0.05). The net benefit of model was improved and cognition resulted in continuous reclassification improvement with NRI of 5.3%. Several clinical dimensions were significant predictors in the XGBoost classification algorithm. Limitations A limited sample size and failure to include sufficient suicide risk factors in the predictive model. Conclusion This study demonstrate that cognitive deficits may serve as an important risk factor to predict suicide attempts in patients with MDD. Combined with other demographic characteristics and attributes drawn from clinical questionnaires, cognitive function can improve the predictive effectiveness of the ML model. Additionally, explanatory ML models can help clinicians detect specific risk factors for each suicide attempter within MDD patients. These findings may be helpful for clinicians to detect those at high risk of suicide attempts quickly and accurately, and help them make proactive treatment decisions."/>
Skip to main content

Table 1 Demographic and clinical features, as well as comparison results without adjusting covariates

From: Can cognition help predict suicide risk in patients with major depressive disorder? A machine learning study

Items DSA* (N = 52) DNS* (N = 61) HC* (N = 98) F/x2 Significant direction
Mean (sd) / N (%) Mean (sd) / N (%) Mean (sd) / N (%) (p-value)
Age 25.54 (8.55) 29.87 (11.50) 29.56 (7.11) 4.160 DSA < DNS、 HC
(0.017)
Gender Male 12 (23.1%) 25 (41.0%) 49 (50.0%) 10.152 DSA: male < female
Female 40 (76.9%) 36 (59.0%) 49 (50.0%) (0.006)
Education level  ≥ 12 years 22 (42.3%) 31 (50.1%) 47 (48.0%) 0.839 (0.657)  
 < 12 years 30 (57.7%) 30 (49.9%) 51 (52.0%)  
suicide attempt lethality High-lethality 9(17.3%) \ \ \ \
low-lethality 43(82.7%)
Suicide attempt recency Within one week 9 (17.3%) \ \ \ \
Not within one week 43 (82.7%)
IGT* performance
 Net total score -13.73 (28.883) -16.53 (33.422) -10.06 (34.148) 0.755(0.471)  
 PartA(Trails 1–40) -4.54 (12.221) -6.27 (13.956) -9.35 (12.315) 2.666(0.072)  
 PartB(Trails 41–100) -9.17 (19.571) -10.03 (20.881) -0.735 (25.015) 3.838(0.023) HC > DSAˎDNS
 Block1(Trials 1–20) -3 (6.669) -3.8 (7.269) -5.27 (6.725) 2.056(0.131)  
 Block2(Trials 21–40) -1.54 (7.868) -2.47 (8.848) -4.08 (7.896) 1.818(0.165)  
 Block3(Trials 41–60) -2.38 (7.118) -4.07 (9.072) -1.55 (9.079) 1.583(0.208)  
 Block4(Trials 61–80) -3.94 (7.092) -3.67 (8.471) -0.51 (9.465) 3.782(0.024) HC > DSAˎDNS
 Block5(Trials 81–100) -2.85 (9.712) -2.34 (9.345) 1.33 (9.877) 4.261(0.015) HC > DSAˎDNS
SST* performance
 positiveRT   726.51(261.209) 690.991(240.908) \ -0.751(0.454)  
 negativeRT   741.257(256.722) 686.08(246.231) \ -1.164(0.247)  
 neutralRT   743.001(243.265) 670.407(235.604) \ -1.608(0.111)  
 suicideRT   756.389(281.243) 698.087(253.441) \ -1.159(0.249)  
 "suicide"word RT   756.81(361.531) 684.317(297.816) \ -1.153(0.251)  
 HAMD-24*   36.212(6.241) 29.016(6.828) \ -5.807(0) DSA > DNA
 CTQ*score   53.94(12.299) 47.33(14.048) \ -2.640(0.009) DSA > DNA
  1. *DSA Depressed suicide attempter, DNS Depressed non-suicide attempter, HC Healthy control
  2. IGT IOWA gambling task, SST Suicide stroop task, HAMD-24 Hamilton Depression Scale-24 items, CTQ Childhood Trauma Questionnaire