Antipsychotics modified COVID-19 prevalence in hospitalized patients diagnosed with mental illnesses

After the Coronavirus disease 2019 (COVID-19) pandemic began three years ago, we now know that Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection is not only characterized by respiratory and gastrointestinal complications. Several studies have found that about one-third of patients suffer neurological or psychiatric symptoms, which are even prevalent in the so-called "Long-COVID." (there is ambiguity about its definition, however here it will be considered as the presence of symptoms related to COVID-19 three weeks after infection [1]). This can be explained by the tropism of SARS-CoV-2 for brain cells [2]. Therefore, it can be considered that patients diagnosed with mental illness are vulnerable to COVID-19, and this assertion is supported by the high prevalence, presence of severe symptoms, and mortality described in this population. However, this susceptibility depends on multiple factors, including the specific psychiatric diagnosis and pharmacotherapy, including antipsychotics [3]. Our working group has previously documented variations in COVID-19 prevalence due to specific diagnoses in 135 hospitalized patients diagnosed with mental illnesses, but we were unable to study the effects of pharmacotherapy [4].

Antipsychotics have anti-inflammatory, antioxidant, and neurotrophic properties beyond their canonical monoaminergic mechanisms, suggesting a potential protective effect of these drugs against COVID-19 [2]. This hypothesis is settled on various preliminary studies carried out in hospitalized patients with a mental illness diagnosis, in which a lower prevalence of COVID-19 severe symptoms was detected in inpatients under atypical or second-generation antipsychotic (SGA) treatment [[5], [6], [7]]. But also some investigations have failed to replicate these findings [8]. With this background, we aimed to investigate how COVID-19 prevalence in hospitalized patients with mental illness was affected by exposure to various antipsychotic classes, including typical or first-generation (FGA), SGA, or a combination of both (methods are described in detail in Supplementary material 1 and information on the hospital population can be found in Supplementary material 2). In the general hospital setting, it was found that inpatients exposed to SGA had a higher prevalence of COVID-19 (Odds ratio (OR) 2.11, 95% CI: 1.13–3.86, P = 0.0207; Fig. 1 ), and this effect was only maintained in inpatients aged 18–65 years (OR 2.38, 95% CI: 1.2–4.92, P = 0.0156; Fig. 1) and inpatients diagnosed with intellectual disabilities (OR 2.87, 95% CI: 1.3–6.19, P = 0.0108; Fig. 1, for detailed information see Supplementary material 3). Thus, it can be suggested that the protective effect of SGA against COVID-19 may depend on multiple factors such as treatment adherence and the specific type of SGA as previously detected [5,6,8], and our results add age and specific diagnosis as factors to consider.

Fig. 1.

COVID-19 prevalence in inpatients exposed to distinct types of antipsychotics (AP), including first-generation antipsychotics (FGA), second-generation antipsychotics (SGA), first- and second-generation antipsychotics (F&SGA), and without antipsychotic (No AP) treatment. In inpatients of the general hospital setting, aged 18–65 years, and diagnosed with intellectual disability (ID) exposed to SGA, increased COVID-19 prevalence was detected. In contrast, in inpatients exposed to FGA aged 18–65 years and diagnosed with ID, decreased COVID-19 prevalence was detected. Bold values indicate statistical significance.

Interestingly, in inpatients exposed to FGA a downward trend in the prevalence of COVID-19 was detected (OR 0.49, 95% CI: 0.24–1.0, P = 0.0756; Fig. 1), which was evident in inpatients aged 18–65 years (OR 0.37, 95% CI: 0.17–0.81, P = 0.0202; Fig. 1) and diagnosed with intellectual disabilities (OR 0.23, 95% CI: 0.16–0.55, P = 0.0039; Fig. 1). Antiviral properties of FGA have been documented since the 1980s. Haloperidol and phenothiazine derivates disrupt Corona viruses infection through the inhibition of cell-cell fusion and clathrin-mediated endocytosis [9]. Also, FGA are capable of acting against Spike protein (S)/angiotensin-converting enzyme (ACE) interaction for SARS-CoV-2 cell infection [10]. These FGA properties may explain the observed reduction in COVID-19 prevalence. The present results are congruent with a study carried out in the French population in which the FGA chlorpromazine (a phenothiazine derivate) reduced the prevalence of severe COVID-19 symptoms [11]. All of this information reinforces the hypothesis of the protective role of FGA for COVID-19 outcomes [12], which also can be beyond the short-term ones, since nowadays there is myriad information about the long-term consequences of SARS-CoV-2 infection known as Long COVID.

Regarding Long COVID prevalence, it was detected in 10.39% of the general hospital setting (Supplementary material 1), which is lower than the estimations in the general population (∼50%) [13]. This phenomenon may be attributed to antipsychotics because of their anti-inflammatory and antiviral mechanisms, but further studies must explore this hypothesis.

The present results suggest that COVID-19 prevalence in hospitalized patients with mental illness diagnosis is modified by the distinct type of antipsychotic according to specific factors: the FGA reduced the prevalence of the disease in inpatients of 18–65 years of age and inpatients diagnosed with intellectual disabilities, while the SGA increased COVID-19 prevalence in the general hospital population, disease in inpatients 18–65 years of age and in inpatients diagnosed with intellectual disabilities. Also, reduced Long COVID prevalence, in comparison with the general population, was detected in these inpatients. This reported data support the hypothesis of the protective role of FGA for COVID-19 outcomes, including long-term ones.

Although preliminary, our findings considerably add to the corpus of knowledge about how antipsychotics alter COVID-19 outcomes not only because of their antiviral properties, but also due to their antioxidant and anti-inflammatory properties, which may also have effects on the mental illnesses' pathophysiological mechanisms [2]. Even though the comparison of FGA and SGA effects on COVID-19 prevalence is the study's greatest strength, it is crucial to keep in mind that the study did not explore the time and dose of the antipsychotic treatment. Also, the research was only conducted on a small group of inpatients, and even smaller subgroups with specific diagnoses, in a healthcare facility under closely controlled and supervised sanitary conditions. To further explore the possibilities examined here, thorough investigations must be conducted in the future.

Funding

This study was supported by CONACYT grant (252808) for G.F. H.T.-B. is funded by CONACYT's "Estancias Posdoctorales por México" program (662350). None of the funding organizations played any other part in the planning of the study, data collecting, analysis, interpretation, report writing, or the choice to submit the work for publication.

Conflict of interest

None.

Appendix A. Supplementary data

Supplementary material 1

Supplementary material 2

Supplementary material 3

Data availability

The data that support the findings of this study are available from the corresponding authors, upon reasonable request.