First author [ref.] | Study design/setting | Hypothesis/objectives | Participants | Participant types | Intervention and methodology | Measures and outcomes | Findings | Limitations |
Qin [26] | Retrospective nested case-controlled study | To determine the suicide risk for allergic disorders (including asthma) in patients with hospital contact | 27096 suicide cases 467571 controls (1981–2006) | Adults and children <90 yrs | Asthma status: Danish general hospital registry and Danish psychiatric central registry suicides: cause of death registry | Completed suicides | IRR risk of suicide for allergy rhinitis with bronchial asthma risk of completed suicide is 1.66 (95% CI 1.46–1.89)# p<0.01 | Only for patients who have had hospital contact (316 out of 27096 suicides) Does not have information on whole life time of patient Medications not analysed |
Kuo [10] | Prospective cohort study, community based | To determine the association between asthma and suicide mortality in high school students | 162766 (12-yr follow-up) | 11–16 yrs | Asthma status: questionnaires completed by patients and parents Suicide: national mortality database | Suicide mortality and overall deaths | Current asthma versus no asthma Cox hazard ratio 2.26 (95% CI 1.43–3.43) Chi-squared p<0.001¶ PAF of suicide accounted for by asthma was 7% | Measures of asthma limited to self-reporting Suicides not recorded as cause of death in official registers Asthma status taken at start of the study, no account of changes in asthma status Potential unmeasured confounders such as anxiety disorders |
Clarke [24] | Prospective case-controlled observational study | To examine the association between asthma and suicide ideation with and without attempts among adults in the USA | 5692 | ≥18 yrs | Asthma diagnosis and suicide ideation and attempts interviews: NCS-R surveys | Suicide ideation and suicide attempts | Significant association between asthma and suicidal ideation with attempts OR 1.53 (95% CI 1.06–2.21), but not suicide ideation alone (without attempts)+ Multiple logistic regression analysis | Self-reporting of asthma status (no validation by medical professional) Exclusion of people who died of suicide Lack of timing information Cross sectional not longitudinal |
Goodwin [5] | Cross-sectional and longitudinal observational study (three time-periods: 1981, 1982 and 1-yr follow-up 1993–1996) | To determine association between asthma and suicidal ideation, asthma and suicidal attempt, and death by suicide | 1981: 3481 completed interviews 1982: 2768 (79% of first time-period) 1-yr follow-up 1920 (69% of second time-period) | Adults >18 yrs (community) | Asthma diagnosis: self-report and follow-up questions Suicide ideation and attempts: DSMIII interviews and questions Suicide deaths: national death index | Suicidal ideation and suicidal attempts, completed suicide (lifetime longitudinally and cross sectional) | Suicidal ideation and asthma OR 2.33 (95% CI 1.03–5.25) p<0.5§ Suicide attempt and asthma OR 3.54 (95% CI 1.4–8.99) p<0.5§ Multiple logistic regression | Asthma diagnosis limited to self-reporting No data on lung function tests No information on specific forms of asthma treatment, frequency and timing Small numbers of participants in subset groups limit significance in findings |
Goodwin [23] | Data from a previous cross-sectional epidemiological study [27, 28] | To determine the association between asthma and suicide ideation among youths in the community | 1285 | Youths 9–17 yrs | One child and one parent/guardian in each family were interviewed, detailed information on the child including psychiatric diagnosis was obtained | Suicide ideation | Hospitalised patients for asthma have higher levels of suicide ideation compared those without asthma OR 3.25 (95% CI 1.04–10.1)ƒ p<0.01## Logistic regression | Patients who have been hospitalised Used epidemiological diagnostic survey instruments rather than clinical administered diagnostic interview Asthma diagnosis based on mothers' self-reports |
Goodwin [25] | Case controlled | An association between asthma and increased risk of mental disorders in primary care and suicide ideation | 1005 patients recruited | Adults 18–70yrs (waiting room of internal medicine clinic) | Asthma diagnosis: ICD-9 code on billing information Suicide ideation: questioning Mental disorders: questionnaire (PRIME-MD PHQ) | Suicide ideation (other outcomes, risk of mental disorders) | Asthma and suicide ideation association (statistically significant p=0.003)¶¶ OR 1.9 (95% CI 1.03–3.4)++ (multivariate logistic regression analysis) | Patients with severe suicide ideation were excluded Did not use objective respiratory measures No consideration of effects of medication No information on cigarette smoking or psychiatric disorders Subjects were of lower socioeconomic status and poorly educated |
NCS-R: National Comorbidity Survey-Replication; IRR: incidence rate ratios; PAF: population attributable fraction; DSM-III: Diagnostic and Statistical Manual of Mental Disorders, 3rd edition; ICD: International Classification of Disease; PRIME-MD: Primary Care Evaluation of Mental Disorders; PHQ: patient health questionnaire. #: adjusted for age, sex and calendar time (p-value, Wald test); ¶: adjusted for sex, age, cigarette smoking by at least one member of the family, cigarette smoking and allergic rhinitis; +: adjusted for socio-demographic, age, sex, race/ethnicity/smoking and nicotine dependence, depression, anxiety, alcohol dependence and mental health factors; §: adjusted for age, sex, lifetime major depression and asthma treatment (p-values, Fischer exact test); ƒ: adjusted for sex, age calendar time through matching; ##: p-value, Wald test; ¶¶: Chi-squared test; ++: adjusted for age, sex, race, marital status, education and comorbid mental disorders.