SAFETYLIT WEEKLY UPDATE

We compile citations and summaries of about 400 new articles every week.
RSS Feed

HELP: Tutorials | FAQ
CONTACT US: Contact info

Search Results

Journal Article

Citation

Steinberg JR, Laursen TM, Adler NE, Munk-Olsen T. Lancet Psychiatry 2020; 7(3): 237.

Affiliation

Aarhus Universitet, Aarhus, Denmark.

Copyright

(Copyright © 2020, Elsevier Publishing)

DOI

10.1016/S2215-0366(20)30031-6

PMID

32087805

Abstract

In response to Hirofumi Hirakawa and Nobuyoshi Ishii's Correspondence,1 we are happy to clarify what we controlled for in our Article. We did control for diagnoses of bipolar disorder in the covariate of previous psychiatric contacts; this variable turned out to be the strongest risk factor for non-fatal suicide attempts. Specifically, the covariate of previous psychiatric contacts coded whether women had an ICD-10 mental health diagnosis given by mental health professionals; it included bipolar disorder, which has an ICD-10 code of F31.X.
We did not, however, control for antiepileptic (or anticonvulsant) drugs that are sometimes used to treat bipolar disorder, although the efficacy of these drugs to treat bipolar disorder has been questioned.2 Adding to this issue, we argue that we do not expect that our main findings would have changed if there had been additional adjustment for use of anticonvulsants. That is, even controlling for anticonvulsants, there would likely have been only minor and limited change in the association between having an abortion and non-fatal suicide attempts, and there would very likely not be a different risk of suicide attempts in the year before and the year after the abortion, relative to women who had no abortion. Thus, controlling for anticonvulsants would not influence our main conclusion that the increased risk of non-fatal suicide attempts in women having abortions could not be due to the abortion. We base this interpretation on our findings that there was no change in relative risk of non-fatal suicide attempts from the year before an abortion to the year after an abortion, both in the unadjusted model or when we controlled for, first, age and calendar year or, second, age, calendar year, previous psychiatric contacts, previous antidepressant medication use, previous anti-anxiety medication use, previous antipsychotic medication use, previous parental psychiatric contacts, previous physical health, previous childbirth, and parents' education when the individual was aged 15 years.


Language: en

NEW SEARCH


All SafetyLit records are available for automatic download to Zotero & Mendeley
Print