Is there a causal link between mental health problems and risk of COVID-19 infection?


The COVID-19 pandemic is having marked effects on mental health, with high levels of depression, anxiety and mental distress. Studies initiated during the lockdown in early 2020 show elevated levels of symptomatology, and these findings have been corroborated with longitudinal studies that have been able to make comparisons with mental health before the COVID-19 crisis (Daly, Sutin, & Robinson, 2020; Pierce et al., 2020).

People with existing mental heath conditions, ethnic minorities and young adults seem particularly susceptible (Iob, Frank, Steptoe, & Fancourt, 2020). However, this research has been concerned with our encounters with a frightening, widespread and dangerous infection, and with the experience of lockdown and social isolation. It has not focused on COVID-19 infection itself.

Now two new studies have investigated the possibility that people with mental illness are at increased risk of infection with COVID-19, and in one case that infection may stimulate the development of mental illness among people with no previous psychiatric history.

These issues are difficult to study because although this infection has led to widespread illness and death across the world, the prevalence of COVID-19 is quite low statistically, making it difficult to evaluate associations with mental health except in very large study samples. Wang et al. (2020) and Taquet at al. (2020) both used electronic health record (EHR) datasets that included millions of people, allowing them to evaluate links on a very large scale. This brings both advantages and limitations.

The COVID-19 pandemic has led to increases in emotional distress throughout the population, but people with existing mental health problems are particularly vulnerable.

The COVID-19 pandemic has led to increases in emotional distress throughout the population, but people with existing mental health problems are particularly vulnerable.


Both studies were carried out with electronic health records in the USA. Wang et al analysed an EHR database of around 61 million adults, while the Taquet study involved a dataset derived from 69.8 million people, so between them they assessed a substantial proportion of adults in the USA. These anonymised databases are populated from a variety of sources including hospitals, primary care and health care organisations, and include basic demography along with the codes for health conditions and their timing. Both studies had a case-control design:

  • Wang et al. compared COVID-19 infection rates among people with and without a recent psychiatric diagnosis (within the last year), adjusting statistically for age, sex, ethnicity, and eight comorbid conditions such as diabetes, obesity and cancer.
  • Taquet et al. used more complex methods involving propensity scores to compare people with and without a recent psychiatric diagnosis, matched on a slightly wider range of factors.


The Wang et al. study analysed just over 1.3 million patients who had a recent psychiatric diagnosis, and there were 15,110 cases of COVID-19 up to the end of July 2020.

  • The analysis showed a markedly higher risk of COVID-19 infection among people with recent psychiatric illness, with a more than 7-fold increase in risk for people with depression and schizophrenia diagnoses (adjusted odds ratios of 7.64 and 7.34).
  • Diagnoses of bipolar disorder and attention-deficit/hyperactivity disorder (ADHD) also conferred increased risk.
  • African Americans with a recent psychiatric diagnosis had a greater risk of COVID-19 infection than Caucasian people, and women were at higher risk than men.
  • There was also evidence of greater hospitalisation rates and mortality among COVID-19 cases with a previous psychiatric history.

The findings reported by Taquet et al. are less dramatic but show a similar pattern. They compared 1,729,837 individuals with a psychiatric diagnosis in the previous year with a matched sample of the same size with no psychiatric history.

  • The relative risk of COVID-19 infection was 1.65, indicating a 65% increase in risk for people with a recent psychiatric history after taking comorbidities, sex and ethnicity into account.
  • Older people (aged over 75) were at higher risk, but unlike the Wang study there were no differences between men and women, and no clear differences between different psychiatric diagnoses.

The second aim of the research by Taquet et al. was to test whether COVID-19 infection increases risk of psychiatric disorder over subsequent weeks. This analysis compared people who had documented COVID-19 with a series of matched cohorts of individuals with other health problems such as infection with influenza, other upper respiratory tract infections, and skin infections.

  • In each case, the likelihood of being diagnosed for the first time with a psychiatric condition over the next 90 days was greater among people who suffered from COVID-19.
  • The relative risks were around 2.0, varying with the precise comparison group being analysed.
  • The greatest risks were for diagnoses of anxiety disorder, insomnia, dementia, and mood disorders, with only a very small increase in diagnosis of psychotic disorders.
These studies (Wang et al., 2020; Taquet at al., 2020) suggest that people with diagnosed mental health problems are at increased risk of COVID-19 infection.

These studies (Wang et al., 2020; Taquet at al., 2020) suggest that people with diagnosed mental health problems are at increased risk of COVID-19 infection.


These two studies point to the possibility that poor mental health may be a risk factor for COVID-19 infection; additionally, Taquet et al indicate that falling victim to the virus may increase risk of future psychiatric problems. If it is genuinely the case that poor mental health increases risk for COVID-19 infection, what might the mechanisms be? There are several candidates:

Exposure to infection

People with mental health problems may be exposed to a greater extent to people with COVID-19, or to higher doses of virus, than are those without psychiatric conditions. This would lead to a greater likelihood of becoming infected. Issues such as living in crowded conditions, difficulty with social distancing, problems with understanding the health advice, or reduced motivation to protect oneself through hand-washing and other preventive actions, could all be relevant.

Recognition of symptoms

Many COVID-19 symptoms are nonspecific, so might not be recognised as originating with the virus, particularly by people with mental health problems who are prone to multiple physical symptoms. This could lead to delays in recognition of infection, perhaps allowing it to develop more vigorously.

Lifestyle factors

There is no strong evidence to date that physical activity or alcohol have effects on COVID-19 risk. But obesity, substance use and smoking can enhance the severity of COVID-19 illness, so could increase the likelihood of hospitalisation (Popkin et al., 2020; Wang, Kaelber, Xu, & Volkow, 2020). These lifestyle factors are all more common among people with psychiatric disorders (Gold et al., 2020), so could link poor mental health with poorer outcomes after COVID-19 infection.

Biological mediators

An intriguing possibility is that biological processes such as systemic inflammation link poor mental health with COVID-19 infection. Several psychiatric conditions such as depression, bipolar disorder and schizophrenia are related to heightened inflammation, and inflammation could have been further elevated by the stress of lockdown and social distancing in psychiatric patients. It is possible that inflammation is one mechanism through which obesity increases vulnerability to COVID-19 infection, and there is strong evidence that inflammation exacerbates disease progression and is associated with risk of admission to intensive care and of death (Khinda et al., 2020).

Health care issues

The role of health care in these links is difficult to evaluate. One simple possibility is that people with mental health problems are more likely than others to attend health care facilities where they could come into contact with a range of fellow patients, and with health care staff who are themselves potentially infectious. It is also possible that mental health care staff lack training in identifying and managing COVID-19 infections, leading to delays in care.

COVID-19 infection and future mental health problems

The mechanisms outlined here have focused on possible links between psychiatric illness and future COVID-19 risk. But some would equally apply to the reverse relationship described by Taquet et al, between COVID-19 infection and future psychiatric problems. For example, unhealthy lifestyle factors could increase risk of mental health problems, and the inflammation elicited by COVID-19 may affect neural circuits contributing to depression or bipolar disorder. Changes in health care following COVID-19 infection might lead to more careful surveillance of symptoms, some of which might be recognised as signs of psychiatric disturbance, thereby increasing the likelihood of formal diagnoses. With the growing understanding of long COVID syndrome, it is becoming clear that a sizeable proportion of patients’ experience ‘brain fog’, fatigue and persistent tiredness, and these could form the substrate for mood disorders in susceptible individuals.

The mechanisms for poor mental health increasing our risk for COVID-19 infection could be extremely varied; from lifestyle factors to biological mediators.

The mechanisms for poor mental health increasing our risk for COVID-19 infection could be extremely varied; from lifestyle factors to biological mediators.

Strengths and limitations

The authors of both papers are cautious about the interpretation of their results, pointing out the limitations of electronic health records (EHR). These are based on the logging of the health events in registries accumulated through inputs from multiple healthcare facilities. Much COVID-19 infection took place before widespread testing was in place and asymptomatic infections may not have been registered, so the amount of infection with the virus may have been underestimated. Importantly, EHRs include very limited sociodemographic and behavioural data, so important confounders will not have been taken into account. There are no strong measures of socioeconomic status such as education, income or occupation, and factors such as smoking were not assessed. Both are observational studies, so cannot establish causality.

These two studies used quite similar designs to address the notion that poor mental health increases risk for COVID-19. They both involved very large health record databases that provide precise information about diagnoses and their timing. They were both able to take into account many of the comorbid health problems that are known to increase susceptibility, such as diabetes and chronic obstructive pulmonary disease. A notable feature is that although both demonstrate that a recent psychiatric diagnosis is associated with increased risk of COVID-19 infection, the magnitude of the relationships was very different. A 7-fold increase in risk was reported by Wang et al, while the risk observed by Taquet et al, was less than 2-fold. The explanation for the difference probably lies in the statistical methods adopted, with a more exact matching of cases with controls in the latter study. A 7-fold association is very unusual in disease epidemiology, and raises suspicions about confounding. There could be unmeasured confounding factors that relate to risk both of mental health problems and COVID-19 infection. For example, lower socioeconomic position is associated with increased incidence of many psychiatric diagnoses while also being relevant to COVID-19 infection risk, and to death following infection (ONS, 2020). Smaller and more crowded housing conditions, more use of public transport, and work in frontline jobs that bring more contact with others will all increase exposure, and are characteristic of lower socioeconomic status groups (Anderson, Frank, Naylor, Wodchis, & Feng, 2020). Unfortunately, neither study was able to incorporate these factors into their analyses in a comprehensive way.

Lower socioeconomic status increases risk of COVID-19, and was not fully taken into account in these studies

Lower socioeconomic status increases risk of COVID-19, and was not fully taken into account in these studies.

Implications for practice

These two studies have thrown light on an important issue in the COVID-19 landscape. The focus of most research on mental health and wellbeing has been on the impact of social distancing, lockdown and lack of physical contact in the population in general. Important though this work is, these studies indicate that links between mental health and COVID-19 infection itself should be studied more seriously. We need to understand how important these relationships are, and provide estimates of the size of associations to be expected. Poor mental health may join the growing list of factors that increase risk of infection. The care of people who are recovering from COVID-19 may also need to incorporate more precise monitoring of mental wellbeing so that serious psychiatric problems can be identified at the earliest stage. However, longitudinal studies that take account of a wider range of covariates are needed before firm policy implications can be drawn.

Mental health needs to be taken into account when evaluating risk of COVID-19 infection.

Mental health needs to be taken into account when evaluating risk of COVID-19 infection.

Conflicts of interest

None declared.


Primary papers

Taquet M, Luciano S, Geddes JR, et al. Bidirectional associations between COVID-19 and psychiatric disorder: retrospective cohort studies of 62,354 COVID-19 cases in the USA. Lancet Psychiatry. 2020. doi:10.1016/S2215-0366(20)30462-4

Wang Q, Xu R, Volkow ND. Increased risk of COVID-19 infection and mortality in people with mental disorders: analysis from electronic health records in the United States. World Psychiatry. 2020. doi:10.1002/wps.20806

Other references

Anderson, G., Frank, J. W., Naylor, C. D., Wodchis, W., & Feng, P. (2020). Using socioeconomics to counter health disparities arising from the covid-19 pandemic. BMJ, 369, m2149. doi:10.1136/bmj.m2149

Daly, M., Sutin, A., & Robinson, E. (2020). Longitudinal changes in mental health and the COVID-19 pandemic: Evidence from the UK Household Longitudinal Study. Psychol Med, 1-37. doi:10.1017/S0033291720004432

Gold, S. M., Kohler-Forsberg, O., Moss-Morris, R., Mehnert, A., Miranda, J. J., Bullinger, M., . . . Otte, C. (2020). Comorbid depression in medical diseases. Nat Rev Dis Primers, 6(1), 69. doi:10.1038/s41572-020-0200-2

Iob, E., Frank, P., Steptoe, A., & Fancourt, D. (2020). Levels of severity of depressive symptoms among at-risk groups in the UK during the COVID-19 pandemic. JAMA Netw Open, 3(10), e2026064. doi:10.1001/jamanetworkopen.2020.26064

Khinda, J., Janjua, N. Z., Cheng, S., van den Heuvel, E. R., Bhatti, P., & Darvishian, M. (2020). Association between markers of immune response at hospital admission and COVID-19 disease severity and mortality: A meta-analysis and meta-regression. J Med Virol. doi:10.1002/jmv.26411

Office for National Statistics (2020). Deaths involving COVID-19 by local area and socioeconomic deprivation: deaths occurring between 1 March and 31 July 2020.

Pierce, M., Hope, H., Ford, T., Hatch, S., Hotopf, M., John, A., . . . Abel, K. M. (2020). Mental health before and during the COVID-19 pandemic: a longitudinal probability sample survey of the UK population. Lancet Psychiatry, 7(10), 883-892. doi:10.1016/S2215-0366(20)30308-4

Popkin, B. M., Du, S., Green, W. D., Beck, M. A., Algaith, T., Herbst, C. H., . . . Shekar, M. (2020). Individuals with obesity and COVID-19: A global perspective on the epidemiology and biological relationships. Obes Rev, 21(11), e13128. doi:10.1111/obr.13128

Wang, Q. Q., Kaelber, D. C., Xu, R., & Volkow, N. D. (2020). COVID-19 risk and outcomes in patients with substance use disorders: analyses from electronic health records in the United States. Mol Psychiatry. doi:10.1038/s41380-020-00880-7

Photo credits

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+