With an overall increase of 21.35%, traffic accidents remained the primary cause, while accidental damage increased by 57.18% overall. There were significant increases in suicide, 141.50%, and legal intervention/military operations, 384.62%, while external causes experienced a considerable decrease, 87.50%. Suicide rates were higher among females (18.6% to 28.2%) compared to males, with percentages varying by sex (Table 1).

Table 1.

The number of deaths due to injuries by cause, sex, age, and percent change in West Azerbaijan province, 2013-2024

	Death, 2013			Death, 2024			Change from 2013 to 2024(%)
	Male N (%)	Female N (%)	Both N (%)	Male N (%)	Female N (%)	Both N (%)
Age	37.2 ± 19.9	36.2 ± 24.2	36.9 ± 21.2	38.7 ± 18.4	39.7 ± 22.4	39.0 ± 19.3	5.69
Traffic Accident*	541(53.5)	143(37.9)	684(49.2)	633(43.8)	197(41.7)	830(43.3)	21.35
Car crashes	406 (75.0)	106 (74.1)	512 (74.9)	581 (91.8)	179 (90.9)	760 (91.6)	48.4
Motorcycle accidents	110 (20.3)	35 (24.5)	145 (21.2)	21 (3.3)	12 (6.1)	33 (4.0)	-77.2
Bicycle accidents	22 (4.1)	2 (1.4)	24 (3.5)	23 (3.6)	1 (0.5)	24 (2.9)	0.0
Pedestrian injuries	3 (0.6)	0 (0.0)	3 (0.4)	8 (1.3)	5 (2.5)	13 (1.6%)	333.3
Accidental Damage**	240(23.7)	108(28.6)	348(25.1)	430(29.8)	117(24.8)	547(28.5)	57.18
Suicide	77(7.6)	70(18.6)	147(10.6)	222(15.4)	133(28.2)	355(18.5)	141.50
Complications of medical and surgical care	9(0.9)	8(2.1)	17(1.2)	4(0.3)	4(0.8)	8(0.4)	-52.9
Murder	73(7.2)	22(5.8)	95(6.8)	72(5.0)	18(3.8)	90(4.7)	-5.26
Consequences of external causes of illness and death	22(2.2)	18(4.8)	40(2.9)	4(0.3)	1(0.2)	5(0.3)	-87.50
Event of undetermined intent	50(5)	8(2.2)	58(4.1)	79(5.4)	2(0.4)	81(4.2)	39.66
Total	1012	377	1389	1444	472	1916	37.94
*Total of Car crashes, Motorcycle accidents, Bicycle accidents, and Pedestrian injuries

**Falls from a height, exposure to smoke, fire, and flames, drowning or immersion in water, contact with heat and hot materials

The crude mortality rate observed in this study was 7 per 10,000 for males, 2 per 10,000 for females, and 5 per 10,000 overall. Age-standardized rates per 100,000 population rose from 6.0 (95% CI: 5.8–6.1) in 2013 to a peak of 8.3 (95% CI: 8.1–8.5) in 2023 for both sexes, before declining slightly to 7.8 (95% CI: 7.6-8.0) in 2024. The increase in males was constant at 8.6 (95% CI: 8.0-9.1) to 11.5 (95% CI: 10.9–12.1), while female rates fluctuated between 2.8 and 4.0. Total YLL reached 473,700 (349,688 in males, 124,012 in females), with a mean YLL per 1,000 population of 17 in males, 6 in females, and 12 overall; annual YLL per 1,000 peaked at 20 (males) and 7 (females) in 2023 (Table 2).

Table 2.

Crude and standardized mortality rate (per 100,000 population) and YLL due to injuries by sex and year in West Azerbaijan province, 2013–2024

Year	No. death			Crude mortality rate Per (10’000)			ASR (95%CI) Per (100’000 population)			No. YLL			YLL Per (1000)
	Male	Female	Both sex	Male	Female	Both sex	Male	Female	Both sex	Male	Female	Both sex	Male	Female	Both sex
2013	1012	377	1389	6	2	4	8.6(8.0-9.1)	3.3(3.0-3.6)	6.0(5.8–6.1)	24,799	10,845	35,644	16	7	11
2014	1084	377	1461	7	2	5	9.1(8.6–9.7)	3.3(2.9–3.6)	6.2(6.1–6.4)	26,233	10,086	36,319	16	6	11
2015	1160	387	1547	7	2	5	9.7(9.1–10.3)	3.4(3.0-3.7)	6.6(6.4–6.8)	26,036	10,855	36,891	16	7	11
2016	1223	390	1613	7	2	5	10.2(9.6–10.7)	3.4(3.0-3.7)	6.8(6.6-7.0)	29,000	10,542	39,542	17	7	12
2017	1267	347	1614	8	2	5	10.4(9.8–11.0)	3.0(2.7–3.3)	6.8(6.6–6.9)	29,833	9526	39,359	18	6	12
2018	1195	389	1584	7	2	5	9.7(9.2–10.3)	3.3(3.0-3.6)	6.6(6.4–6.7)	27,142	10,259	37,401	16	6	11
2019	1293	422	1715	8	3	5	10.5(9.9–11.1)	3.6(3.2–3.9)	7.1(6.9–7.3)	30,536	11,226	41,762	18	7	12
2020	1145	340	1485	7	2	4	9.2(8.7–9.8)	2.8(2.5–3.2)	6.1(6.0-6.2)	27,478	9150	36,628	16	5	11
2021	1303	391	1694	7	2	5	10.5(9.9–11.1)	3.3(2.9–3.6)	6.9(6.8–7.1)	29,337	9466	38,803	17	6	11
2022	1508	422	1930	9	2	6	12.1(11.5–12.7)	3.5(3.2–3.8)	7.9(7.7-8.0)	32,577	9827	42,404	19	6	12
2023	1577	478	2055	9	3	6	12.6(12.0-13.2)	4.0(3.6–4.3)	8.3(8.1–8.5)	34,670	10,981	45,651	20	6	13
2024	1444	472	1916	8	3	5	11.5(10.9–12.1)	3.9(3.5–4.2)	7.8(7.6-8.0)	32,047	11,249	43,296	18	6	12
Total	15,211	4792	20,003	7	2	5	10.3(10.2–10.5)	3.4(3.3–3.5)	6.9(6.8-7.0)	349,688	124,012	473,700	17	6	12

The burden of YLL due to injuries was highest in the 15–29-year age range, with peak values in the 15–19 and 20–24-year groups, and was consistently higher among males than females in almost all age categories (A).

The main age groups for suicide were those aged 15–34, and there were significant contributions from the 15–19, 20–24, and 25-29-year categories. For both males and females over 50 years old, suicide-related YLL was minimal. In a study of 731 female suicides (ages 15–35), 51% were rural residents and 49% were urban. The most common demographic characteristics were high school education (26%) and homemaking as a primary occupation (56%) (B).

Young and middle-aged adults, particularly those in the 15–19, 20–24, 25–29, and 30–34-year age groups, were most affected by accidental damages, with the biggest impact on the YLL 1–4 age group (C).

Males had the greatest percentage of YLL due to traffic accidents, with a high value in the 15–19, 20–24, and 25–29-year age groups, followed by a gradual decrease in older ages. The age pattern for female traffic accident YLL was similar but at a smaller magnitude, indicating that adolescents and young adults are at greater risk (D) (Fig. 1).

Fig. 1.

The total number of YLL due to injuries, suicide, accidental damages, and traffic accidents in West Azerbaijan Province by gender in age groups, 2013-2024

Trends in injuries, traffic accidents, suicides, and accidental damages were predicted for an interval of 36 months after the observations. Prediction with ARIMA (313) suggests that the total number of deaths from injuries will range between 120 and 182 per month in the next three years (A). The number of people who die due to traffic accidents is predicted to fluctuate between 43 and 78 per month, based on the ARIMA (312) model (B). Furthermore, prediction with ARIMA (111) shows the number of suicide deaths will fluctuate between 26 and 35 per month (C). In the end, there will be a variation of 36 to 49 deaths from accidental damage per month, based on the prediction with ARIMA (415) (D). After all the models were fitted (for injuries, traffic accidents, suicides, and accidental damages), the partial autocorrelation function graph confirmed that the selected model was suitable (Fig. 2). Moreover, the correctness of the selected models was confirmed by verifying their assumptions, including the normality of residuals, the uncorrelatedness of residuals, the stability of residual variance, and the randomness of residuals (Supplementary File).

Fig. 2.

The total time series of death due to accidents and incidents, suicide, and accidental damages with their partial autocorrelation in West Azerbaijan Province, 2013–2024

Discussion

The findings of this study indicate that mortality due to injuries in West Azerbaijan province during the period from 2013 to 2024 was associated with an overall increase of 37.94%. The number of deaths rose from 1,389 cases in 2013 to 1,916 cases in 2024, and the age-standardized rate (ASR) increased from 6.0 to 7.8 per 100,000 population. This upward trend is consistent with national studies in Iran; Taherian and colleagues, in their analysis of the trend of traffic accident mortality in Iran, reported that after a decade of decline, the death rate began to rise again from 2017 and continued until 2022 [10]. Globally, the WHO, in its Global Status Report on Road Safety 2023, announced that road traffic deaths have reached 1.19 million per year, with over 92% of deaths occurring in low- and middle-income countries (LMICs) [21]. This pattern aligns with the observed increase in West Azerbaijan province as part of a middle-income country. Furthermore, Kisa and colleagues, in their Global Burden of Disease 2017 study, emphasized that road injuries are one of the leading causes of YLL worldwide and predominantly affect young males. This finding is consistent with the total YLL burden in our study (473,700 years) [22]. This increase in mortality indicates the ineffectiveness of preventive policies at the provincial level, particularly in rural areas where 37.3% of deaths occurred. Focusing on socio-economic factors such as infrastructure development could reverse this trend, as LMICs like Iran often face similar challenges that require coordinated national and international interventions.

Traffic accidents remained the leading cause of death in this study, accounting for approximately 43–49% of fatalities, with a 21.35% increase over the study period. They generated the highest YLL in the 15–29 age group. Godthelp and colleagues have shown that the road death rate in LMICs is three times higher than in high-income countries, and over half of the deaths involve pedestrians, cyclists, and motorcyclists [23]. This pattern aligns with the concentration of deaths in urban (59.7%) and rural areas observed in our study. Weaver and colleagues, in their GBD analysis, reported that unintentional injuries, including road traffic injuries, impose a high burden of YLL in developing regions and predict that without interventions, this burden will increase by 2030 [24]. Furthermore, Abbasi and colleagues examined the burden of transport injuries in western provinces and reported a significant increase in West Azerbaijan, a finding consistent with the fluctuation in the crude death rate (from 4 to 5 per 10,000 population) in our study [25].Therefore, traffic accidents not only impose a substantial economic burden but also lead to the loss of young labor force. This has a more pronounced impact in provinces like West Azerbaijan with an agriculture-based economy. I suggest that focusing on safety education and road improvements could reduce YLL.

An important finding of the present study was the high number of deaths among children aged 1–4 years in both boys and girls, which was largely attributable to accidents and unintentional injuries [37]. This pattern indicates that early childhood injuries constitute a considerable share of preventable mortality in the province. Similar evidence has shown that, among children aged 1–4 years, traffic injuries and drowning are the leading causes of injury death, followed by falls, burns, and poisoning [26]. The high burden observed in our study may reflect inadequate child safety measures at home and in public environments, insufficient parental awareness, and limited implementation of injury-prevention programs. This finding highlights the urgent need for targeted interventions such as parental education, safer home environments, child-focused traffic safety policies, and strengthened community-based prevention programs. Because mortality in this age group leads to a substantial number of years of life lost, preventive strategies in early childhood should be considered a major public health priority.

One of the concerning findings of this research is the significant 141.5% increase in deaths from suicide, particularly the rise in the proportion of women from 18.6% to 28.2%, with the peak YLL occurring in the 15–34 age group. Hushmandi and colleagues, in forecasting the suicide trend in Iran, have highlighted an increase among women and youth, identifying socio-economic factors as key drivers [27]. Globally, Weaver and colleagues, in their GBD study, examined the global burden of suicide, which accounts for over 700,000 annual deaths, with 73% occurring in LMICs. They noted a higher rate among men but an increase in some female subgroups, a pattern consistent with the gender disparity observed in our study [24]. Bertuccio and colleagues, in their analysis of youth suicide trends from 1990 to 2020, reported an overall global decline but emphasized regional fluctuations, particularly in the Middle East [28]. Furthermore, the WHO, in its 2021 global suicide report, estimated that suicide is the third leading cause of death among individuals aged 15–29 years, aligning with the concentration of YLL among youth in our study [29]. This rise in suicide indicates a mental health crisis among women and youth, potentially linked to cultural and economic factors. Psychological support programs, such as online counseling, could help control this trend, as global studies have demonstrated the success of such interventions in reducing YLL.

Notably, the increase in suicide among females aged 15–35 years represents one of the most significant findings of the present study. The concentration of YLL within adolescent girls and young adult women suggests the presence of age-specific psychosocial vulnerabilities that may differ from those observed in males or older women. Factors such as unemployment, educational pressures, family conflicts, gender-related social expectations, early marriage, domestic violence, and limited access to mental health services may contribute to this increasing trend in LMIC settings [30]. Therefore, conducting more detailed age-specific analyses among adolescent females (15–19 years), young adults (20–24 years), and women aged 25–35 years could provide more precise evidence for designing targeted suicide prevention programs. Such interventions may include school-based mental health education, university counseling services, women-centered psychosocial support programs, and early identification of high-risk individuals within primary healthcare systems. In addition, several qualitative studies in Iranian settings have emphasized the role of family problems, marriage and love, social stigma, pressure of high expectations, poverty, and other sociocultural pressures in suicidal behaviors among women [31, 32]. However, further qualitative and mixed-methods research may still be necessary to better understand the underlying cultural and psychosocial determinants of this increasing trend in West Azerbaijan province.

The predominance of mortality among males (76%) and the concentration of the highest YLL in the 15–29 age group is a pattern consistent with numerous national and global studies. Farzadfar, in their analysis of the disease burden in Iran 2022, has shown that unintentional injuries, particularly road traffic injuries, impose a high YLL burden on young males [33]. Globally, Lou and colleagues, in their GBD study, indicated a slight global decline in the DALY rate due to unintentional injuries but emphasized that progress is slower in LMICs, with the primary burden falling on the youth [34]. Similarly, Yan and colleagues, in their GBD trend analysis, reported a decrease in suicide deaths among youth but noted that the burden remains high in LMICs [35]. The gender and age differences in mortality appear to reflect the role of high-risk behaviors among men and social pressures compared to women. Therefore, focusing on gender-specific prevention programs could reduce YLL.

The 87.5% decrease in mortality from other external causes may be related to improvements in data recording or changes in classification. However, the overall YLL burden in this study aligns with national and global estimates. A study on the burden of falls among the elderly in Iran revealed provincial inequalities, with western provinces bearing a higher burden [36]. Globally, the 2021 GBD study identified unintentional injuries as one of the leading causes of death in developing regions [37]. ARIMA model-based forecasts in this study indicate continued fluctuations in monthly mortality, ranging from 120 to 182 cases for total deaths, highlighting the necessity for preventive interventions. Therefore, ARIMA models can be utilized for planning, and integration with global systems such as GBD can enhance accuracy.

The increase in the mean age at death from 36.9 to 39.0 years may indicate demographic shifts or improved survival rates, yet the primary burden remains on the youth. Alicandro and colleagues, in their analysis of global suicide trends, reported a decrease among older age groups but noted that the burden remains high among youth in LMICs [38]. Compared to neighboring countries, the burden of road traffic injuries in Iran is higher, and Segui-Gomez and colleagues have highlighted the slow progress in road safety within LMICs [39]. Finally, this study underscores the necessity for targeted prevention programs, road safety education, psychological support, and improved data recording, which could reduce the YLL burden [40]. Consequently, age-related changes indicate population aging, but focusing on youth remains critical. Integrating national policies with global experiences can reduce the overall burden, with success in high-income countries serving as a model for LMICs.

Strengths and limitations

One of the major strengths of this study is the long-term evaluation of mortality trends due to accidents and external causes over a 12-year period at the provincial level, combined with the calculation of YLL and time-series forecasting using ARIMA models. The use of population-based registry data also provided broad coverage of mortality events across both urban and rural areas of West Azerbaijan province. Furthermore, the stratification of mortality by age, sex, cause of death, and geographic location enabled a more detailed understanding of vulnerable population groups and temporal trends.

However, several limitations should also be considered when interpreting the findings. First, this study relied on registry-based mortality data, and such data may be affected by underreporting, incomplete records, delayed registration, or misclassification of causes of death. Second, changes in coding practices or improvements in data recording systems over time may have influenced some of the observed trends, especially regarding other external causes. Third, the study lacked access to detailed socio-economic, psychological, and behavioral variables that could further explain the observed increase in suicide and accident-related mortality. In addition, qualitative information regarding contextual risk factors, family conditions, and mental health determinants was not available in the registry system. Finally, because this study was conducted at the provincial level, caution should be exercised when generalizing the findings to other provinces or national populations with different demographic and socio-economic characteristics. Despite these limitations, the study provides valuable epidemiological evidence for policymakers and public health planners regarding the burden of injuries and external causes of mortality in West Azerbaijan province.

Conclusion

This study demonstrated that mortality due to injuries in West Azerbaijan province exhibited a concerning increasing trend during the study period, with traffic accidents remaining the leading cause of death, while suicide, particularly among women, has emerged as a significant and growing factor. The burden of YLL primarily affected young males, a pattern consistent with national trends in Iran and the high burden of unintentional injuries in low- and middle-income countries. These findings underscore the profound impact of socio-economic, infrastructural, and mental health factors on premature deaths, and forecasts also indicate continued fluctuations in mortality in the coming years, highlighting the necessity for immediate and coordinated action. Consequently, the burden of mortality and YLL due to accidents and injuries in West Azerbaijan province is not only a public health issue but also a serious developmental challenge that threatens the productive force of society. Therefore, it is proposed that road accident prevention programs be strengthened through improved road infrastructure, stricter enforcement of laws, and targeted educational campaigns; mental health services, especially for women and youth, be expanded to include accessible counseling centers and regular screening programs; the mortality registration system be enhanced with modern technologies to enable more effective monitoring; community-based interventions in rural areas be implemented for home safety education and improved emergency access; longitudinal and qualitative research be conducted to identify local risk factors; and inter-sectoral collaboration among relevant organizations be strengthened to formulate a comprehensive prevention plan. Through the coordinated implementation of these measures and continuous monitoring, the current trend can be reversed, and the burden of premature deaths and Years of Life Lost can be significantly reduced, ultimately contributing to the enhancement of public health and sustainable development in the province.

Authors’ contributions

Research idea and study design: NA, MZ; data acquisition: ER, FB; data analysis/ interpretation: MN, FB; statistical analysis: MN; supervision or mentorship: NZ, MZ. the manuscript draft writing: KM, SM. All authors revised and approved the draft of the manuscript.

Funding

Not Applicable.

Data availability

The datasets generated and/or analysed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Declarations

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.