| Received: April 15 2023; Accepted: April 15 2023 |
We read Finsterer & Scorza's letter1 regarding our paper on Long COVID in the Brazilian Journal of Psychiatry2 with great interest.
There is now evidence that Long COVID may exacerbate preexisting type 2 diabetes and cause the onset of insulin resistance (IR).2 In our paper, IR was estimated using the homeostasis model assessment 2 (HOMA2IR) index, which is an adequate measure of IR as assessed with the glycemic clamp.2 We found that HOMA2IR scores were associated with increased severity of depression due to Long COVID.
Finsterer & Scorza raise concerns that this association may not be causal, as it has not been previously reported. If this were a criterion, it would never be possible to make any new discoveries. Secondly, they state that the sample size for the study was small, although the study is well powered. We recently confirmed in a study cohort of patients with COVID-19 that 3-6 months after infection, patients with Long COVID have higher HOMA2IR scores than those without Long COVID.3
According to the authors, a third argument against causality is the absence of alternative causes. This is yet another peculiar remark, given that the study identified additional "causes," such as inflammatory mediators. Furthermore, in reference 2 the HOMA2IR score contributed to the phenome of Long COVID in addition to the effects of inflammatory mediators of Long COVID (CRP and tryptophan catabolites) and the acute infectious phase (peak body temperature and lowered SpO2).3
According to Finsterer & Scorza, the fourth argument is that there would be no information on other Long COVID symptoms that "have contributed to depression." This is incomprehensible because we2 reported that all symptoms of Long COVID, including depression, anxiety, chronic fatigue, cognitive disorders, and physio-somatic symptoms (including gastrointestinal symptoms), belong to one factor, indicating that all these symptoms are manifestations of the "physio-affective phenome" of Long COVID. In addition, we3 reported that the HOMA2IR score contributes to the Long COVID physio-affective phenome. Previous papers have frequently shown evidence that IR plays a role in the pathophysiology of major depression.2-6 The HOMA2IR score is associated with type 2 diabetes mellitus-related depression and anxiety,4 and in unstable angina, the HOMA2IR score accurately predicts the physio-affective phenome.5 In severe depression, HOMA2IR contributes to the physio-affective phenome and these effects are partially mediated by neuronal and astroglial projection biomarkers.6
The authors then state that 50% of our patients had metabolic syndrome, concluding that IR is not associated with Long COVID but rather with metabolic syndrome. This argument is completely flawed since the aforementioned percentage is only an estimate based on the results of HOMA2IR.2 We do not believe that measuring C peptide, HbA1C, insulin antibodies, or amylases would alter any of the observed associations between HOMA2IR and the Long COVID phenome. The authors appear to imply that an autopsy would be fascinating, but we are not sure.
We have no interest in how the condition is labelled, whether as Long COVID or any other term. All of these binary diagnoses are man-made constructs, based on symptoms and illness duration that have not been statistically validated as constructs. Instead of using non-validated binary diagnoses, researchers should employ our validated quantitative, physio-affective phenome scores, which are calculated using our precision nomothetic medicine approach.2-6 Based on prior knowledge in depression and other IR-related conditions, our findings suggest that IR has a causal effect on depression due to Long COVID via amplification of neurotoxic pathways.2

PDF
Print
Send this article by email
How to cite this article
Submit a comment
Mendeley
Pocket




