Combining clinical variables to optimize prediction of antidepressant treatment outcomes

Raquel Iniesta; Karim Malki; Wolfgang Maier; Marcella Rietschel; Ole Mors; Joanna Hauser; Neven Henigsberg; Mojca Zvezdana Dernovsek; Daniel Souery; Daniel Stahl; Richard Dobson; Katherine J. Aitchison; Anne Farmer; Cathryn M. Lewis; Peter McGuffin; Rudolf Uher

doi:10.1016/j.jpsychires.2016.03.016

Combining clinical variables to optimize prediction of antidepressant treatment outcomes

Raquel Iniesta, Karim Malki, Wolfgang Maier, Marcella Rietschel, Ole Mors, Joanna Hauser, Neven Henigsberg, Mojca Zvezdana Dernovsek, Daniel Souery, Daniel Stahl, Richard Dobson, Katherine J. Aitchison, Anne Farmer, Cathryn M. Lewis, Peter McGuffin, Rudolf Uher

Research output: Contribution to journal › Article › peer-review

115 Citations (Scopus)

420 Downloads (Pure)

Abstract

The outcome of treatment with antidepressants varies markedly across people with the same diagnosis. A clinically significant prediction of outcomes could spare the frustration of trial and error approach and improve the outcomes of major depressive disorder through individualized treatment selection. It is likely that a combination of multiple predictors is needed to achieve such prediction. We used elastic net regularized regression to optimize prediction of symptom improvement and remission during treatment with escitalopram or nortriptyline and to identify contributing predictors from a range of demographic and clinical variables in 793 adults with major depressive disorder. A combination of demographic and clinical variables, with strong contributions from symptoms of depressed mood, reduced interest, decreased activity, indecisiveness, pessimism and anxiety significantly predicted treatment outcomes, explaining 5–10% of variance in symptom improvement with escitalopram. Similar combinations of variables predicted remission with area under the curve 0.72, explaining approximately 15% of variance (pseudo R2) in who achieves remission, with strong contributions from body mass index, appetite, interest-activity symptom dimension and anxious-somatizing depression subtype. Escitalopram-specific outcome prediction was more accurate than generic outcome prediction, and reached effect sizes that were near or above a previously established benchmark for clinical significance. Outcome prediction on the nortriptyline arm did not significantly differ from chance. These results suggest that easily obtained demographic and clinical variables can predict therapeutic response to escitalopram with clinically meaningful accuracy, suggesting a potential for individualized prescription of this antidepressant drug.

Original language	English
Pages (from-to)	94-102
Number of pages	9
Journal	Journal of psychiatric research
Volume	78
Early online date	1 Apr 2016
DOIs	https://doi.org/10.1016/j.jpsychires.2016.03.016
Publication status	Published - Jul 2016

Keywords

Depression
Outcome
Antidepressant
Prediction
Machine learning
Statistical learning

Access to Document

10.1016/j.jpsychires.2016.03.016Licence: CC BY-NC-ND

Combining clinical variables to_INIESTA_Acc30Mar2016Epub1Apr2016_GOLD VoR(CC BY NC ND)Final published version, 523 KBLicence: CC BY-NC-ND

Cite this

Iniesta, R., Malki, K., Maier, W., Rietschel, M., Mors, O., Hauser, J., Henigsberg, N., Dernovsek, M. Z., Souery, D., Stahl, D., Dobson, R., Aitchison, K. J., Farmer, A., Lewis, C. M., McGuffin, P., & Uher, R. (2016). Combining clinical variables to optimize prediction of antidepressant treatment outcomes. Journal of psychiatric research, 78, 94-102. https://doi.org/10.1016/j.jpsychires.2016.03.016

@article{6d0fedb996aa42cca6269ba0c240023e,

title = "Combining clinical variables to optimize prediction of antidepressant treatment outcomes",

abstract = "The outcome of treatment with antidepressants varies markedly across people with the same diagnosis. A clinically significant prediction of outcomes could spare the frustration of trial and error approach and improve the outcomes of major depressive disorder through individualized treatment selection. It is likely that a combination of multiple predictors is needed to achieve such prediction. We used elastic net regularized regression to optimize prediction of symptom improvement and remission during treatment with escitalopram or nortriptyline and to identify contributing predictors from a range of demographic and clinical variables in 793 adults with major depressive disorder. A combination of demographic and clinical variables, with strong contributions from symptoms of depressed mood, reduced interest, decreased activity, indecisiveness, pessimism and anxiety significantly predicted treatment outcomes, explaining 5–10% of variance in symptom improvement with escitalopram. Similar combinations of variables predicted remission with area under the curve 0.72, explaining approximately 15% of variance (pseudo R2) in who achieves remission, with strong contributions from body mass index, appetite, interest-activity symptom dimension and anxious-somatizing depression subtype. Escitalopram-specific outcome prediction was more accurate than generic outcome prediction, and reached effect sizes that were near or above a previously established benchmark for clinical significance. Outcome prediction on the nortriptyline arm did not significantly differ from chance. These results suggest that easily obtained demographic and clinical variables can predict therapeutic response to escitalopram with clinically meaningful accuracy, suggesting a potential for individualized prescription of this antidepressant drug.",

keywords = "Depression, Outcome, Antidepressant, Prediction, Machine learning, Statistical learning",

author = "Raquel Iniesta and Karim Malki and Wolfgang Maier and Marcella Rietschel and Ole Mors and Joanna Hauser and Neven Henigsberg and Dernovsek, {Mojca Zvezdana} and Daniel Souery and Daniel Stahl and Richard Dobson and Aitchison, {Katherine J.} and Anne Farmer and Lewis, {Cathryn M.} and Peter McGuffin and Rudolf Uher",

year = "2016",

month = jul,

doi = "10.1016/j.jpsychires.2016.03.016",

language = "English",

volume = "78",

pages = "94--102",

journal = "Journal of psychiatric research",

issn = "0022-3956",

publisher = "Elsevier",

}

Iniesta, R , Malki, K, Maier, W, Rietschel, M, Mors, O, Hauser, J, Henigsberg, N, Dernovsek, MZ, Souery, D, Stahl, D , Dobson, R , Aitchison, KJ , Farmer, A , Lewis, CM , McGuffin, P & Uher, R 2016, 'Combining clinical variables to optimize prediction of antidepressant treatment outcomes', Journal of psychiatric research, vol. 78, pp. 94-102. https://doi.org/10.1016/j.jpsychires.2016.03.016

TY - JOUR

T1 - Combining clinical variables to optimize prediction of antidepressant treatment outcomes

AU - Iniesta, Raquel

AU - Malki, Karim

AU - Maier, Wolfgang

AU - Rietschel, Marcella

AU - Mors, Ole

AU - Hauser, Joanna

AU - Henigsberg, Neven

AU - Dernovsek, Mojca Zvezdana

AU - Souery, Daniel

AU - Stahl, Daniel

AU - Dobson, Richard

AU - Aitchison, Katherine J.

AU - Farmer, Anne

AU - Lewis, Cathryn M.

AU - McGuffin, Peter

AU - Uher, Rudolf

PY - 2016/7

Y1 - 2016/7

N2 - The outcome of treatment with antidepressants varies markedly across people with the same diagnosis. A clinically significant prediction of outcomes could spare the frustration of trial and error approach and improve the outcomes of major depressive disorder through individualized treatment selection. It is likely that a combination of multiple predictors is needed to achieve such prediction. We used elastic net regularized regression to optimize prediction of symptom improvement and remission during treatment with escitalopram or nortriptyline and to identify contributing predictors from a range of demographic and clinical variables in 793 adults with major depressive disorder. A combination of demographic and clinical variables, with strong contributions from symptoms of depressed mood, reduced interest, decreased activity, indecisiveness, pessimism and anxiety significantly predicted treatment outcomes, explaining 5–10% of variance in symptom improvement with escitalopram. Similar combinations of variables predicted remission with area under the curve 0.72, explaining approximately 15% of variance (pseudo R2) in who achieves remission, with strong contributions from body mass index, appetite, interest-activity symptom dimension and anxious-somatizing depression subtype. Escitalopram-specific outcome prediction was more accurate than generic outcome prediction, and reached effect sizes that were near or above a previously established benchmark for clinical significance. Outcome prediction on the nortriptyline arm did not significantly differ from chance. These results suggest that easily obtained demographic and clinical variables can predict therapeutic response to escitalopram with clinically meaningful accuracy, suggesting a potential for individualized prescription of this antidepressant drug.

AB - The outcome of treatment with antidepressants varies markedly across people with the same diagnosis. A clinically significant prediction of outcomes could spare the frustration of trial and error approach and improve the outcomes of major depressive disorder through individualized treatment selection. It is likely that a combination of multiple predictors is needed to achieve such prediction. We used elastic net regularized regression to optimize prediction of symptom improvement and remission during treatment with escitalopram or nortriptyline and to identify contributing predictors from a range of demographic and clinical variables in 793 adults with major depressive disorder. A combination of demographic and clinical variables, with strong contributions from symptoms of depressed mood, reduced interest, decreased activity, indecisiveness, pessimism and anxiety significantly predicted treatment outcomes, explaining 5–10% of variance in symptom improvement with escitalopram. Similar combinations of variables predicted remission with area under the curve 0.72, explaining approximately 15% of variance (pseudo R2) in who achieves remission, with strong contributions from body mass index, appetite, interest-activity symptom dimension and anxious-somatizing depression subtype. Escitalopram-specific outcome prediction was more accurate than generic outcome prediction, and reached effect sizes that were near or above a previously established benchmark for clinical significance. Outcome prediction on the nortriptyline arm did not significantly differ from chance. These results suggest that easily obtained demographic and clinical variables can predict therapeutic response to escitalopram with clinically meaningful accuracy, suggesting a potential for individualized prescription of this antidepressant drug.

KW - Depression

KW - Outcome

KW - Antidepressant

KW - Prediction

KW - Machine learning

KW - Statistical learning

U2 - 10.1016/j.jpsychires.2016.03.016

DO - 10.1016/j.jpsychires.2016.03.016

M3 - Article

SN - 0022-3956

VL - 78

SP - 94

EP - 102

JO - Journal of psychiatric research

JF - Journal of psychiatric research

ER -

Combining clinical variables to optimize prediction of antidepressant treatment outcomes

Abstract

Keywords

Access to Document

Fingerprint

Cite this