Impact of social disparities on risk factors for suicidal ideation and suicide attempt among commercially insured youth and adults in the US

BACKGROUND: To study the role of social disparities on the effect of risk factors on suicidal ideation (SI) and suicide attempt (SA) among commercially insured youth and adults in the US. METHODS: A national level retrospective cohort study was conducted using health insurance claims data from four major insurance companies in the US. The cohort was defined by patients having a mental health or substance use disorder (MH/SUD) related outpatient encounter (index encounter). We used Cox proportional hazards models to evaluate the impact of social disparities on the effect of risk factors on patients’ future SI and SA. Risk factors considered in the models consist of a mix of long-, mid-, and short-term prior comorbidities and prescriptions that are identified as important factors for predicting SA by Simon et al. 2018. Social disparities were summarized by the social deprivation index (SDI) at patients’ zip code level, and then categorized into five quintiles; higher SDI quintiles indicated higher levels of social deprivation. Patients were censored if they did not have an SI or SA diagnosis before 9/30/2015 (ICD-9 to ICD-10 switch date) or until the last day of the insurance plan enrollment, whichever came first. RESULTS: Between 2014-2015, we identified 317,383 patients <65 years old with an index encounter. Among them, 124,424 aged <25 (youth and young adults; hereinafter youth) and 192,959 aged between 25-64 (other adults; hereinafter adults). Prevalence of SI and SA were both higher in youth (7.09% and 1.86%, respectively, vs. 3.00% and 0.79% in adults). Survival analysis showed that SDI impacted the behaviors of SI and SA differently for youth and adults. Among youth, SDI impacted the effects of risk factors for both SI and SA after controlling for demographic variables. For SI, SDI impacted the effects of age, alcohol use disorder diagnosis in the past 3 years, SA in the past year, and MH emergency department (ED) visits in the past 3 months. Overall, the effect of age on SI decreased as SDI quintile increased. Alcohol use disorder diagnosis in the past 3 years had a positive impact on SI for patients from neighborhoods in the 2nd SDI quintile (HR=1.25). SA in the past year had a negative impact on SI for patients from neighborhoods in the 3rd SDI quintile (HR=0.55). MH ED visits in the past 3 months had a positive impact on SI for patients from neighborhoods in the 2nd and 4th SDI quintiles (HR=1.30 and 1.63, respectively). For SA, SDI impacted the effects of SA in the past 3 years, SA and schizophrenia in the past 3 years, MH ED visits in the past year, and Benzodiazepine prescription in the past 3 months. Overall, the effect of SA in the past 3 years on SA increased as SDI increased. SA and schizophrenia in the past 3 years had a positive impact on SA for patients from neighborhoods in the 2nd SDI quintile (HR=3.09). MH ED visits in the past year had a positive impact on SA for patients from neighborhoods of 1st SDI quintile (HR=1.60). Benzodiazepine prescription in the past 3 months had a positive impact on SA for patients from neighborhoods in the 2nd SDI quintile (HR=1.91). Among adults, patients in poorer neighborhoods were positively associated with having SI (HRs=1.09, 1.13, 1.19, and 1.25 for SDI quintiles 2nd, 3rd, 4th, 5th vs. 1st), however having SA was not affected by the social deprivation level of patients’ neighborhoods (p=0.21). DISCUSSION: Among commercially insured patients in the US, youth had higher prevalence of SI and SA when compared to adults. Neighborhood level social disparities impacted the behaviors of SI and SA differently for youth and adults. Among youth, social disparities nonlinearly impacted the effects of risk factors for both SI and SA. Among adults, social disparities were significantly associated with SI, but not with SA.