A Population-Based Outcome-Wide Association Study of the Comorbidities and Sequelae Following COVID-19 Infection

Background Immense attention has been given to the outcome of COVID-19 infection. However, comprehensive studies based on large populational cohort with long-term follow-up are still lacking. This study aimed to investigate the risk of various short-term comorbidities (within one month) and long-term sequelae (above one month) after COVID-19 infection. Methods In this large prospective cohort study with 14 months follow-up information based on UK biobank, we included 16,776 COVID-19-positive participants and 58,281 COVID-19-negative participants matched for comparison. The risk of each comorbidity and sequela was evaluated by multivariable logistic regression analysis and presented as hazard ratio (HR) and 95% confidence interval (95% CI). Results COVID-19-positive individuals had a higher risk of 47 types of comorbidities within one month following COVID-19 infection, especially those who were older, male, overweight/obese, ever-smoked, with more pre-existing comorbidities and hospitalized. About 70.37% of COVID-19 patients with comorbidities had more than one co-occurring comorbidities. Additionally, only 6 high-risk sequelae were observed after one month of COVID-19 infection, and the incidence was relatively low (< 1%). Conclusion In addition to long-term sequelae following COVID-19 infection, plenty of comorbidities were observed, especially in patients with older age, male gender, overweight/obese, more pre-existing comorbidities and severe COVID-19, indicating that more attention should be given to these susceptible persons within this period. Supplementary Information The online version contains supplementary material available at 10.1007/s44197-023-00161-w.


Introduction
Over the past 3 years, corona virus disease 2019 (COVID- 19) has spread globally and has become a growing public health concern, posing a serious threat to human survival.As of November 2022, over 637 million cases of COVID-19 have been confirmed globally, of which about 1.04% have died [1].In the context of widespread infection, the possible systemic comorbidities and sequelae of COVID-19 may cause panic.
The aim of this study is to perform a comprehensive outcome-wide association analysis of COVID-19 comorbidities (short-term) and sequelae (long-term), and individual susceptibilities based on the large prospective UK Biobank (UKB) cohort.

Study Design and Data Source
This prospective population-based outcome-wide association study was conducted in the UKB, which is a national prospective cohort that recruited over 500,000 participants aged 40-69 years from 22 assessment centers across the United Kingdom between 2006 and 2010.UKB obtained ethical approval from the North West Multicenter Research Ethics Committee [17], and has collected extensive detailed baseline and long-term follow-up data, including real-time updated extensive clinical records.SARS-CoV-2 testing results were obtained from Public Health England (PHE), Public Health Scotland (PHS) and Secure Anonymized Information Linkage (SAIL).Hospitalization information of participants from England, Scotland, and Wales were acquired from Hospital Episode Statistics, the Scottish Morbidity Record, and the Patient Episode Database for Wales, respectively.Mortality data were extracted from the National Health Service (NHS) Digital and NHS Central Register.

Participants and Exposure
A total of 106,760 UKB participants who were tested for SARS-CoV-2 during March 2020 and February 2021 were included.Among the 17,832 participants with at least one positive SARS-CoV-2 test (COVID-19-positve participants), those who were tested with positive result after 2021-2-23 (had limited follow-up time, n = 1041), or withdrew during follow-up (n = 15) were excluded.We further defined the severity of COVID-19 according to the presence of death or hospitalization due to COVID-19.For these 88,928 participants without any positive SARS-CoV-2 test result (COVID-19-negative participants), we excluded those who were reported to die or hospitalize due to COVID-19 (n = 303), died before the pandemic of COVID-19 (2020-1-31, n = 1,095) or withdrew during follow-up (n = 14).Then for each COVID-19-positive participant, we matched up to 4 COVID-19-negative participants according to birth year, sex, and Townsend deprivation index (TDI) by propensity scores.Finally, we excluded COVID-19-negative participants who died before the SARS-CoV-2 test date of their matched COVID-19-positive counterparts.These inclusion and exclusion criteria, and matching process resulted in a final cohort of 16,776 COVID-19-positive participants and 58,281 COVID-19-negative participants (Supplementary Figs.1-2).

Outcome and Covariates
Comorbidities and sequelae, defined using 3-digit ICD-10 codes (International Classification of Diseases, 10th revision, excluding codes used for special purposes, injury, poisoning and certain other consequences of external causes, factors influencing health status and contact with health services, as well as external causes of morbidity and mortality) from medical records, referred to newly onset illnesses < 1 month and ≥ 1 month after the diagnosis of COVID-19, respectively [18][19][20].We then reclassified the eligible (hazard ratio [HR] > 1, p < 0.05, and case number > 10) diseases into more broadly defined comorbidities and sequelae (Supplementary Tables 1-2).In the case of multiple identical records for the same individual, the date of diagnosis was derived from the earliest record.Ethnicity, body mass index (BMI), smoking status, and Charlson comorbidity index (CCI, without age calculation, Supplementary Table 3) were included as covariates [21][22][23].In subgroup analysis, we also defined another covariate, current age, as the age on January 31st, 2020.

Follow-Up
The start date of follow-up was the test date for COVID-19positive participants.For COVID-19-negative participants, it was the same as that for their matched counterparts.Follow-up ended on the day of (1) specific disease diagnosis, (2) death, or (3) end of follow-up (March 31st, 2021), whichever came first.The longest follow-up duration was 14 months.

Statistics Analyses
Student's t-tests and Chi-square tests were performed as appropriate to assess the differences among groups.Univariable conditional Cox proportional hazards models were performed to identify eligible diseases for comorbidity and sequela reclassification, and outcome-wide association analyses for the risk of COVID-19-related comorbidities and sequelae were conducted using multivariable conditional Cox proportional hazards models adjusting for ethnicity, BMI, smoking status, and CCI [24].In the specific analysis of each disease, participants diagnosed with corresponding disease before COVID-19 were excluded.In the analysis of death in COVID-19-positive participants, we also excluded individuals who died on the day of diagnosis.To explore the impact of severity of COVID-19 on COVID-19-related sequelae, we additionally adjusted for age, sex, and TDI.In subgroup analyses, participants were stratified by current age (< 65 and ≥ 65 years), sex (female and male), BMI (< 25 kg/ m 2 and ≥ 25 kg/m 2 ), smoking status (ever-smoker and neversmoker), and CCI (≤ 1 and ≥ 2).
All analyses were performed using R software (version 3.6.3,https:// www.r-proje ct.org/), and a two-tailed p < 0.05 was considered statistically significant.

Subgroup Analyses of COVID-19-Related Comorbidities by Age, Sex, BMI, Smoking Status and CCI
To understand whether the burden of COVID-19-related comorbidities differed among various populations, we further carried out subgroup analyses stratified by current age, sex, BMI, smoking status and CCI, respectively (Fig. 2).

Discussion
In this prospective cohort study, we conducted comprehensive outcome-wide association analyses to identify COVID-19-related comorbidities and sequelae in a large population.Overall, 47 types of COVID-19 related comorbidities that occurred within one month after COVID-19 infection were identified, by incidence from high to low, including lower respiratory infection, respiratory failure, electrolyte imbalance, renal failure, hypertension and other heart disease, etc.We also observed that COVID-19-related comorbidities tended to co-occur, especially in severe COVID-19 patients.Besides, older age, male gender, obese/overweight, smoking history, higher CCI scores and severe COVID-19 were risk factors for experiencing more types of comorbidities.Meanwhile, we identified 6 types of COVID-19 related sequelae that began to appear after one month following COVID-19 infection, such as lower respiratory infection, immobility, interstitial pulmonary disease, fecal abnormalities, decubitus ulcer and urinary incontinence.Nonetheless, the incidence rates of these COVID-19-related sequelae were all relatively low (< 1%).Therefore, the public should be urged not to worry too much about these low-morbidities and low-seriousness sequelae.
Previous studies showed that COVID-19 patients had a higher risk of having sequelae, such as myalgia, sexual dysfunction, hearing loss and disturbances of smell and taste [25][26][27][28], however, these sequelae were not prominent in our results, as we only observed 6 types of COVID-19-related sequelae with low incidences.However, we observed plenty of comorbidities involving multiple organs after COVID-19 infection, such as respiratory, neurological, circulatory and urinary systems.Besides, the majority of COVID-19 patients developing comorbidities showed more than one comorbidity.Thus, comorbidity was a more prominent issue for COVID-19 patients.
It has been established that SARS-CoV-2 can upregulate the expression of the type 2 angiotensin converting enzyme (ACE-2), and can bind ACE-2 receptors on the surface of the host cells for cell entrance in many organs, which may explain the comorbidities we observed in COVID-19 patients, such as hypertension, diabetes and COPD [29][30][31].COVID-19 is associated with a high inflammatory burden and SARS-CoV-2 can affect the myocardium and cardiac biomarker level and lead to myocarditis and heart failure [32][33][34][35].Moreover, interleukin-mediated modulation of phosphokinases and phosphatases, NOD-like receptor thermal protein domain associated protein 3 (NLRP3) inflammasome-mediated inflammation and pathological accumulation   [36][37][38][39].In addition, the activation of the RAS, hemodynamic changes and secondary infection of the urinary tract following COVID-19 infection are associated with the comorbidities and sequelae of urinary system in COVID-19 patients [40][41][42].Hearing loss was found in the comorbidities of severe COVID-19 patients, which may be associated with brainstem dysfunction resulting from neuroinflammatory mechanisms.Cytokine storm after COVID-19 infection could damage the auditory glial cells and might play a role in hearing loss [43].
We observed that participants with advanced age, male sex, smoking status or excessive obesity were at higher risks of COVID-19-related comorbidities.Old and obese participants were usually characterized by more pre-existing comorbidities, weaker immune defense, and higher levels of proinflammatory cytokines, which may contribute to their more comorbidities [44,45].Besides, the discrepancy of COVID-19 related outcomes between male and female participants could be attributed to the differences in sex hormones, expression levels of ACE2 and Transmembrane protease serine 2 (TMPRSS2), and lifestyles [46].Moreover, smoking is associated with a higher expression level of ACE2 in airway epithelial cells, which may induce the occurrence of COVID-19 related comorbidities and sequelae [47].These populations should be paid special attention as they were more susceptible to COVID-19-related comorbidities.
Although the advent of our study provided new insight into the comorbidities and sequelae of COVID-19 patients, a few limitations still existed.First, due to the data limitation, we defined the severity of COVID-19 according to the hospitalization status or death cause, which may lead to partial bias but was acceptable [48].Second, due to the limited data on COVID-19 medications, we could not assess the effect of COVID-19 medications on COVID-19-related outcomes.Third, our observations were mainly of comorbidities and sequelae associated with the SARS-CoV-2 Alpha variant, which was the main strain in the UK between January 31, 2020 and March 31, 2021, but not the Delta and Omicron variants of SARS-CoV-2, which began to emerged and spread in the UK from March and October 2021, respectively.The SARS-CoV-2 Alpha variant was considered relatively more pathogenic but less infectious than Delta and Omicron variant, therefore, the incidence and severity of sequelae of Delta and Omicron variants might be lower than, also be different from, those of Alpha variants [49][50][51][52].Unfortunately, to date, we are unable to obtain the lagging comorbidities and sequelae data of Delta and Omicron variant for analysis.
In conclusion, 47 types of high-risk comorbidities might occur within one month after COVID-19 infection, especially in patients with older age, overweight/obese, more pre-existing comorbidities and severe COVID-19.And only 6 types of COVID-19-related sequelae appeared after one month following COVID-19 infection, indicating that more attention and health care should be given to these susceptible populations after COVID-19 infection.use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.To view a copy of this licence, visit http:// creat iveco mmons.org/ licen ses/ by/4.0/.

Fig. 2
Fig. 2 Differences in adjusted hazards ratio of clinical comorbidities among COVID-19-positive participants compared with COVID-19-negative participants stratified by age, sex, BMI, smoking status, and CCI, respectively.Note: Only the top 20 comorbidities in terms

Table 1
The incidence and hazard ratio of comorbidity in COVID-19 individuals and matched negative comparisons, adjusted for ethnicity, BMI, smoking status, and CCI

Table 2
Clinical comorbidity of COVID-19 individuals with different severity (mild/non-hospitalized and severe/hospitalized) and matched negative comparisons, adjusted for ethnicity, BMI, smoking status, and CCI

Table 2
COVID-19 corona virus disease 2019, BMI body mass index, CCI Charlson comorbidity index, HR hazard ratio, CI confidence interval, COPD chronic obstructive pulmonary disease.Bold indicates p values less than 0.05

Table 3
Clinical comorbidity in COVID-19 individuals and matched negative comparisons by age, adjusted for ethnicity, BMI, smoking status, and CCI Several studies have found abnormalities in brain structures in COVID-19 patients, such as reduced grey matter thickness, tissue-contrast in the cortex and gyrus, and reduced overall brain size