Original Article

Exploring Change in Trend of Homicide Incidence Rate in Iran from 2006 to 2016: Applying Segmented Regression Model

Introduction

Murder or homicide is the act of killing a person by another to cause death or serious harm to her/him ¹ . In other words, homicide means that someone is being killed illegally ² . According to the WHO report, about 470,000 homicides occur each year in the world ³ . Every year about 200,000 murders occur among youth 10-29 yr which accounts for about 43% of all homicides that is the fourth leading cause of death for people in this age group. About 84% of homicides occur in males aged 10-29 years. Therefore, the incidence rate of homicide among men was more than women around the world. During 2000-2012 incidence rate of homicide among youth has decreased in most of the countries. However, this decrease is mostly related to high and middle-income countries ⁴ . The highest incidence rate of homicide was in El Salvador, Honduras, Venezuela and Colombia, respectively ⁵ .

Moreover, homicide in Iran occurred in 3630 people (4.1 per 100,000) in 2015 that it was more common in men (6.8 per 100,000) compared to women. While women more than 60 yr old had highest incidence rate of homicide (6.5 per 100,000) ⁶ . In Iran in 2019, about 80.3% of the victims were men and the mean age of victims were 32.4 year. Besides, the most root cause of murder was quarrel and most of murders were carried out using a weapon like firearms (46%) and stab wounds (29%), respectively ⁷ .

According to the report of variety studies in Iran and other countries regarding homicide different risk factors like gender, age, psychiatric disorders, climatic and ecological conditions, socio-economical level, cultural factors, poverty, individual violence and family breakdown are among the most mentioned risk factors of Homicide or murder ^8-12 . Moreover, among Europeans, annual alcohol consumption positively associated with the rate of mortality due to homicide ¹³ . In Iran a study in Isfahan showed the living in rural area, living in crowding family, lack or low level of literacy and low-level income, history of violence and drugs in family were predisposing risk factors for homicide ¹⁴ .

According to the past studies about homicide, the incidence rate of homicide has varied over the years and has been influenced by many economic, social and other factors ^15-19 . In Iran all of previous studies about homicide were cross-sectional that conducted during one year, the trend of incidence rate changes of homicide in the country has not been investigated so far ^14,20 .

According to the importance of homicide in the world especially in Iran, we were going to depict the epidemiology of homicide in Iran and investigate any changes in the homicide’s incidence rate trend by applying joint point regression during 2006-2016.

Methods

The study was a cross-sectional panel (pseudo-panel) case. The homicide data were extracted from Iranian legal medicine organization from 2006 to 2016. According to the national laws in Iran on legal cases of death, all suspicious deaths should be referred to as the Legal Medicine Organization and death certificate is just issued by this organization. We extracted information of all available and registered homicide cases from Iranian legal medicine organization during the years 2006 to 2016 and the annual mortality rates were calculated. Moreover, to adjust for any differences in age distribution during the study period and also for the sake of comparison with other countries, directly standardized rates using the latest standard population of WHO were also measured ²¹ .

First, the homicide data were checked for completeness and any inconsistencies. The entered data were cleaned and edited before subsequent analysis. Second, homicide rates during 2006 to 2016 were estimated by dividing the number of homicides by population.

We obtained census population counts for Iran for 2006, 2012 and 2017 and carried out a linear interpolation of the census population counts to obtain population denominators for the intervening years ²² . All rates were calculated per 100,000 populations. We subsequently, directly standardized overall rates by age to the world (WHO 2000-2015) standard population to enable comparability between years relevant to this analysis and also rates from other countries.

Third, we applied the Joinpoint regression model for determining changes in the trend of the homicide incidence rate during the study period. The number of change points were selected based on the permutation test with 5000 iterations. Trends of homicide incidence were summarized by annual percent change (APC) and average annual percent change (AAPC). Joinpoint (V.4.6) software was used for trend analysis at 0.05 level of significance. Jointpoint regression model was fitted to the data using maximum likelihood estimation and for choosing the best model (comparing linear and log-linear models) Bayesian Information Criterion (BIC) were used. Summary statistics such as frequency and percentage were computed for qualitative data and also mean and standard deviation were computed for quantitative data. Annual percent change in a log-transformed model measures the annual percentage change in rate. It is estimated according to the following formula:

Log (Rate[y]) = b0 + b1 *y where log(rate[y]) is the natural log of rate in year [y] and the APC from year [y] to year [y+1] is equal to: (e ^b1 -1)*100.

This study was registered at Legal Medicine Research Center in Iran and was approved by the Ethics Committee of the Legal Medicine Organization (IR.LMO.REC.IR.LMO. REC.1396.22).

Results

Totally, 26918 homicide cases (81.1% in males and 18.9% in females) occurred during 2006 to 2016. The average age of the homicide cases during the study period was 32.88 ±(0.09) and most of the victims were 15-29 and 30-49 year. During the study period the relative frequency of homicide in men was more than women.

The highest and lowest incidence rate of homicide was in 2008 and 2016 respectively (). After analyzing homicide mortality rate by age groups, results showed that in 2016 highest and lowest occurrence were in 15-29 yr (4.37 per 100,000) and 0-4 yr (0.66 per 100,000) age-group, respectively. Age-group of 15-29 and 30-49 yr had the highest homicide mortality rates compared to other age-group (Table 1).

Close

Figure 1. Incidence rate of Homicide in Iran from 2006 to 2016 (per 100,000)

During 11 years, most of the homicides occurred using firearms and after that cold weapon like knife and sharp things had the second rank. The frequency of firearms was 45.2 % and knife and sharp things was 32.7% from 2006 to 2011 and from 2012 to 2016 frequency of homicide by firearms was 45.28% and knife and sharp things drop to 27.7%. The method with the lowest frequency was electrocution (0.03%).

After calculating the incidence rates by provinces of Iran in 2016, was observed that the four provinces of Sistan & Baluchistan, Khuzestan, Kerman and Ilam had the highest incidence rates in 2016, respectively (Table 2).

After comparing the age-standardized incidence rate by gender, was observed that generally, homicide rates were significantly higher among men compared to women (4.17, 95% CI: 4.04, 4.3, incidence rate- ratio of men to female).

During the study period, joint regression analysis showed the incidence rate of homicide in Iran and men have been significantly decreased (APC: -2.8% (95% CI: -3.9, -1.7) and -3.2% (95% CI: - 4.5, -1.8) respectively (P <0.001). However, the incidence rate of homicide in women was almost constant and the trend in this subgroup was not statistically significant (P =0.100). In other words, the incidence rate of homicide 2.8% and 3.2% significantly decreased per year in the whole country and men, respectively but incidence rate of homicide in women was almost constant. (Table 3).

We also estimated the trend of homicide in the four provinces with highest homicide incidence rates including Sistan and Baluchistan, Ilam, Kerman and Khuzestan. After analyzing the four mentioned provinces, a model with one joint point was chosen as the best model for Kerman province but for other provinces models without any break-point were chosen as the best model. In Kerman, the study period was divided into two periods from 2006 to 2009 and 2009 to 2016. The first slope from 2006 to 2009 revealed that incidence rate, about 16% decreased (P =0.030). In other four provinces, the slopes of trends were not statistically significant and incidence rates were constant during the study period ().

Close

Figure 2. Trend of Homicide logarithmic changes in four provinces of Iran with high incidence rate from 2006 to 2016

According to analysis of homicide mortality trends by age strata (Table 3), zero joint point model were found to be the best fitted model in the age groups 15-29 and ≥70 yr, whereas changes in two age strata 0-4 and 5-14 yr were not statistically significant. In 30-49 and 50-69 yr, 1 joint point models were obtained as the best models. For age group, 30-49 yr, homicide mortality rates presented a significant decrease of about 3% per year from 2006 to 2014 and non-significant decline from 2014 onwards. Estimation by Joint point analysis in 50-69 yr determined as a significant decline of 0.5% per year from 2006 to 2012 but a non-significant decrease from 2012 onwards.

Discussion

The current study showed trend of homicide incidence rate in Iran during 2006 to 2016. Mean age of homicide cases during 2006 to 2016 was 32.88 ±(0.09) yr and most of victims were 15-29 and 30-49 yr old. Most of the homicide cases were occurred in 15-44 yr’ age group in East-Mediterranean ²³ . In addition, in Isfahan, most of victims were 15-29 yr old ²⁴ . In high-income countries most of homicide cases occurred in the age group of 25- 34 yr old and after that in 75 yr and more ²⁵ . In the most of countries in the world, most of homicide occurs among youth and middle-aged. The reasons for this are complex but economic and social pressures, drug and alcohol abuse, media violence access, weapon availability especially cold weapon, puberty age among teens and modern cultures which values material goods and personal success and leads to unhealthy competitions among youth and teens maybe affect occurrence of homicide among mentioned age group.

In this study, similar to other studies conducted in parts of Europe and Asia, homicide occurred much more among males (66% to 87%) ^26-30 . According to the WHO reports in 2016, about 83% of homicide occurred among males which their results are in agreement with the current study ⁴ . The most frequent method of homicide was using firearms and this results agreed with another study that was conducted in United-States ³¹ while another studies in Iran reported that the most common homicide’s method was cold weapon like knife ^24,9 . The reason for this difference is that current study has been done on the national level but previous studies have been conducted in the provincial levels.

According to the results of break-point analysis, generally, incidence rate of homicide in Iran and among men were without break-point and have decreased. Moreover, incidence rate of homicide in country has been significantly decreasing from 3.67 per 100,000 to 2.81 per 100,000 during 2006 to 2016 while in Russia the incidence rate of homicide in this country from 2001 to 2009 was fairly constant ³² . In the United-States, homicide incidence rate among males and age group of 18-49 yr was more than others during 1980 to 2008. Generally, in that country trend of homicide incidence rate has been along with the ups and downs during 1980 to 2008. Therefore, in 1980 it has been increasing, and then peaked in 1991, after which the trend has been decreasing in 2008 ¹⁵ . We can say that homicide trend changes in different countries depend on national factors such as serious socio-economic and political changes, unemployment, alcohol abuse especially in European countries, and tribal traditions, especially in Asian and African countries. Changes in these factors during the time induce some changes in occurrence of crimes such as homicide. About Iran, we can say existence of laws relevant to preventing different types of violence such as children maltreatment laws against child marriage or legal age of women or men’s marriage, youth violence laws against weapons on school permission or gang or criminal group membership, sexual violence laws against rape and victim’s laws about providing for victim compensation can affect the reduce of homicide in the country.

Results of annual percent changes (APC) analysis revealed that homicide incidence rate decreased about 2.8% and 3.2% in the whole of Iran and men, respectively from 2006 to 2016 while the result of Lotufo and Bensenor’s study in Brazil during 12 yr showed that among men 15-44 yr, APC increased 4.7% during 1996 to 2001 and decreased 14.6 % during 2001 to 2007 ³³ .

Furthermore, APC analysis by age groups in current study showed that homicide mortality rates among people in 30-49 yr in Iran have decreased about 3% per year from 2006 to 2014 and in 50-69 yr this rate have declined 0.5% per year from 2006 to 2012. This finding was in contrast to another study’s results in Puerto Rico. The result of APC analysis showed significant increase from 2007 onwards especially in the 20-24 and 25-29 yr’ age groups. Moreover, AAPC was showed significant increase from 2007 onward across 20 to 44 yr’ age group ³⁴ . According to the results of current study, incidence rate of homicide in Iran have been different by provinces so that Sistan and Baluchistan, Khuzestan, Kerman and Ilam provinces had the highest incidence rate in 2016, respectively. This finding is in line with other studies in which, homicide incidence rate in other countries was different at the provincial level. Besides, incidence rate of homicide in western regions of Norway was lower than the overall average in the country ^31,35,36 .

The trend of homicide has been decreasing throughout the country but the increasing trend in some of border provinces was observed which seems to be the main challenge in controlling this phenomenon in these provinces which are in the neighborhood with countries who are at war or are involved with illegal drug and human trading. Another cause of high incidence in the four mentioned provinces of Iran can be tribal conflict, honor killing and high prevalence of consuming drugs and easy access to weapons in the border regions of the country ^37,38 .

Future studies are suggested to be carried out on the following cases:

1. Assessment of incidence rate by provinces to recognize hotspot regions and also spatial analysis

2. Assessment of risk factors related to occurrence of murder in the form of analytical studies for intervention and policy making

Use of the most complete, accurate data of homicide in the country without sampling process is the strength of the study and lack of access to other risk factors of homicide occurrence such drug and alcohol abuse, income status, employment and unemployment status are the weaknesses of the study.

Conclusion

The incidence rate of homicide in Iran has been decreasing but currently, the incidence of homicide in some border provinces is high compared to the overall average in the country. Although, the decline in incidence rate of homicide over the past decade in the country has been good but not enough. Authorities should apply different policies for controlling this issue, like recognition of homicide’s hotspot regions and controlling the risk factors that lead to the occurrence of murder such as attention to social and economic issues, violence status, drug addiction and easy access to it, psychiatric disorders, poverty, unemployment, social inequality and false tribal traditions especially in provinces with higher than average rate of homicide and planning for strategies to prevent youth violence such as, counseling, vocational training, family therapy, training health care workers to identify and refer youths at high risk for violence and activities and polices to mitigate the effect of rapid social change and tackle gun violence among youth.

Acknowledgements

The authors would like to thank the Iranian Legal Medicine Research Center and Iranian Legal Medicine Organization which participated in this project and supported this study and provided the necessary data.

Conflict of interest

None declared.

Funding

This study is financially supported by the Research Center of Legal Medicine Organization of Iran.

Highlights

References

1. Homicide. WHO web site; 2015. [updated 2015; cited 25 Oct 2016]; Available from: https://apps.who.int/violence-info/homicide.
2. Bryant W, Cussen T. Homicide in Australia: 2010–11 to 2011–12: National Homicide Monitoring Program Report. Canberra: Australian Institute of Criminology 2015(23):1-64.
3. Violence and Injury. WHO web site; 2016. [updated May 2016; cited 29 Jun 2016]; Available from: http://www.who.int/violence_injury_prevention/violence/en/.
4. Youth violence. WHO web site; 2016. [update 30 Sep 2016; cited 9 Oct 2017]; Available from: http://www.who.int/mediacentre/factsheets/fs356/en/.
5. Homicide Estimates by country. WHO web site; 2016. [update 17 Apr 2016; cited 25 Oct 2016]; Available from: http://apps.who.int/gho/data/node.main.VIOLENCEHOMICIDE?lang=en.
6. Homicide in Iran. WHO web site; 2015. [updated 2015; cited 13 Apr 2020]; Available from: https://apps.who.int/violence-info/country/IR/.
7. Barzegar A, Ghadipasha M, Aram S, Khademi A. Suicides, Homicides, and Fatal Occupational Accidents in Iran: A Cross-Sectional, Retrospective Study Based on Autopsy Reports; Need to Develop Preventive Strategies. Iran Red Crescent Med J 2019; 21(4):e65672.
8. Murray J, Cerqueira D, Kahn T. Crime and violence in Brazil: Systematic review of time trends, prevalence rates and risk factors. Aggress Violent Behav 2013; 18(5):471-83.
9. Gharehdaghi J, Ghorbani M, Akhlaghi M, Yousefinejad V, Paeezi M. Epidemiology of Homicide by Sharp Force in Tehran, Iran between 2010 and 2011. Indian J Med Forensic Med Toxicol 2013; 7(2):51-6.
10. Males Males, M M. Age, poverty, homicide, and gun homicide: Is young age or poverty level the key issue?. SAGE Open 2015:1-12.
11. Bridges FS. Gun control law (Bill C-17), suicide, and homicide in Canada. Psychol Rep 2004; 94:819-26.
12. Zadali S, Javidi F. Some Socio-Cultural Factors Influencing the Propensity to Murder among Young People in Ahwaz, Iran. International Journal of Economics, Management and Social Sciences 2014; 3(1):54-9.
13. Rossow Rossow, I I. Alcohol and homicide: a cross‐cultural comparison of the relationship in 14 European countries. Addiction 2001; 96:77-92.
14. Maghsoodloo S, Ghodousi A, Kafashian A, Bahramzadeh S, Attari A. Committing Murder in Female Murderers: A Qualitative Research. Iranian Journal of Psychiatry and Clinical Psychology 2010; 15(4):388-97.
15. Cooper A, Smith EL. Homicide trends in the United States, 1980-2008. Bur Justice Stat Spec Rep 2011:1-37.
16. Jazi ME, Hajidehabadi MA. Evaluation of the Role of Individual and Social Factors Effective on Homicide. International Journal of Humanities and Cultural Studies 2016:498-509.
17. Manso BP. Homicide in Sa˜o Paulo: An Examination of Trends from 1960–2010. Springer International Publishing. 2005.
18. McPhedran S, Baker J. Quantifying homicide trends in Australia: a methodological caution. J Inj Violence Res 2012; 4(1):44.
19. Parker KF, Mancik A, Stansfield R. American crime drops: Investigating the breaks, dips and drops in temporal homicide. Soc Sci Res 2017; 64:154-70.
20. Maes M, Cosyns P, Mektzer HY. Seasonality in violent suicide but not in nonviolent suicide or homicide. Am J Psychiatry 1993; 150(9):1380-5.
21. World standard population (2000-2025). WHO web site; 2001. [updated 2001; cited 22 Mar 2017]; Available at: https://apps.who.int/healthinfo/statistics/mortality/whodpms/definitions/pop.htm.
22. Population and housing census. Statistical Center of Iran; 2017. [updated 2 Sep 2017; cited 22 Dec 2017]; Available at: https://www.amar.org.ir/english/Population-and-Housing-Censuses.
23. Rezaeian M. Suicide/homicide ratios in countries of the Eastern Mediterranean Region. East Mediterr Health J 2008; 14(6):1459-65.
24. Karimi J, Koehler SA, Karimi R, Soleymanpour A, Mohammad K, Holakouie-Naieni K. A forensic epidemiological study of homicide in Isfahan Province and review of the literature. Am J Forensic Med Pathol 2017; 38(3):233-40.
25. Pampel FC, Williamson JB. Age patterns of suicide and homicide mortality rates in high-income nations. Social Forces 2001; 80(1):251-82.
26. Smit PR, Bijleveld CC J H, van der Zee S. Homicide in the Netherlands An exploratory study of the 1998 cases. Homicide Stud 2001; 5:293-310.
27. Mouzos J. Homicide in the course of other crime in Australia. Trends & issues in crime and criminal justice 2003; (252):1-6. Available at: https://aic.gov.au/publications/tandi/tandi252.
28. Weaver GS, Wittekind JEC, Huff-Corzine L, Corzine J, Petee TA, Jarvis JP. Violent Encounters: A Criminal Event Analysis of Lethal and Nonlethal Outcomes. J Contemp Crim Justice 2004; 20(4):348-68.
29. Azmak D, Altun G, Bilgi S, Yilmaz A. Firearm fatalities in Edirne, 1984-1997. Forensic Sci Int 1998; 95(3):231-9.
30. Henderson JP, Morgan SE, Patel F, Tiplady ME. Patterns of non-firearm homicide. J Clin Forensic Med 2005; 12(3):128-32.
31. Grinshteyn E, Hemenway D. Violent death rates: the US compared with other high-income OECD countries, 2010. Am J Med 2016; 129(3):266-73.
32. Lysova A, Shchitov N. What is Russia’s real homicide rate? Statistical reconstruction and the ‘decivilizing process’. Theor Criminol 2015; 19(2):257-77.
33. Lotufo PA, Bensenor IM. Income inequality and male homicide rates: Sa˜o Paulo, Brazil, 1996–255 2007. Eur J Public Health 2009; 19(6):602-4.
34. Zavala-Zegarra DE, López-Charneco M, Garcia-Rivera EJ, Concha-Eastman A, Rodriguez JF, Conte-Miller M. Geographic distribution of risk of death due to homicide in Puerto Rico, 2001–2010. Rev Panam Salud Publica 2012; 32(5):321-9.
35. Shaw M, Tunstall H, Dorling D. Increasing inequalities in risk of murder in Britain: trends in the demographic and spatial distribution of murder, 1981–2000. Health place 2005; 11(1):45-54.
36. Paul J Brantingham, Patricia L Brantingham. Notes on the geometry of crime. London: Routledge; 2017.
37. Farahmand M, Najafi K. A Study of Factors Affecting Inter-Ethnic Conflicts of Laks and Lors in Khorramabad City. Strategic Research on Social Problems in Iran 2018; 7(10):43-64.
38. Zare Shah Abadi A, Bondori M. A Survey of the Role of Exiles on the Sense of Security of Citizens in Izeh City. Strategic Research on Social Problems in Iran 2019; 7(4):19-40.