Introduction

This report examines suicide, drug-related and alcohol-specific deaths in Northern Ireland, building on earlier research that analysed the socio-demographic factors associated with drug-related deaths (PDF, 455 KB) and alcohol-specific deaths (PDF, 449 KB) up to 2016 and 2017, respectively.

Background

Suicide, drug-related deaths and alcohol-specific deaths remain major public health challenges in Northern Ireland, with profound and lasting impacts on individuals, families and communities. Over the past decade, mortality rates from these causes have risen (see references 1, 2, 3), with males consistently more affected across all categories. However, the age profile varied: drug-related deaths were more common among adults under 45 years; alcohol-specific deaths most affected those aged 35-64 years; and suicide affected people across the life course, from adolescents to older adults.

This study used a linked research dataset, the Northern Ireland Mortality Study (NIMS) (see reference 4), covering the period 2011 to 2022, to examine the personal, household and area factors associated with these deaths. The aim is to provide robust evidence to inform prevention strategies and public health policy.

Definitions

Suicides include self-inflicted injuries and deaths of undetermined intent.
Drug-related deaths include poisonings, drug abuse or dependence.
Alcohol-specific deaths result from health conditions that are a direct consequence of problematic alcohol usage, e.g. alcoholic liver disease.
Deaths due to drugs or alcohol, where these have been determined by the Coroner to be non-accidental are also classified as suicides. These definitions are consistent with Accredited Official Statistics.

Due to data availability, Section 1 and Annexes 4 and 5 present data based on the year of registration of death. Sections 2 and 3 present data based on year of death occurrence, which is more timely but subject to revision. Median registration times (January 2021 - December 2023, sourced from Vital Statistics NISRA) were 222 days for suicides, 237 days for drug-related deaths and 6 days for alcohol-specific deaths.

Full definitions are provided in Annex 1.


Note: A review of suicide deaths (see reference 5) in Northern Ireland (covering the period from 2015-2020) led to the reclassification of some suicide deaths initially recorded as undetermined or accidental causes, resulting in a reduction in the number of deaths previously categorised as suicides. Since suicide deaths prior to 2015 were not reviewed, suicide trend data before 2015 are not directly comparable, and are therefore not included in this report.

Report aims

This report addresses three main objectives:

  1. Summarise trends in suicide, drug-related and alcohol-specific mortality (Section 1);
  2. Compare personal, household and area characteristics of those who died with the general population (Section 2); and
  3. Identify factors associated with increased mortality risk (Section 3).

Key findings

Section 2: Socio-demographic characteristics

  • Suicides and drug-related deaths were more common among those aged 16-44 years, while alcohol-specific deaths were more prevalent among individuals aged 35-54 years (Table 2).

  • Compared to the general population, all three mortality causes were more prevalent among males, the economically inactive, the unemployed, those with no qualifications, individuals with self-reported poor mental health or disability, social renters, residents of Greater Belfast, or the 20% most deprived areas (Tables 2&3).

  • Additionally, compared to the general population, higher levels of all three causes were observed among individuals living alone (Table 3), while male children from lone-parent households had higher levels of suicide and drug-related deaths (Table 5).

See (Annex 1) for detailed definitions of socio-demographic and household variables.

Section 3: Modelling mortality risk

The most at-risk groups (male versus female comparisons), after accounting for other factors (Figures 3A-3C) & Tables 8A-8C:

Education and employment

  • No qualifications: associated with a raised risk of suicide and drug-related deaths in males.
  • Unemployment: strongly associated across all causes in males and with suicides and alcohol-specific deaths in females.
  • Economic inactivity: increased risk of alcohol-specific mortality in males and across all three causes in females.

Health conditions

  • Poor mental health (with or without other conditions): strongly associated with suicide and drug-related deaths for males and females.
  • Having a disability: associated with increased mortality risk for suicide and alcohol-specific deaths for males and across all three causes for females.

Living situation and relationships

  • Living alone: a key factor associated with increased mortality across all causes in males and with suicides in females.
  • Being divorced/separated: associated with higher mortality risk from drug-related and alcohol-specific deaths in males and being single or divorced/separated was associated with increased risk from drug-related and alcohol-specific deaths in females.
  • Male children in lone-parent households: at increased risk of suicide and alcohol-specific mortality.
  • Male lone parents in a household: increased risk of drug-related mortality.
  • Urban living: associated with increased mortality risk for drug-related deaths in males and across all causes for females.

Where to go for help

If you are struggling to cope, please call one of the organisations below. There is help available around the clock, every single day of the year, providing a safe place for anyone struggling to cope, whoever they are, however they feel.

Minding Your Head - Find out more about mental health and the issues that can affect it; early warning signs that a mental health issue may be developing; tips on how to maintain good mental health.

Lifeline - A free 24 hour crisis response helpline for people who are experiencing distress or despair, where trained counsellors will listen and help immediately on the phone and follow up with other support if necessary. Phone 0808 808 8000

Samaritans – A registered charity aimed at providing emotional support to anyone in emotional distress, struggling to cope, or at risk of suicide throughout UK and Ireland, often through their telephone helpline or online chat. Freephone 116 123

Information for the media - There is strong evidence that sensationalist media reports about suicide and the nature of suicide deaths can lead to subsequent additional suicidal behaviours (suicides and suicide attempts) or indeed increase the likelihood of copycat deaths. Media professionals should exercise caution in reporting on suicide, balancing the public’s “right to know” against the risk of causing harm. It is therefore important that those reporting on suicide statistics adhere to the guidelines of safe reporting from the World Health Organisation (WHO), the International Association for Suicide Prevention (IASP) and the Samaritans.

Section 1: Mortality trends

This section presents a brief overview of the leading causes of deaths in Northern Ireland by sex from 2015-2023 and examines recent trends in suicides, drug-related deaths and alcohol-specific deaths.

Cancer and circulatory conditions were the leading causes of death among individuals aged 15-64 years in Northern Ireland. For both males and females, alcohol-specific deaths ranked third. For males suicide ranked fourth and drug-related deaths ranked fifth; for females, drug-related deaths ranked seventh and suicide ranked eighth Annex 3, Figure 4A.

Among those aged 15-49 years Annex 3, Figure 4B, suicide was the leading cause of death for males (19.2%), followed by drug-related deaths (18.3%), both of which exceeded cancer (14.5%). For females in this age group, cancer remained the most common cause (32.4%), followed by drug-related deaths (11.6%) and suicide (11.0%). Alcohol-specific deaths were also a leading cause, ranking fifth for both males (10.1%) and females (10.5%).

Age-standardised mortality rates (ASMRs) from 2013-2023 for suicide, drug-related and alcohol-specific deaths are presented in Figure 1, for individuals aged 15-49 years and 15-64 years.
Annual registration-based figures can fluctuate, particularly in smaller populations like Northern Ireland, due in part to procedural delays. To provide a clearer picture of underlying trends, three-year rolling average ASMRs are presented in Figure 1. Further details on how ASMRs are calculated are available in Annex 2.

Figure 1: ASMRs (per 100,000 population) for suicide, drug-related and alcohol-specific deaths, 2011-2013 to 2021-2023 (three-year rolling averages)

line plots showing age-specific mortality rates for suicide, drugs and alcohol for 15-49 years and 15-64 years in Northern Ireland from 2013 to 2023.

Note: Three-year suicide trends are reported from 2017 onwards. This reflects the impact of a review of suicide deaths (2015-2020), which led to the reduction in the number of deaths previously recorded as suicides. As deaths prior to 2015 were not reassessed, any three-year trend data incorporating years before 2015 are not directly comparable.

Definition

Age-standardised mortality rates (ASMRs) Since mortality rates generally increase with age, populations with older age structures tend to experience higher overall death rates. To enable meaningful comparisons over time and between countries, ASMRs are used as they adjust for differences in age distribution.


Key trends

  • Among those aged 15-49, three-year average suicide ASMRs peaked in 2020 (reflecting the 2018-2020 average). For the 15-64 age group, suicide ASMRs remained relatively stable between 2017 and 2023.

  • Drug-related ASMRs showed a sustained upward trend across both age groups, peaking in 2021 (2019-2021 average).

  • Alcohol-specific ASMRs also followed a rising trend from 2013 across both age groups, reaching a peak in 2021 (2019-2021 average), before declining slightly in the most recent years.

Annex 4, Figures 5A & 5B present mortality trends for suicide, drug-related deaths, and alcohol-specific deaths across UK countries based on the registration year of death from 2015-2023. Among UK countries, England has the lowest suicide rate. Scotland has the highest rate of drug-related mortality in the UK, while Northern Ireland, England and Wales have shown relatively similar rates. Alcohol-specific mortality rates are considerably higher in Northern Ireland and Scotland, which up to 2022, exceeded those in Wales and England. Regional disparities within countries are also evident. For instance, mortality rates are higher in the north of England compared to the south (see reference 6).

Note: Drug-related deaths in Northern Ireland were compared to drug poisonings in England, Wales and Scotland.

Section 2: Socio-demographic characteristics

This section provides a descriptive overview of the personal, household and area characteristics of individuals who died due to suicide, drug-related and alcohol-specific deaths in Northern Ireland. Comparisons are made against the general population aged 16 to 64 years using data from the Northern Ireland Mortality Study (NIMS).

(i) Data source

The Northern Ireland Mortality Study (see reference 4), is a population-based research dataset that links 2011 Census records to registered deaths from April 2011 to September 2022. This linkage allows robust analysis of individual, household and area characteristics associated with mortality.

Figure 2: NIMS study population, ages 16-64 years

Figure 2 shows the Northern Ireland Mortality study population over 16 years. It starts with baseline population at March 2011 and highlight the number of subsequent suicide, drug and alcohol deaths registered in Northern Ireland until September 2022
Note: Example of calculating coverage rates for deaths on NIMS. Coverage rates are shown in brackets in Figure 2. Between April 2011 and September 2022, a total of 2,564 suicides were registered in Northern Ireland. Of these, 1,649 suicides were included on the NIMS dataset, resulting in a coverage rate of 64.3% (1,649/2,564).


Key strengths of NIMS include:

  1. Nationally representative data enabling comparisons of characteristics between those that died and the general population.
  2. Sufficient sample sizes for detailed subgroup analyses.
  3. Supports modelling to quantify mortality risk across diverse groups.

(ii) Characteristics of those who died compared with the general population

Annex 5, Table 2 outlines personal characteristics while Annex 5, Table 3 details household and area characteristics of individuals who died by suicide, drug-related and alcohol-specific causes, in each case compared to the general population. Data disaggregated by sex are available in Annex 6, Table 4 and Annex 6, Table 5.

Males aged 16-64 years made up 48.9% of Northern Ireland’s population but accounted for 75.9% of suicides, 70.8% of drug-related deaths and 66.5% of alcohol-specific deaths.

Unless notable differences between males and females were observed, overall percentages are reported below.

Age differences

  • Suicides (74.0%) and drug-related deaths (85.0%) were disproportionately higher among 16-44 year olds (61.1% of the general population).
  • Alcohol-specific deaths were over-represented among 35-54 year olds (73.9% vs 43.5% of the population).

Marital status

  • Single males (which made up 44.1% of the overall male population) were over-represented in suicides (61.6% of all male suicides occurred among single males) and drug-related deaths (78.9% of all male drug-related deaths occurred among single males).
  • Divorced or separated individuals had a higher proportion of alcohol-specific deaths (31.0%) compared to the population (10.2%).

Qualifications and employment status

  • Those without formal qualifications (21.4% in the population) were over-represented in suicide (33.0%), drug-related (40.3%), and alcohol-specific (41.4%) deaths.
  • Economically inactive individuals (27.8% in the population) had higher rates of suicide (39.8%), drug-related (53.1%), and alcohol-specific (50.5%) deaths.

Health conditions

  • Among individuals with poor mental health (7.7% in the population), deaths by suicide (24.4%), drug-related (32.7%), and alcohol-specific (29.2%) were notably over-represented.
  • Similarly, among those with a disability (17.2% in the population), suicides (31.4%), drug-related (41.9%), and alcohol-specific (50.5%) deaths were disproportionately high.

Housing and living arrangements

  • Social housing residents (12.8% in the population) accounted for a higher share of deaths by suicide (25.5%), drugs (39.3%) and alcohol (35.5%).
  • Lone-parent females in a household (15.0% in the population) had elevated levels of suicide (27.8%), drug-related (34.4%) and alcohol-specific (21.8%) deaths.
  • Male children from lone parent households (8.7% in the population) had higher proportions of suicide (16.4%) and drug-related deaths (20.8%).
  • Individuals living alone (10.6% in the population) faced increased rates of suicide (22.5%), drug-related (27.4%) and alcohol-specific (40.2%) deaths.

Area factors

  • Greater Belfast (34.0% in the population), recorded a higher proportion of suicides (40.2%), drug-related (45.8%), and alcohol-specific (39.6%) deaths.
  • The 20% most deprived areas accounted for 27.8% of suicides, 31.4% of drug-related deaths and 27.6% of alcohol-specific deaths.

Section 3: Modelling mortality risk

While Section 2 provided descriptive statistics, this section explores associations between personal, household and area characteristics and cause-specific mortality in greater depth using modelling analysis. Results from the overall models (Annex 8, Tables 8A-8C) for suicide, drug-related and alcohol-specific deaths indicated higher mortality risks for males compared to females - over three times higher for suicide, more than twice as high for drug-related deaths and 78% higher for alcohol-specific deaths.

To account for differences between males and females, separate models were developed for each sex. These models account for differences in characteristics (e.g. having educational qualifications or not) and enable the simultaneous evaluation of multiple factors. Further detail on the methods can be found in Annex 2.

Figures 3A, 3B and 3C illustrate sex-specific Hazard Ratios for suicides, drug-related deaths and alcohol-specific deaths respectively, highlighting the characteristics most strongly associated with increased mortality risk. Each characteristic is compared with its specific reference group (e.g. married, employed, no disability). Full modelling details can be found in (Annex 8, Tables 8A-8C).

The variables included in the models were selected based on well-established associations from previous research (see references 7, 8, 9). While these findings identify factors linked to higher mortality risk, they should not be interpreted as evidence of causation but rather as an insight into broader patterns associated with each cause of death.

Suicides

Figure 3A: Mortality hazard ratios (with 95% confidence intervals) for suicides, 16-64 years by sex: April 2011 to September 2022

Figure 3A shows which groups had higher risk of suicide for males and females aged 16–64 years in Northern Ireland (April 2011–September 2022).

Note and interpretation
Hazard ratios (HRs) show how much more likely death is for a given group compared to its reference group, after adjusting for other factors. For example, males with poor mental health alone (HR 2.93) were 2.93 times more likely to die by suicide than males without poor mental health while economically inactive females (HR 1.46) had a 46% increased risk of suicide compared to employed females.

These associations do not imply causation but highlight key patterns for targeting prevention efforts.


After accounting for other factors (such as age and area deprivation) (see Table 8A), higher suicide risk was associated with the following factors:

  • Reporting poor mental health alone was linked to a significantly increased risk (2.93 times higher for males; 4.14 times higher for females). When poor mental health was reported alongside other conditions, risk remained elevated (2.11 times higher for males; 3.33 times higher for females).
  • Living alone (compared to adults living in a couple household) was associated with nearly double the risk (1.93 for males; 1.99 for females).
  • Having no formal qualifications (vs degree-level) was associated with a 90% increased risk for males and 29% for females (difference for the latter not statistically significant). Intermediate-level qualifications, as the highest level of educational attainment, was also associated with a 53% increase in risk among males.
  • Having a disability was associated with higher risk (36% for males: 25% higher risk for females - not statistically significant).
  • Unemployment (vs employment) was associated with a 30% increased risk for males and 89% for females.

Additional associations:

  • Among males, higher risk was associated with being a child in a lone parent household (47% increase).
  • Among females, elevated risk was associated with economic inactivity (46%) and urban residence (28%).

Alcohol-specific deaths

Figure 3C: Mortality hazard ratios (with 95% confidence intervals) for alcohol-specific deaths, 16-64 years by sex: April 2011 to September 2022

Figure 3B shows which groups had higher risk of alcohol-specific death for males and females aged 16–64 years in Northern Ireland (April 2011–September 2022).

Note and interpretation
Hazard ratios (HRs) show how much more likely death is for a given group compared to its reference group, after adjusting for other factors. For example, unemployed males (HR: 2.15) were 2.15 times more likely to die than employed males while economically inactive females (HR: 1.63) had a 63% increased risk of alcohol-specific death compared to employed females.

These associations do not imply causation but highlight key patterns for targeting prevention efforts.


After accounting for other factors (such as age and area deprivation) (see Table 8C), higher risk of alcohol-specific death was associated with:

  • Unemployment (vs employment) more than doubled the risk for males (HR: 2.15) and females (HR: 2.51).
  • Economic inactivity (vs employment) was also associated with elevated risk - 89% higher for males and 63% higher for females.
  • Living in social-rented accommodation (vs owner-occupied) was associated with increased risk for both males (80%) and females (51%).
  • Being divorced or separated (vs married) was associated with a 75% increase in risk for males and more than doubled the risk for females (HR: 2.39).
  • Having a disability increased the risk by 63% for males and by more than double for females (HR: 2.12).

Additional associations:
Among males (all compared to adults living in couple households):

  • Living alone was associated with more than double the risk (HR: 2.44), and
  • Being a child in a lone parent household was associated with a 70% higher risk.
  • Other living arrangements (e.g. unrelated individuals sharing accommodation) was associated with a 52% increased risk.

Among females:

  • Being single (86% increase),
  • Urban residence (42% increase), and
  • Reporting poor mental health alone (i.e. without other conditions - 40% increase).

Summary

This study provides new insights into suicide, drug-related, and alcohol-specific deaths in Northern Ireland, examining population level risks over recent years. The findings highlight that these deaths disproportionately impact males, individuals with no educational qualifications, the economically inactive, people with poor mental health or disabilities, and those living in social housing. Geographic factors also play a role, with higher mortality risks in the most deprived or urban areas. For males, living alone was associated with increased mortality risk, while for females, being single, separated or divorced was one of several important risk factors. The study also found that male children in lone-parent households had an elevated risk of suicide and alcohol death, while male lone parents in a household had an increased risk of drug-related mortality.

Mental health conditions were a major contributing factor. Northern Ireland has 25% higher levels of reported mental illness among both young people (see reference 10) and adults (see reference 11) compared to the rest of the UK, a disparity often linked to the legacy of conflict and socio-economic challenges. The study found strong associations between poor mental health, disability and all three mortality types examined in this study. Loneliness (see reference 12) and poverty (see reference 13) may further exacerbate these risks. Living alone or being single, divorced or separated were associated with increased mortality risk, although it is important to note that not everyone in these groups necessarily experiences loneliness or social isolation. Loneliness, in particular, is a concern among young people in Northern Ireland, with survey data indicating that nearly half of 16-24 year olds report feeling lonely often (see reference 14). This issue has been recognised as a key societal challenge, and loneliness is now included as an indicator in the Programme for Government (see references 15, 16) in Northern Ireland.

In response to these challenges, Northern Ireland has introduced several strategies including the ten-year Mental Health Strategy (2021-2031) (see reference 17), the ’Preventing Harm, Empowering Recovery - Substance Use Strategy (2021-2031) (see reference 18), and the cross-departmental suicide prevention plan, Protect Life 2 (2019-2027) (see reference 19). These strategies align with the broader Making Life Better Public Health strategy (2013-2023) (see reference 20, which aimed to promote mental well-being, reduce stigma, encourage responsible alcohol use, and address substance use through education and harm reduction. The Programme for Government (see reference 15) also aims to reduce health inequalities, build resilience, and mitigate the social impact of these issues. In addition, Minimum Unit Pricing (MUP) for alcohol is currently under review (see reference 21).

Limitations

While the study provides important insights, several limitations should be noted. The analysis is based on socio-economic, health and household data from the 2011 Census, which may not reflect more recent individual circumstances. Although a NIMS 2021 study is now available, NIMS 2011 allowed for a much longer follow-up period for assessing mortality outcomes. The modelling analysis also excludes individuals living in communal establishments such as hostels, hospitals and prisons. The linked Census - death registrations dataset covered 63% of suicides, 60% of drug-related deaths and 84% of alcohol-specific deaths, suggesting that the impact of socio-economic factors may be underestimated, particularly among hard-to-reach populations such as those engaging in harmful substance use. Future work should aim to include these groups for a more complete picture. An upcoming study (see reference 22) of mortality among ex-prisoners in Northern Ireland will offer further insights into one such population. The inclusion of household income as a poverty measure could also enhance the analysis, given the well-established link (see reference 23) between poverty and suicide, drugs and alcohol deaths. Despite these limitations, the study enhances understanding of the factors contributing to these deaths, with the inclusion of household data adding particular insight.

What next?

The research team is collaborating with policy stakeholders, including the Department of Health and Public Health Agency to maximise the policy and operational relevance of this work.

Further information

Acknowledgements

The help provided by the staff of the Northern Ireland Mortality Study (NIMS) and the NILS Research Support Unit is acknowledged. The NIMS is funded by the Health and Social Care Research and Development Division of the Public Health Agency (HSC R&D Division) and NISRA. The NILS-RSU is funded by the Economic & Social Research Council which is part of the UK Research and Innovation and by the Northern Ireland Government. We also thank colleagues in NISRA, the Department of Health and the Public Health Agency who contributed to the development of this report.

The Office for Research Ethics Committees Northern Ireland has ratified the usage of NIMS for approved research. Access to the NIMS dataset is limited to researchers and research teams who are accredited under the Digital Economy Act (DEA) and subject to approval by each of the data providers and the Research Approvals Group. The NIMS dataset was accessed in the NISRA safe setting and records were rendered anonymous for analysis purposes.

About ADR UK

ADR UK (Administrative Data Research UK) is a partnership transforming the way researchers access the UK’s wealth of public sector data, to enable better informed policy decisions that improve people’s lives. By linking together data held by different parts of government and facilitating safe and secure access for accredited researchers to these newly joined-up and de-identified data sets, ADR UK is creating a sustainable body of knowledge about how our society and economy function – tailored to give decision makers the answers they need to solve important policy questions. ADR UK is funded by the Economic and Social Research Council (ESRC), part of UK Research and Innovation.

Contact details

Research team John Hughes, Brian Foley, Jana Ross, Carmel Colohan, and Deborah Lyness (all Administrative Research Unit, NISRA)

For further information on this research, please contact

Accessibility contact

Please contact Dissemination Branch for assistance with accessibility requirements or alternative formats. Contact details are:

Email:

Telephone: +44 (0)300 200 7836

Dissemination Branch
NISRA
Colby House
Stranmillis Court
BELFAST
BT9 5RR

Annexes

Annex 1 - Data and definitions

Due to data availability, comparative trends in Section 1 and Annexes 4 and 5 are based on the year of registration of death. In contrast, all data in Sections 2 and 3 are based on the year of death occurrence, as registration-based data can be impacted by administrative processes and procedural delays. Median registration times (January 2021 - December 2023, sourced from Vital Statistics NISRA) were 222 days for suicides, 237 days for drug-related deaths and 6 days for alcohol-specific deaths. It should be noted that occurrence-based data are therefore more timely but subject to future revision, particularly for the most recent years.

The definitions in this report align with those used in official statistics and reflect cases where the underlying cause recorded on the death certificate meets the definition for suicide, drug-related and alcohol-specific deaths. Deaths were categorised using International Classification of Disease Codes (ICD-10 codes) (see reference 24), as detailed in Table 1.

Annex 2 - Methodology

Age-standardised mortality rates reflect the mortality rates that would occur if the population of Northern Ireland had the same age distribution (the proportion of total population in each five year age group) as a hypothetical European population, known as the European Standard Population (ESP) 2013 European Standard Population (PDF, 6 MB). ASMRs are presented per 100,000 people and allow comparisons of mortality rates across different age distributions, whether over time or between countries.

Cox Proportional Hazards regression (see reference 30) is a method used in time-to-event analysis (or survival analysis) to evaluate the duration between the beginning of a study and the occurrence of an event (e.g. alcohol-specific death). This model, a form of multiple regression analysis, estimates the effect of a number of predictor variables on the time until an event, e.g. death. Unlike logistic regression, the Cox proportional hazards model is better suited for analysing the risk of death from a specific cause as it incorporates not only whether the individual dies but also the timing of the death and competing risks from other causes. Similar hazard ratios were observed when running the analyses using logistic regression models.

Variable selection: All Northern Ireland residents were initially analysed to examine the impact of each individual predictor variable, adjusted for age and sex. Weaker predictor variables (based on the significance level of p>.05) were then removed from the initial selection of models and the model was rerun to examine all remaining predictor variables simultaneously.

Hazard Ratio (HR) refers to a likelihood ratio derived from the Cox proportional hazards model. It compares the hazard (or risk) of an event occurring in different categories of a variable, relative to the designated reference category, which always has a HR of one.

  • A HR of one for the comparison group indicates no difference in the likelihood of the event occurring between the reference category and the comparison group.
  • A HR of greater than one suggests that individuals in the comparison group are more likely to experience the event (e.g. death) than those in the reference group, after accounting for other factors. For example, in the suicide model (Table 8A), a HR of 1.73 for ‘no qualifications’ indicates that this group has a 73% higher risk of suicide compared to those with a degree, after taking the other model variables into account.
  • A HR less than one indicates that the comparison group is less likely to experience the event than the reference group. For example, a HR of 0.27 (as shown in Table 8B) for individuals aged 55-64 compared to those aged 35-44 implies a 73% lower likelihood of drug-related death for individuals aged 55-64 compared to 35-44 year-olds, after taking the other model variables into account.

95% confidence intervals (cis) are a range of likely values around the hazard ratio. Cis that do not cross one are statistically significant while cis that do cross one are not statistically significant.

Annex 3 - Leading Causes of death

For all deaths registered from 2015 to 2023, Figure 4A shows the leading causes of death, for ages 15-64 years, for males and females.

Figure 4A. Leading causes of death in Northern Ireland from 2015 to 2023: by sex for those aged 15-64 years (%)

Figure 4A shows the leading causes of death among males and femlaes aged 15-64 years in Northern Ireland from 2015 to 2023. Percentages for each cause of death are shown in horizontal bars.
Note: Some deaths are categorised as both suicides and drug-related, so percentages may not sum to 100%. For further details on definitions and cause of death categories used in Figure 4A, see Annex 1.



For all deaths registered from 2015 to 2023, Figure 4B shows the leading causes of death, for ages 15-49 years, for males and females.

Figure 4B. Leading causes of death in Northern Ireland from 2015 to 2023: by sex for those aged 15-49 years (%)

Figure 4B shows the leading causes of death among males and femlaes aged 15-49 years in Northern Ireland from 2015 to 2023. Percentages for each cause of death are shown in horizontal bars.


Note: Some deaths are categorised as both suicides and drug-related, so percentages may not sum to 100%. For further details on definitions and cause of death categories used in Figure 4B, see Annex 1.

Annex 4 - UK Trends

Figure 5A: UK ASMR (per 100,000) trends 2015-2023, 15-49 years

Figure 5A shows line plots of suicide, drug-related and alcohol-specific deaths for people aged 15-49 years in England, Northern Ireland, Scotland and Wales> Rates are shown per 100,000 population.

Figure 5B: UK ASMR (per 100,000) trends 2015-2023, 15-64 years

Figure 5B shows line plots of suicide, drug-related and alcohol-specific deaths for people aged 15-64 years in England, Northern Ireland, Scotland and Wales> Rates are shown per 100,000 population.

Figures for drug-related deaths in Northern Ireland are comparable to drug poisonings in England, Wales and Scotland. It should be noted, however, that cross country comparisons will be affected by information available in coroners’ systems and what in turn is made available to create the official data. Note: the lower number of suicides reported in Wales (see reference 31) in 2021 is partly attributed to delays in the registration process caused by the Covid-19 pandemic and associated lockdowns.

Annex 5 - Characteristics

Table 2: Personal characteristics (%) of the Northern Ireland population and among those who died from suicide, drug-related or alcohol-specific deaths: 16-64 years

Characteristic
Category
General
Suicide
Drugs
Alcohol
population (%)
deaths (%)
deaths (%)
deaths (%)
N = 1,107,664
N = 1,649
N = 1,094
N = 2,375
Sex Male 48.9 75.9 70.8 66.5
Sex Female 51.1 24.1 29.2 33.5
Age 16-24 18.7 26.5 33.5 2.0
Age 25-34 20.7 24.4 28.0 9.8
Age 35-44 21.7 23.1 23.5 30.2
Age 45-54 21.8 19.8 11.6 43.7
Age 55-64 17.1 6.2 3.3 14.3
Marital status Single 41.1 59.7 72.2 38.3
Marital status Married 47.1 26.0 11.6 27.3
Marital status Divorced/separated 10.2 13.4 15.5 31.0
Marital status Widowed 1.6 0.9 0.8 3.4
Educational attainment No qualifications 21.4 33.0 40.3 41.4
Educational attainment Intermediate level 52.6 54.0 51.4 43.9
Educational attainment Degree-level 26.0 13.0 8.4 14.8
Economic activity Employed (including students) 66.8 49.4 32.2 38.6
Economic activity Economically inactive 27.8 39.8 53.1 50.5
Economic activity Unemployed 5.5 10.8 14.7 10.9
Poor mental health No 92.3 75.6 67.3 70.8
Poor mental health Yes 7.7 24.4 32.7 29.2
Having a disability No 82.8 68.6 58.1 49.5
Having a disability Yes 17.2 31.4 41.9 50.5
Note 1: Intermediate level: school-level qualification, other vocational qualification or apprenticeship.
Note 2: The proportions of deaths were weighted by age and sex to correct for any under-representation. See Annex 7 for further details.

Table 3: Household and area characteristics (%) of the Northern Ireland population and among those who died from suicide, drug-related or alcohol-specific deaths: 16-64 years

Characteristic
Category
General
Suicide
Drugs
Alcohol
population (%)
deaths (%)
deaths (%)
deaths (%)
N = 1,081,735
N = 1,611
N = 1,055
N = 2,314
Housing tenure Owner-occupied 73.1 56.6 39.4 47.0
Housing tenure Private-rental 14.1 17.9 21.3 17.6
Housing tenure Social-rental 12.8 25.5 39.3 35.5
Living arrangements Adult in a couple household 53.5 31.4 17.9 33.1
Living arrangements Child in a couple household 15.4 16.8 16.9 4.2
Living arrangements Lone parent in a household 8.8 8.3 12.6 10.4
Living arrangements Child in a lone parent household 7.2 14.9 17.0 6.3
Living arrangements Living alone 10.6 22.5 27.4 40.2
Living arrangements Other household types 4.5 6.2 8.1 5.8
Local Government District Greater Belfast 34.0 40.2 45.8 39.6
Local Government District Derry City and Strabane 8.2 7.6 7.7 10.8
Local Government District Rest of Northern Ireland 57.8 52.3 46.6 49.6
Area deprivation Top 20% of deprived areas 19.5 27.8 31.4 27.6
Area deprivation Bottom 80% of deprived areas 80.5 72.2 68.6 72.4
Area of residence Urban 63.8 73.0 83.4 78.6
Area of residence Rural 36.2 27.0 16.6 21.4
Note 1: Adult in a couple household (i.e. a member of a couple living together).
Note 2: Child living in a couple or lone parent household includes both dependent and adult children (i.e. individuals aged 18 years or over still living at home).
Note 3: Other household types e.g. unrelated individuals living together, such as young professionals.
Note 4: Greater Belfast encompasses Belfast, Antrim and Newtownabbey and Lisburn and Castlereagh Local Government Districts (see reference 27).
Note 5: The proportions of deaths were weighted by age and sex to correct for any under-representation. See Annex 7 for further details.

Annex 6 - Characteristics by sex


Table 4: Personal characteristics (%) of the Northern Ireland population and among those who died from suicide, drug-related or alcohol-specific deaths by sex

Males
Females
Characteristic
Category
General population (%)
Suicide (%)
Drugs (%)
Alcohol (%)
General population (%)
Suicide (%)
Drugs (%)
Alcohol (%)
Age 16-24 19.3 27.1 37.4 1.9 18.1 24.7 24.0 2.0
Age 25-34 20.3 24.8 30.3 10.0 21.1 23.0 22.5 9.6
Age 35-44 21.4 22.6 22.5 30.0 22.0 24.8 26.0 30.5
Age 45-54 21.7 19.3 8.4 43.4 21.8 21.3 19.3 44.5
Age 55-64 17.3 6.2 1.4 14.7 16.9 6.2 8.1 13.4
Marital status Single 44.1 61.6 78.9 43.0 38.3 53.8 55.8 29.2
Marital status Married 46.9 26.9 10.0 24.8 47.3 23.1 15.5 32.2
Marital status Divorced/separated 9.1 11.5 11.1 32.3 14.4 23.1 28.8 38.6
Educational attainment No qualifications 22.9 33.4 41.8 44.2 20.0 31.5 36.7 35.7
Educational attainment Intermediate level 54.2 54.6 51.6 43.1 51.1 52.4 50.7 45.4
Educational attainment Degree-level 23.0 12.0 6.6 12.7 28.9 16.1 12.6 18.9
Economic activity Employed (including students) 69.6 52.1 32.7 38.1 64.0 40.9 30.9 39.8
Economic activity Economically inactive 22.9 36.1 49.0 49.1 32.4 51.3 62.9 53.3
Economic activity Unemployed 7.5 11.8 18.2 12.8 3.6 7.8 6.2 6.9
Poor mental health No 93.5 79.2 70.2 71.3 91.1 64.5 60.3 69.9
Poor mental health Yes 6.5 20.8 29.8 28.7 8.9 35.5 39.7 30.1
Having a disability No 83.2 70.3 61.1 49.1 82.5 63.2 50.6 50.2
Having a disability Yes 16.8 29.7 38.9 50.9 17.5 36.8 49.4 49.8
Note 1: Intermediate level: School-level qualification, other vocational qualification or apprenticeship.
Note 2: The proportions of deaths were weighted by age and sex to correct for any under-representation. See Annex 7 for further details.


Table 5: Household and area characteristics (%) of the Northern Ireland population and among those who died from suicide, drug-related or alcohol-specific deaths by sex

Males
Females
Characteristic
Category
General population (%)
Suicide (%)
Drugs (%)
Alcohol (%)
General population (%)
Suicide (%)
Drugs (%)
Alcohol (%)
Housing tenure Owner-occupied 74.4 59.4 40.1 44.4 71.9 47.5 37.7 52.1
Housing tenure Private-rented 13.6 17.3 21.2 18.6 14.5 19.6 21.4 15.4
Housing tenure Social-rented 12.0 23.2 38.7 37.0 13.6 32.8 40.9 32.4
Living arrangements Adult in a couple household 53.3 32.1 15.2 29.5 53.7 29.0 24.4 40.3
Living arrangements Child in a couple household 18.3 19.2 20.2 5.0 12.8 9.2 8.8 2.7
Living arrangements Lone parent in a household 2.2 2.0 3.7 4.6 15.0 27.8 34.4 21.8
Living arrangements Child in a lone parent household 8.7 16.4 20.8 7.9 5.7 10.0 7.7 3.0
Living arrangements Living alone 12.3 23.9 30.3 46.4 9.1 18.4 20.4 27.8
Living arrangements Other household types 5.3 6.3 9.7 6.6 3.8 5.7 4.2 4.4
Local Government District Greater Belfast 33.4 39.3 46.3 40.4 34.5 42.8 44.4 37.9
Local Government District Derry City and Strabane 8.1 7.6 7.6 9.4 8.3 7.4 7.9 13.6
Local Government District Rest of Northern Ireland 58.5 53.1 46.1 50.1 57.2 49.7 47.7 48.5
Area deprivation Top 20% of deprived areas 19.4 27.6 30.4 28.2 19.6 28.3 33.9 26.2
Area deprivation Bottom 80% of deprived areas 80.6 72.4 69.6 71.8 80.4 71.7 66.1 73.8
Area of residence Urban 62.7 71.5 83.6 78.5 64.8 77.8 82.9 78.7
Area of residence Rural 37.3 28.5 16.4 21.5 35.2 22.2 17.1 21.3
Note 1: Adult in a couple household (i.e. a member of a couple living together).
Note 2: Child living in a couple or lone parent household includes both dependent and adult children (i.e. individuals aged 18 years or over still living at home).
Note 3: Other household types e.g. unrelated individuals living together, such as young professionals.
Note 4: Greater Belfast encompasses Belfast, Antrim and Newtownabbey and Lisburn and Castlereagh Local Government Districts (see reference 27).
Note 5: The proportions of deaths were weighted by age and sex to correct for any under-representation. See Annex 7 for further details.

Annex 7 - Weighting

The study analysed 1,107,664 individuals aged 16-64 years enumerated in the 2011 Census. In the NIMS dataset, 1,649 suicides, 1,094 drug-related deaths and 2,375 alcohol-specific deaths were registered between 1 April 2011 and 30 September 2022, representing 64.3%, 61.7% and 86.3% of these deaths respectively. Differences between official deaths and those in the research dataset were due to factors such as migration, differences in personal details (e.g. name, date of birth, address) and Census non-enumeration.

For household and area analyses in Tables 3 and 5 and modelling analyses in Section 3, individuals in communal establishments were excluded due to household-level non-response. Other exclusions were made for missing or edited data on long-term illness, urban residency or education level resulting in 38 suicides, 39 drug-related deaths and 61 alcohol-specific deaths being omitted. Weighting adjustments were applied in Tables 2-5 to correct for under-representation of deaths by sex and age group, calculated by dividing official death counts by corresponding research dataset counts (Tables 6 and 7).

Table 6: Sex and age specific weights applied to deaths: residents of households and communal establishments

Suicide
Drugs
Alcohol
Age Male Female Male Female Male Female
16-24 2.55 2.79 2.46 2.79 1.66 1.30
25-34 1.59 1.41 1.58 1.37 1.66 1.30
35-44 1.44 1.30 1.56 1.28 1.47 1.23
45-54 1.26 1.17 1.32 1.12 1.28 1.08
55-64 1.17 1.09 1.09 1.10 1.16 1.04

Table 7: Sex and age specific weights applied to deaths: residents of households only

Suicide
Drugs
Alcohol
Age Male Female Male Female Male Female
16-24 2.63 2.97 2.54 2.79 1.71 1.30
25-34 1.66 1.42 1.67 1.43 1.71 1.30
35-44 1.46 1.32 1.67 1.30 1.49 1.23
45-54 1.29 1.19 1.33 1.14 1.33 1.09
55-64 1.19 1.10 1.15 1.10 1.20 1.06

Note: In Tables 6 and 7, a combined weight for individuals aged 16-34 was generated for alcohol-specific deaths due to the small number of alcohol-specific deaths in the 16-24 age group.

Annex 8 - Full modelling results

Table 8A: Hazard ratios with 95% confidence intervals (cis) for suicides, overall and by sex

Category (i) Overall
(95% CI)		 (ii) Male
(95% CI)		 (iii) Female
(95% CI)
Male (vs female) 3.17 (2.81,3.58) NA NA
16-24 (vs 35-44) 0.92 (0.76,1.12) 0.92 (0.74,1.15) 0.94 (0.63,1.40)
25-34 (vs 35-44) 1.10 (0.94,1.28) 1.12 (0.94,1.34) 1.02 (0.75,1.37)
45-54 (vs 35-44) 0.84 (0.73,0.98) 0.85 (0.72,1.01) 0.83 (0.62,1.10)
55-64 (vs 35-44) 0.60 (0.50,0.72) 0.59 (0.48,0.73) 0.61 (0.43,0.88)
Single (vs married) 1.12 (0.90,1.40) 1.06 (0.83,1.37) 1.33 (0.86,2.05)
Divorced/separated (vs married) 1.17 (0.93,1.48) 1.09 (0.82,1.44) 1.37 (0.88,2.14)
Intermediate (vs degree) 1.40 (1.21,1.63) 1.53 (1.27,1.83) 1.09 (0.82,1.43)
No qualifications (vs degree) 1.73 (1.46,2.05) 1.90 (1.55,2.33) 1.29 (0.93,1.78)
Economically inactive (vs employed) 1.20 (1.05,1.38) 1.15 (0.98,1.36) 1.46 (1.13,1.89)
Unemployed (vs employed) 1.39 (1.16,1.66) 1.30 (1.06,1.58) 1.89 (1.25,2.86)
Poor mental health alone (vs none) 3.29 (2.77,3.90) 2.93 (2.36,3.63) 4.14 (3.09,5.54)
Poor mental health + other conditions (vs none) 2.45 (2.05,2.93) 2.11 (1.70,2.62) 3.33 (2.41,4.60)
Having a disability (vs not) 1.32 (1.14,1.53) 1.36 (1.15,1.62) 1.25 (0.94,1.65)
Child in a couple household (vs adult in a couple household) 0.91 (0.70,1.17) 0.98 (0.74,1.31) 0.61 (0.34,1.10)
Lone parent household (vs adult in a couple household) 1.42 (1.09,1.84) 1.20 (0.78,1.85) 1.35 (0.89,2.04)
Child in a lone parent household (vs adult in a couple household) 1.40 (1.09,1.79) 1.47 (1.11,1.95) 1.19 (0.69,2.07)
Living alone (vs adult in a couple household) 1.90 (1.53,2.36) 1.93 (1.51,2.48) 1.99 (1.30,3.06)
Living other (vs adult in a couple household) 1.23 (0.94,1.63) 1.19 (0.86,1.63) 1.50 (0.86,2.63)
Private-rented (vs owner occupied) 0.94 (0.81,1.10) 0.95 (0.80,1.14) 0.92 (0.69,1.24)
Social-rented (vs owner-occupied) 1.11 (0.97,1.28) 1.08 (0.92,1.27) 1.20 (0.92,1.57)
Top 20% deprived areas (vs all other areas) 1.17 (1.04,1.31) 1.16 (1.02,1.32) 1.18 (0.95,1.48)
Urban residence (vs rural) 1.20 (1.07,1.35) 1.18 (1.04,1.35) 1.28 (1.01,1.62)
Note: 95% confidence intervals (cis) are a range of likely values around the hazard ratio. Cis that do not cross 1 are statistically significant, while cis that do cross 1 are not.

Table 8B: Hazard ratios with 95% confidence intervals (cis) for drug-related deaths, overall and by sex

Category (i) Overall
(95% CI)		 (ii) Male
(95% CI)		 (iii) Female
(95% CI)
Male (vs female) 2.16 (1.87,2.50) NA NA
16-24 (vs 35-44) 1.36 (1.10,1.69) 1.56 (1.20,2.02) 1.04 (0.68,1.57)
25-34 (vs 35-44) 1.33 (1.11,1.59) 1.49 (1.19,1.85) 1.03 (0.75,1.42)
45-54 (vs 35-44) 0.47 (0.38,0.58) 0.36 (0.27,0.48) 0.69 (0.51,0.95)
55-64 (vs 35-44) 0.27 (0.21,0.36) 0.15 (0.10,0.23) 0.52 (0.35,0.77)
Single (vs married) 2.00 (1.50,2.66) 1.72 (1.19,2.49) 2.73 (1.75,4.27)
Divorced/separated (vs married) 2.37 (1.76,3.21) 1.92 (1.28,2.89) 3.05 (1.95,4.78)
Intermediate (vs degree) 1.54 (1.24,1.93) 1.91 (1.41,2.59) 1.17 (0.84,1.63)
No qualifications (vs degree) 2.09 (1.64,2.66) 2.82 (2.04,3.90) 1.28 (0.89,1.86)
Economically inactive (vs employed) 1.64 (1.38,1.94) 1.59 (1.28,1.98) 1.72 (1.30,2.29)
Unemployed (vs employed) 2.11 (1.72,2.59) 2.10 (1.66,2.64) 1.53 (0.93,2.51)
Poor mental health alone (vs none) 2.41 (1.96,2.96) 2.73 (2.10,3.55) 1.97 (1.39,2.79)
Poor mental health + other conditions (vs none) 2.53 (2.08,3.07) 2.44 (1.88,3.15) 2.54 (1.87,3.45)
Having a disability (vs not) 1.99 (1.67,2.37) 1.71 (1.36,2.13) 2.47 (1.84,3.32)
Child in a couple household (vs adult in a couple household) 0.75 (0.56,1.01) 0.90 (0.63,1.30) 0.40 (0.22,0.74)
Lone parent household (vs adult in a couple household) 1.35 (1.02,1.80) 2.09 (1.30,3.37) 1.02 (0.69,1.49)
Child in a lone parent household (vs adult in a couple household) 1.09 (0.81,1.45) 1.26 (0.88,1.80) 0.68 (0.39,1.20)
Living alone (vs adult in a couple household) 1.73 (1.35,2.23) 2.16 (1.56,2.99) 1.23 (0.82,1.86)
Living other (vs adult in a couple household) 1.23 (0.90,1.69) 1.53 (1.04,2.23) 0.71 (0.38,1.33)
Private-rented (vs owner occupied) 1.14 (0.95,1.37) 1.17 (0.94,1.47) 1.18 (0.86,1.61)
Social-rented (vs owner-occupied) 1.65 (1.40,1.94) 1.75 (1.43,2.13) 1.58 (1.20,2.10)
Top 20% deprived areas (vs all other areas) 1.14 (0.99,1.30) 1.07 (0.91,1.27) 1.26 (1.01,1.59)
Urban residence (vs rural) 1.86 (1.57,2.20) 1.94 (1.58,2.39) 1.68 (1.26,2.24)
Note: 95% confidence intervals (cis) are a range of likely values around the hazard ratio. Cis that do not cross 1 are statistically significant, while cis that do cross 1 are not.

Table 8C: Hazard ratios with 95% confidence intervals (cis) for alcohol-specific deaths, overall and by sex

Category (i) Overall
(95% CI)		 (ii) Male
(95% CI)		 (iii) Female
(95% CI)
Male (vs female) 1.78 (1.62,1.95) N/A N/A
16-24 (vs 35-44) 0.05 (0.04,0.08) 0.05 (0.03,0.08) 0.06 (0.03,0.12)
25-34 (vs 35-44) 0.32 (0.27,0.38) 0.32 (0.26,0.41) 0.32 (0.24,0.42)
45-54 (vs 35-44) 1.35 (1.21,1.51) 1.39 (1.22,1.60) 1.29 (1.08,1.54)
55-64 (vs 35-44) 1.00 (0.88,1.14) 1.13 (0.97,1.33) 0.82 (0.66,1.02)
Single (vs married) 1.61 (1.34,1.94) 1.45 (1.13,1.85) 1.86 (1.39,2.49)
Divorced/separated (vs married) 1.98 (1.65,2.38) 1.75 (1.37,2.23) 2.39 (1.81,3.15)
Intermediate (vs degree) 1.17 (1.03,1.33) 1.20 (1.02,1.42) 1.15 (0.94,1.40)
No qualifications (vs degree) 1.18 (1.03,1.35) 1.23 (1.03,1.47) 1.13 (0.90,1.41)
Economically inactive (vs employed) 1.80 (1.59,2.03) 1.89 (1.61,2.21) 1.63 (1.35,1.97)
Unemployed (vs employed) 2.25 (1.93,2.62) 2.15 (1.79,2.57) 2.51 (1.86,3.39)
Poor mental health alone (vs none) 1.33 (1.14,1.56) 1.29 (1.05,1.58) 1.40 (1.09,1.79)
Poor mental health + other conditions (vs none) 1.32 (1.16,1.49) 1.36 (1.16,1.59) 1.25 (1.01,1.54)
Having a disability (vs not) 1.81 (1.61,2.03) 1.63 (1.41,1.89) 2.12 (1.77,2.56)
Child in a couple household (vs adult in a couple household) 0.94 (0.70,1.26) 1.15 (0.81,1.64) 0.74 (0.43,1.28)
Lone parent household (vs adult in a couple household) 0.87 (0.70,1.08) 1.34 (0.96,1.87) 0.61 (0.46,0.81)
Child in a lone parent household (vs adult in a couple household) 1.35 (1.06,1.72) 1.70 (1.26,2.29) 0.91 (0.57,1.47)
Living alone (vs adult in a couple household) 1.93 (1.62,2.31) 2.44 (1.94,3.07) 1.32 (1.00,1.74)
Living other (vs adult in a couple household) 1.27 (1.00,1.61) 1.52 (1.13,2.04) 1.01 (0.67,1.52)
Private-rented (vs owner occupied) 1.26 (1.11,1.44) 1.34 (1.14,1.57) 1.13 (0.91,1.42)
Social-rented (vs owner-occupied) 1.69 (1.51,1.89) 1.80 (1.56,2.07) 1.51 (1.25,1.82)
Top 20% deprived areas (vs all other areas) 1.05 (0.95,1.15) 1.03 (0.92,1.17) 1.07 (0.91,1.26)
Urban residence (vs rural) 1.46 (1.32,1.62) 1.49 (1.31,1.70) 1.42 (1.20,1.69)
Note: 95% confidence intervals (cis) are a range of likely values around the hazard ratio. Cis that do not cross 1 are statistically significant, while cis that do cross 1 are not.