To save content items to your account,
please confirm that you agree to abide by our usage policies.
If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account.
Find out more about saving content to .
To save content items to your Kindle, first ensure firstname.lastname@example.org
is added to your Approved Personal Document E-mail List under your Personal Document Settings
on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part
of your Kindle email address below.
Find out more about saving to your Kindle.
Note you can select to save to either the @free.kindle.com or @kindle.com variations.
‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi.
‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
Posttraumatic stress symptoms (PTSS) are common following traumatic stress exposure (TSE). Identification of individuals with PTSS risk in the early aftermath of TSE is important to enable targeted administration of preventive interventions. In this study, we used baseline survey data from two prospective cohort studies to identify the most influential predictors of substantial PTSS.
Self-identifying black and white American women and men (n = 1546) presenting to one of 16 emergency departments (EDs) within 24 h of motor vehicle collision (MVC) TSE were enrolled. Individuals with substantial PTSS (⩾33, Impact of Events Scale – Revised) 6 months after MVC were identified via follow-up questionnaire. Sociodemographic, pain, general health, event, and psychological/cognitive characteristics were collected in the ED and used in prediction modeling. Ensemble learning methods and Monte Carlo cross-validation were used for feature selection and to determine prediction accuracy. External validation was performed on a hold-out sample (30% of total sample).
Twenty-five percent (n = 394) of individuals reported PTSS 6 months following MVC. Regularized linear regression was the top performing learning method. The top 30 factors together showed good reliability in predicting PTSS in the external sample (Area under the curve = 0.79 ± 0.002). Top predictors included acute pain severity, recovery expectations, socioeconomic status, self-reported race, and psychological symptoms.
These analyses add to a growing literature indicating that influential predictors of PTSS can be identified and risk for future PTSS estimated from characteristics easily available/assessable at the time of ED presentation following TSE.
Most criminological theories are not truly scientific, since they do not yield exact quantitative predictions of criminal career features, such as the prevalence and frequency of offending at different ages. This Element aims to make progress towards more scientific criminological theories. A simple theory is described, based on measures of the probability of reoffending and the frequency of offending. Three offender categories are identified: high risk/high rate, high risk/low rate, and low risk/low rate. It is demonstrated that this theory accurately predicts key criminal career features in three datasets: in England the Offenders Index (national data), the Cambridge Study in Delinquent Development (CSDD) and in America the Pittsburgh Youth Study (PYS). The theory is then extended in the CSDD and PYS by identifying early risk factors that predict the three categories. Criminological theorists are encouraged to replicate and build on our research to develop scientific theories that yield quantitative predictions.
Young people are most vulnerable to suicidal behaviours but least likely to seek help. A more elaborate study of the intrinsic and extrinsic correlates of suicidal ideation and behaviours particularly amid ongoing population-level stressors and the identification of less stigmatising markers in representative youth populations is essential.
Participants (n = 2540, aged 15–25) were consecutively recruited from an ongoing large-scale household-based epidemiological youth mental health study in Hong Kong between September 2019 and 2021. Lifetime and 12-month prevalence of suicidal ideation, plan, and attempt were assessed, alongside suicide-related rumination, hopelessness and neuroticism, personal and population-level stressors, family functioning, cognitive ability, lifetime non-suicidal self-harm, 12-month major depressive disorder (MDD), and alcohol use.
The 12-month prevalence of suicidal ideation, ideation-only (no plan or attempt), plan, and attempt was 20.0, 15.4, 4.6, and 1.3%, respectively. Importantly, multivariable logistic regression findings revealed that suicide-related rumination was the only factor associated with all four suicidal outcomes (all p < 0.01). Among those with suicidal ideation (two-stage approach), intrinsic factors, including suicide-related rumination, poorer cognitive ability, and 12-month MDE, were specifically associated with suicide plan, while extrinsic factors, including coronavirus disease 2019 (COVID-19) stressors, poorer family functioning, and personal life stressors, as well as non-suicidal self-harm, were specifically associated with suicide attempt.
Suicide-related rumination, population-level COVID-19 stressors, and poorer family functioning may be important less-stigmatising markers for youth suicidal risks. The respective roles played by not only intrinsic but also extrinsic factors in suicide plan and attempt using a two-stage approach should be considered in future preventative intervention work.
This study is performed to figure out how the presence of diabetes affects the infection, progression and prognosis of 2019 novel coronavirus disease (COVID-19), and the effective therapy that can treat the diabetes-complicated patients with COVID-19. A multicentre study was performed in four hospitals. COVID-19 patients with diabetes mellitus (DM) or hyperglycaemia were compared with those without these conditions and matched by propensity score matching for their clinical progress and outcome. Totally, 2444 confirmed COVID-19 patients were recruited, from whom 336 had DM. Compared to 1344 non-DM patients with age and sex matched, DM-COVID-19 patients had significantly higher rates of intensive care unit entrance (12.43% vs. 6.58%, P = 0.014), kidney failure (9.20% vs. 4.05%, P = 0.027) and mortality (25.00% vs. 18.15%, P < 0.001). Age and sex-stratified comparison revealed increased susceptibility to COVID-19 only from females with DM. For either non-DM or DM group, hyperglycaemia was associated with adverse outcomes, featured by higher rates of severe pneumonia and mortality, in comparison with non-hyperglycaemia. This was accompanied by significantly altered laboratory indicators including lymphocyte and neutrophil percentage, C-reactive protein and urea nitrogen level, all with correlation coefficients >0.35. Both diabetes and hyperglycaemia were independently associated with adverse prognosis of COVID-19, with hazard ratios of 10.41 and 3.58, respectively.
We investigated risk factors associated with COVID-19 by conducting a retrospective, frequency-matched case-control study, with three sampling periods (August–October 2020). We compared cases completing routine contact tracing to asymptomatic population controls. Multivariable analyses estimated adjusted odds ratios (aORs) for non-household community settings. Meta-analyses using random effects provided pooled odds ratios (pORs). Working in healthcare (pOR 2.87; aORs 2.72, 2.81, 3.08, for study periods 1–3 respectively), social care (pOR 4.15; aORs 2.46, 5.06, 5.41, for study periods 1–3 respectively) or hospitality (pOR 2.36; aORs 2.01, 2.54, 2.63, for study periods 1–3 respectively) were associated with increased odds of being a COVID-19 case. Additionally, working in bars, pubs and restaurants, warehouse settings, construction, educational settings were significantly associated. While definitively determining where transmission occurs is impossible, we provide evidence that in certain sectors, the impact of mitigation measures may only be partial and reinforcement of measures should be considered in these settings.
Using nested case–control data from the Lifelines COVID-19 cohort, we undertook a validation study of a clinical and genetic model to predict the risk of severe COVID-19 in people with confirmed COVID-19 and in people with confirmed or self-reported COVID-19. The model performed well in terms of discrimination of cases and controls for all ages (area under the receiver operating characteristic curve (AUC) = 0.680 for confirmed COVID-19 and AUC = 0.689 for confirmed and self-reported COVID-19) and in the age group in which the model was developed (50 years and older; AUC = 0.658 for confirmed COVID-19 and AUC = 0.651 for confirmed and self-reported COVID-19). There was no evidence of over- or under-dispersion of risk scores but there was evidence of overall over-estimation of risk in all analyses (all P < 0.0001). In the light of large numbers of people worldwide remaining unvaccinated and continuing uncertainty regarding vaccine efficacy over time and against variants of concern, identification of people at high risk of severe COVID-19 may encourage the uptake of vaccinations (including boosters) and the use of non-pharmaceutical inventions.
Bullying victimisation has been associated with increased risk of suicide ideation and attempt throughout the lifespan, but no study has yet examined whether it translates to a greater risk of death by suicide. We aimed to determine the association of bullying victimisation with suicide mortality.
Participants were drawn from the 1958 British birth cohort, a prospective follow-up of all births in 1 week in Britain in 1958. We conducted logistic regressions on 14 946 participants whose mothers reported bullying victimisation at 7 and 11 years with linked information on suicide deaths through the National Health Service Central Register.
Fifty-five participants (48 males) had died by suicide between the age 18 and 52 years. Bullying victimisation was associated with suicide mortality; a one standard deviation increases in bullying victimisation linked to an increased odds for suicide mortality [odds ratio (OR) 1.29; 1.02–1.64] during adulthood. The OR attenuated by 11% after adjustment for individual (e.g. behavioural and emotional problems) and familial characteristics (e.g. adverse childhood experiences, 1.18; 0.92–1.51). Analysis of bullying victimisation frequency categories yields similar results: compared with individuals who had not been bullied, those who had been frequently bullied had an increased odds for suicide mortality (OR 1.89; 0.99–3.62).
Our study suggests that individuals who have been frequently bullied have a small increased risk of dying by suicide, when no other risk factors is considered. Suicide prevention might start in childhood, with bullying included in a range of inter-correlated vulnerabilities encompassing behavioural and emotional difficulties and adverse experiences within the family.
Circulating branched-chain amino acids (BCAAs), a subgroup of the nine essential amino acids, have been associated to pancreatic cancer risk. The aim of this study is to estimate the relation between BCAA intake from diet and pancreatic cancer risk.
We analysed data from a multicentric Italian case-control study, including 326 pancreatic cancer cases and 652 controls, matched to cases by study centre, sex and age. A validated food-frequency questionnaire was used to collect the participants’ usual diet before cancer diagnosis (or hospital admission for controls) and to compute dietary intakes of various nutrients, including BCAAs. Odds ratios (ORs) and corresponding confidence intervals (CIs) were computed through logistic regression models conditioned on the matching variables and adjusted for major confounding factors, including total energy intake.
We found a positive association between the BCAA intake and pancreatic cancer risk (OR for the third quartile=1.88, 95% CI=1.08-3.26; OR for the fourth quartile =2.17, 95% CI=1.17-4.06), with a significant trend in risk. The association persisted after excluding subjects with diabetes and family history of pancreatic cancer, and across strata of selected covariates.
These data support and quantify the association between dietary BCAAs and pancreatic cancer, previously suggested by studies on circulating BCAAs.
Little is known about the childhood antecedents and adult correlates of adolescent dual-harm (i.e. co-occurring self- and other-harm). We examine the longitudinal associations between (a) social and psychological risk factors in childhood and adolescent dual-harm and (b) adolescent dual-harm and social and mental health impairments in early adulthood.
Participants (N = 1482) are from a prospective longitudinal community-representative study. Dual-, self-, and other-harm were self-reported at ages 13, 15, and 17. Social and psychological risk factors in childhood were assessed between 7 and 11; early adult correlates at age 20. Groups with dual-harm, self-harm only, other-harm only, and no harm were compared.
Between 13 and 17, 7.2% of adolescents reported dual-harm (self-harm only: 16.2%; other-harm only: 13.3%). Some childhood risk factors (e.g. sensation-seeking, parental divorce, victimization by peers) characterized all harm groups; others were common to the dual- and self-harm (anxiety/depressive symptoms, relational aggression) or dual- and other-harm groups only (low self-control, substance use, delinquency). Adolescents with dual-harm had reported more physical aggression and harsh parenting, and lower school bonding in childhood than any other group. In early adulthood, they reported more anxiety/depressive symptoms, psychopathy symptoms, homicidal ideations, delinquency, and victimization experiences than any other group.
Adolescent dual-harm follows psychological problems and social disconnection in childhood and signals risk of psychopathology and isolation in early adulthood. To curb the burden from dual-harm, interventions must target adolescents, families, peer networks, and school environments. Differentiating youth with dual-harm from those with single-harm is important for developing personalized treatments.
The COVID-19 pandemic has had profound consequences for population mental health. However, it is less clear for whom these effects are sustained.
To investigate the prevalence, incidence, prognosis and risk factors for symptoms of depression and anxiety in a UK cohort over three distinct periods in the pandemic in 2020.
An online survey was completed by a UK community cohort at three points (n = 3097 at baseline, n = 878 completed all surveys): April (baseline), July to September (time point 2) and November to December (time point 3). Participants completed validated measures of depression and anxiety on each occasion, and we prospectively explored the role of sociodemographic and psychological factors (loneliness, positive mood and perceived risk of and worry about COVID-19) as risk factors.
Depression (Patient Health Questionnaire-9 means: baseline, 7.69; time point 2, 5.53; time point 3, 6.06) and anxiety scores (Generalised Anxiety Disorder-7 means: baseline, 6.59; time point 2, 4.60; time point 3, 4.98) were considerably greater than pre-pandemic population norms at all time points. Women reported greater depression and anxiety symptoms than men. Younger age, history of mental health disorder, more COVID-19-related negative life events, greater loneliness and lower positive mood at baseline were all significant predictors of poorer mental health at time point 3.
The negative impact of the COVID-19 pandemic on mental health has persisted to some degree. Younger people and individuals with prior mental health disorders are at greatest risk. Easing of restrictions and resumption of social interaction could mitigate the risk factors of loneliness and positive mood.
Recent dimensional models of adversity informed by a neurobiological deficit framework highlights threat and deprivation as core dimensions, whereas models informed by an evolutionary, adaptational and functional framework calls attention to harshness and unpredictability. This report seeks to evaluate an integrative model of threat, deprivation, and unpredictability, drawing on the Fragile Families Study. Confirmatory factor analysis of presumed multiple indicators of each construct reveals an adequate three-factor structure of adversity. Theory-based targeted predictions of the developmental sequelae of each dimension also received empirical support, with deprivation linked to health problems and cognitive ability; threat linked to aggression; and unpredictability to substance use and sexual risk-taking. These findings lend credibility to utility of the three-dimensional integrative framework of adversity. It could thus inform development of dimensional measures of risk assessment and exploration of multidimensional adversity profiles, sensitive to individual differences in lived experiences, supporting patient-centered, strength-based approaches to services.
The coronavirus disease 2019 (COVID-19) pandemic might affect mental health. Data from population-representative panel surveys with multiple waves including pre-COVID data investigating risk and protective factors are still rare.
In a stratified random sample of the German household population (n = 6684), we conducted survey-weighted multiple linear regressions to determine the association of various psychological risk and protective factors assessed between 2015 and 2020 with changes in psychological distress [(PD; measured via Patient Health Questionnaire for Depression and Anxiety (PHQ-4)] from pre-pandemic (average of 2016 and 2019) to peri-pandemic (both 2020 and 2021) time points. Control analyses on PD change between two pre-pandemic time points (2016 and 2019) were conducted. Regularized regressions were computed to inform on which factors were statistically most influential in the multicollinear setting.
PHQ-4 scores in 2020 (M = 2.45) and 2021 (M = 2.21) were elevated compared to 2019 (M = 1.79). Several risk factors (catastrophizing, neuroticism, and asking for instrumental support) and protective factors (perceived stress recovery, positive reappraisal, and optimism) were identified for the peri-pandemic outcomes. Control analyses revealed that in pre-pandemic times, neuroticism and optimism were predominantly related to PD changes. Regularized regression mostly confirmed the results and highlighted perceived stress recovery as most consistent influential protective factor across peri-pandemic outcomes.
We identified several psychological risk and protective factors related to PD outcomes during the COVID-19 pandemic. A comparison of pre-pandemic data stresses the relevance of longitudinal assessments to potentially reconcile contradictory findings. Implications and suggestions for targeted prevention and intervention programs during highly stressful times such as pandemics are discussed.
This study describes risk factors associated with mortality among COVID-19 cases reported in the WHO African region between 21 March and 31 October 2020. Average hazard ratios of death were calculated using weighted Cox regression as well as median time to death for key risk factors. We included 46 870 confirmed cases reported by eight Member States in the region. The overall incidence was 20.06 per 100 000, with a total of 803 deaths and a total observation time of 3 959 874 person-days. Male sex (aHR 1.54 (95% CI 1.31–1.81); P < 0.001), older age (aHR 1.08 (95% CI 1.07–1.08); P < 0.001), persons who lived in a capital city (aHR 1.42 (95% CI 1.22–1.65); P < 0.001) and those with one or more comorbidity (aHR 36.37 (95% CI 20.26–65.27); P < 0.001) had a higher hazard of death. Being a healthcare worker reduced the average hazard of death by 40% (aHR 0.59 (95% CI 0.37–0.93); P = 0.024). Time to death was significantly less for persons ≥60 years (P = 0.038) and persons residing in capital cities (P < 0.001). The African region has COVID-19-related mortality similar to that of other regions, and is likely underestimated. Similar risk factors contribute to COVID-19-associated mortality as identified in other regions.
To estimate the prevalence of vitamin A deficiency (VAD) in children and associated risk factors.
Analysis of data from a cross-sectional multicentre study performed in the primary care units of the municipalities from January to June 2015. The children’s legal guardians answered a socio-economic questionnaire, and the children’s blood samples were obtained by venipuncture. Plasma retinol was determined by HPLC. Plasma retinol values of <0·70 μmol/l were considered VDA. Poisson multiple regression with robust variance was used. Values of P < 0·05 were considered significant. The data were analysed in the SPSS software, 21.0.
Forty-eight poorest municipalities in the South Region of Brazil.
Children (n 1503) aged 12–59 months.
The prevalence of VAD in the sample was 1·9 % (95 % CI (0·5, 6·8)). The following risk factors were associated with the outcome in the final explanatory model: family received Bolsa Familia program benefits (PR = 3·19; 95 % CI (1·69, 6·02)), child was not being breastfed (PR = 5·22; 95 % CI (1·68, 16·18)) and stunting (PR = 4·75; 95 % CI (2·10, 10·73)).
VAD did not represent a public health problem for children living in socio-economically vulnerable municipalities in the South Region of Brazil, suggesting a new panorama of this nutritional deficiency even in regions of low socio-economic conditions in these three states. Thus, in view of the current nutritional transition scenario, it is necessary to continuously monitor and improve public policies related to vitamin A supplementation in the country.
The study aim was to examine the incidence and risk factors of respiratory syncytial virus (RSV) bronchiolitis hospitalisations and disease severity among infants. We compared demographic and health characteristics of children aged 0–23 hospitalised for RSV bronchiolitis (cases, n = 1227) during 2008–2018 and control children (n = 554) of the same age admitted for non-respiratory disease. RSV antigen was detected in nasal swabs by immunochromatography. Multiple logistic regression models were applied. The average annual incidence of hospitalisation for RSV bronchiolitis was 12.6 per 1000 and 1.7 per 1000 (P < 0.001) among infants and toddlers, respectively, with winter seasonality (November–March). The risk of hospitalisation for RSV bronchiolitis increased among children aged 0–5 months (OR 7.66; 95% CI 5.61–10.45) and 6–11 months (OR 12.88, 95% CI 8.48–19.55), compared to those aged 12–23 months. Additional risk factors were living in low vs. higher socio-economic status towns (OR 1.49; 95% CI 1.14–1.95), having chronic medical conditions (OR 2.75; 95% CI 1.61–4.70), birth month (October–January vs. June–September) (OR 2.19; 95% CI 1.60–2.99) and history of stay in neonatal intensive care unit at birth (OR 2.37; 95% CI 1.27–4.41). Male children and those who had pneumonia were more likely to have severe RSV bronchiolitis. In conclusion, the burden of hospitalisations for RSV bronchiolitis is high, especially in young infants. Effective preventive measures such as RSV active vaccines can reduce the risk of hospitalisations for RSV bronchiolitis among these vulnerable groups.
Cholesteatoma is a benign but destructive epithelial lesion in the middle ear and/or mastoid. It is hard to translate data from previous research to daily clinical practice. In this study, factors influencing recurrence rates in daily clinical practice were identified.
The study included 67 patients who were treated for a cholesteatoma with combined approach tympanoplasty. The average follow-up time was 35 months.
The recurrence rate was 23.3 per cent in adults and 45.5 per cent in children. Predictors of recurrence were younger age and a low tegmen. A cholesteatoma in a child and the simultaneous presence of a low tegmen led to recurrence in 82.8 per cent of the patients.
Patients – especially children – with a low tegmen have an increased risk of recurrence. It is recommended that ENT surgeons be aware of recurrence in children, particularly in the case of a low tegmen.
Evidence on risk factors for postpartum depression (PPD) are fragmented and inconsistent.
To assess the strength and credibility of evidence on risk factors of PPD, ranking them based on the umbrella review methodology.
Databases were searched until 1 December 2020, for systematic reviews and meta-analyses of observational studies. Two reviewers assessed quality, credibility of associations according to umbrella review criteria (URC) and evidence certainty according to Grading of Recommendations-Assessment-Development-Evaluations criteria.
Including 185 observational studies (n = 3 272 093) from 11 systematic reviews, the association between premenstrual syndrome and PPD was the strongest (highly suggestive: odds ratio 2.20, 95%CI 1.81–2.68), followed by violent experiences (highly suggestive: odds ratio (OR) = 2.07, 95%CI 1.70–2.50) and unintended pregnancy (highly suggestive: OR=1.53, 95%CI 1.35–1.75). Following URC, the association was suggestive for Caesarean section (OR = 1.29, 95%CI 1.17–1.43), gestational diabetes (OR = 1.60, 95%CI 1.25–2.06) and 5-HTTPRL polymorphism (OR = 0.70, 95%CI 0.57–0.86); and weak for preterm delivery (OR = 2.12, 95%CI 1.43–3.14), anaemia during pregnancy (OR = 1.47, 95%CI 1.17–1.84), vitamin D deficiency (OR = 3.67, 95%CI 1.72–7.85) and postpartum anaemia (OR = 1.75, 95%CI 1.18–2.60). No significant associations were found for medically assisted conception and intra-labour epidural analgesia. No association was rated as ‘convincing evidence’. According to GRADE, the certainty of the evidence was low for Caesarean section, preterm delivery, 5-HTTLPR polymorphism and anaemia during pregnancy, and ‘very low’ for remaining factors.
The most robust risk factors of PDD were premenstrual syndrome, violent experiences and unintended pregnancy. These results should be integrated in clinical algorithms to assess the risk of PPD.
Self-harm in pregnancy or the year after birth (‘perinatal self-harm’) is clinically important, yet prevalence rates, temporal trends and risk factors are unclear.
A cohort study of 679 881 mothers (1 172 191 pregnancies) was conducted using Danish population register data-linkage. Hospital treatment for self-harm during pregnancy and the postnatal period (12 months after live delivery) were primary outcomes. Prevalence rates 1997–2015, in women with and without psychiatric history, were calculated. Cox regression was used to identify risk factors.
Prevalence rates of self-harm were, in pregnancy, 32.2 (95% CI 28.9–35.4)/100 000 deliveries and, postnatally, 63.3 (95% CI 58.8–67.9)/100 000 deliveries. Prevalence rates of perinatal self-harm in women without a psychiatric history remained stable but declined among women with a psychiatric history. Risk factors for perinatal self-harm: younger age, non-Danish birth, prior self-harm, psychiatric history and parental psychiatric history. Additional risk factors for postnatal self-harm: multiparity and preterm birth. Of psychiatric conditions, personality disorder was most strongly associated with pregnancy self-harm (aHR 3.15, 95% CI 1.68–5.89); psychosis was most strongly associated with postnatal self-harm (aHR 6.36, 95% CI 4.30–9.41). For psychiatric disorders, aHRs were higher postnatally, particularly for psychotic and mood disorders.
Perinatal self-harm is more common in women with pre-existing psychiatric history and declined between 1997 and 2015, although not among women without pre-existing history. Our results suggest it may be a consequence of adversity and psychopathology, so preventative intervention research should consider both social and psychological determinants among women with and without psychiatric history.
Previous research has highlighted the importance of understanding which psychosocial factors distinguish between those with suicide thoughts compared to those who attempt suicide. This study aims to investigate these distinguishing factors further within an ideation-to-action framework and to explore sex differences in suicide risk.
Participants (n = 7546, aged 16+) were from the cross-sectional Adult Psychiatric Morbidity Survey (APMS; 2014) of England. Face-to-face and self-completion questionnaires assessed lifetime suicidal ideation, lifetime suicide attempts, demographic characteristics, life experiences, social support, health and mental illness. Multinomial logistic regression examined factors differentiating between those with suicidal ideation only and suicide attempt histories (with or without suicidal ideation) in men and women.
Overall men were less likely to report suicidal thoughts and attempts, compared to females. More factors differentiated between suicidal thoughts and attempts in women compared to in men; these included hospital admission for mental illness, below degree level qualifications, being single and childhood adversity. In men, factors which significantly differentiated between suicidal thoughts and attempts included self-report of professional diagnosis of mental illness and childhood adversity. Higher levels of social support were associated with being in the suicidal thoughts group v. in the attempts group in men.
This study identified some key differences between men and women in factors associated with suicide attempts compared to suicidal thoughts. The findings support the use of the ideation-to-action framework to investigate sex differences in suicidal behaviour. Future research should examine the extent to which these factors are associated with suicide risk over time.