As the first step, we examined the genetic correlations between the three psychosocial factors and COVID-19 outcomes and found that each of the three psychosocial factors was negatively correlated with each outcome of COVID-19 (Table 2). Then, we evaluated the overall causal effects of EA, intelligence, and income on COVID-19 outcomes and found that they could protect against SARS-CoV-2 infection, hospitalized COVID-19, and critical COVID-19 (Table 3 and Fig. 1). Finally, we tested their independent effects on COVID-19 outcomes and found that only EA could protect against all COVID-19 outcomes independently. Intelligence protected against SARS-CoV-2 infection but not against severe forms of COVID-19, while higher income was found to increase the risks of SARS-CoV-2 infection independently of two other socioeconomic factors (Table 4 and Fig. 1).
To date, three MR studies have investigated the relationships between psychosocial factors and COVID-19. The first MR study reported that higher EA confers a lower risk of severe COVID-19 [20]. The second study revealed that both EA and intelligence exerted causal effects on hospitalized and critical COVID-19 outcomes in either MR or MVMR analysis but did not evaluate the effects of income [22]. The third study, based on MVMR, found that EA exerts protective effects on SARS-CoV-2 infection as well as on hospitalized and critical COVID-19 patients but did not report the effects of intelligence and income on any COVID-19 outcomes [21]. Our study differed from previous studies described above in employing the most recent GWAS datasets and in testing the overall and independent effects of EA, intelligence, and income on each outcome of COVID-19.
When the independent effects of each psychosocial factor were evaluated, some surprising discoveries were made. In particular, we found that genetic predisposition to higher intelligence might reduce the risk of SARS-CoV-2 infection but not severe COVID-19. People with high intelligence may be better able to understand defensive measures, which may translate into a more efficient assessment of personal risks and, as a result, lessening one's chances of being infected. On the other hand, recent evidence suggests that COVID-19 may exert a detrimental effect on memory and intelligence [32–34].
We confirmed previous findings suggesting that the genetic underpinning of high EA contributed to reducing exposure to COVID-19 and the propensity to develop its severe forms. People with higher education may exercise better long-term strategies for the management of their health by maintaining a risk-reducing lifestyle [35–37]. Even if infected, a better baseline physique may prevent significant deterioration of health. A previous observational study found that higher EA was associated with a lower propensity to smoke, to be inactive or obese, or to have hypertension, hypercholesterolemia, hyperglycemia, and high BMI [38, 39]. It was found that comorbidity with other somatic diseases, such as cancer, cardiovascular diseases, cerebrovascular diseases, or diabetes, was associated with increased severity and mortality of COVID-19 [40–43]. In accordance with these observations, we found that intelligence itself cannot protect against the severity of COVID-19, but higher EA may decrease the severity of SARS-CoV-2 disease outcomes independent of intelligence.
Remarkably, the protective effects of income on SARS-CoV-2 infection were reversed in the MVMR analysis. Higher income was associated with an increased risk for SARS-CoV-2 infection, and its protective effects on hospitalization disappeared. Our study suggested that the protective effects of higher income on COVID-19 were mainly mediated by education. A combination of high income with low EA might be counterproductive for efforts against infections for a couple of reasons. First, the high possibility of infection in high-income populations might be due to their higher mobility in the early stage of the pandemic, when they retained the ability to travel between different countries or areas, thus increasing the risks of contracting the virus. Second, high-income populations might have greater access to molecular diagnostics, which increased COVID-19 positivity rates in these groups [44].
During the COVID-19 pandemic, higher rates of infection and death from COVID-19 in low-income or middle-income populations have been reported repeatedly [45]. A study of COVID-19 inpatients in UK hospitals showed that larger social deprivation scores were associated with a higher risk of death [46]. In Sweden, being a man, having a lower personal income, having lower education, and being single increased the risk of death from COVID-19 [47]. The higher rates of hospitalization, severity, and mortality of COVID-19 in the populations of low socioeconomic status may be due to poor housing, physically exhausting jobs, worse hygiene, low food quality, more substance misuse, less access to healthcare systems, and delayed contact with the health care provider, along with difficulties in affording the expense of medical care [15, 48]. Accordingly, in developed countries, the mortality and morbidity of SARS-CoV-2 were significantly higher in ethnic minorities than in Caucasians [49]. Most ethnic minorities were of lower socioeconomic status, and their education levels and income levels were not high.
This study was not free of limitations. First, the current MR analysis employed the summary statistics of GWAS meta-analyses conducted among Europeans, indicating that the causality inferred from these datasets might apply to Europeans only. Second, we only analyzed EA, intelligence, income, and COVID-19, while other sociodemographic factors and clinical parameters were not examined. Third, we realized that the outcomes of SARS-CoV-2 infection depended on the individuals’ overall situations, which included sociodemographic characteristics, comorbidities, immune status, and anthropometrics rather than being solely due to genetics.
To summarize, our study affirmed that EA had a beneficial impact on reducing vulnerability to SARS-CoV-2 and lessening the severity of COVID-19. Importantly, these effects were not influenced by intelligence and income. Nevertheless, our study did not find evidence supporting the individual effects of intelligence or income on the severity of COVID-19.