Miriam E. Tucker
July 09, 2019
There was an approximate 35% reduction in deaths over the study period for all antidepressant classes except for reversible inhibitors of monoamine oxidase A (RIMA).
"Since the incidence of major depressive disorder among individuals with diabetes is significantly greater than the general population, and diabetes and depression each independently contribute to elevated total mortality, we suggest clinicians need to screen for depression among patients with diabetes," lead researcher Vincent Chin-Hung Chen, MD, PhD, professor of medicine at Chang Gung Medical University and staff psychiatrist at Chiayi Chang Gung Hospital, Taiwan, told Medscape Medical News.
"Clinicians can cooperate with psychiatrists to help those patients, including prescription of antidepressants," Chen added.
However, he also cautioned that this is "an association study and does not represent a causal relationship. Further studies are warranted," to replicate the findings, especially in other countries or areas, he emphasized.
The results are from a retrospective analysis of over 50,000 individuals with comorbid diabetes and major depression or dysthymia from a nationwide Taiwanese database. The study was published online July 2 in the Journal of Clinical Endocrinology & Metabolism by Chen and colleagues.
Inverse Relationship Between Antidepressant Dose and Early Death
Using 2000-2013 data from the National Health Insurance Research Database in Taiwan, researchers identified 53,412 patients with newly diagnosed diabetes — the type of diabetes could not be ascertained from the database — and subsequently diagnosed depression (major depressive disorder or dysthymic disorder). Those with antidepressant use prior to the diagnosis of diabetes were excluded.
Of these, 50,532 people were taking antidepressants and 2880 were not.
Those taking antidepressants were divided into three dose subgroups based on cumulative defined daily dose (cDDD) or the assumed average maintenance dose per day for a drug used for its main indication in adults: < 28 DDDs (19.3% of patients), 28-84 DDDs (18.3%), and 84-364 DDDs (62.4%).
Over a 10-year follow-up (average 9.2-9.9 years), death rates ranged from 1113.7/100,000 person-years in the highest dose group to 1963.7/100,000 person-years in the lowest dose group (P < .001).
After adjustments, total mortality decreased as cumulative total dose increased. This became statistically significant at a cDDD of > 28 (hazard ratio [HR], 0.91), and the significance became greater at cDDD ≥ 84 (HR, 0.65).
When each of seven classes of antidepressants was analyzed separately, there was no significant effect on overall mortality with cDDD < 83.
When cDDD was ≥ 84, there were significant reductions in mortality for selective serotonin reuptake inhibitors (HR, 0.63), serotonin-norepinephrine reuptake inhibitors (HR, 0.58), norepinephrine-dopamine reuptake inhibitors (HR, 0.20), mirtazapine (HR, 0.60), tricyclic/tetracyclic antidepressants (HR, 0.73), and trazodone (HR, 0.52).
In contrast, RIMAs were linked to a significant increase in total mortality (HR, 1.48).
"In Taiwan, RIMA is usually used as the second-line drug for depression when most of the antidepressants do not work. The [higher] risk [for death] may be due to the indication bias to treat the treatment-resistant depressed patients with poor prognosis," Chen suggested to Medscape Medical News.
The mechanism for the mortality reduction with antidepressants is not known, but the authors hypothesize that it could be related to an inflammation modulation effect.
Summarizing, Chen and colleagues say: "To our knowledge, this is the first large population-based cohort study to identify an inverse association between antidepressant use and mortality among individuals diagnosed with DM and comorbid depression."
And they stress that the data provide "further rationale for the screening and treating of depression in persons who have [diabetes]."
The study was supported by grants from Changhua Christian Hospital, Taiwan. Chen has reported no relevant financial relationships.