{"paper_id":"6e452bea-7966-49f1-a5e6-8be88a1ef1cf","body_text":"Risk of long COVID and associated symptoms after acute SARS-COV-2 infection in ethnic \nminorities: a Danish nationwide cohort study \nGeorge F. Mkoma1*, Charles Agyemang2,3, Thomas Benfield4,5, Mikael Rostila6,7, Agneta \nCederström6,7, Jørgen H. Petersen8, Marie Norredam1,4 \nAffiliations \n1Danish Research Centre for Migration, Ethnicity and Health, Section of Health Services \nResearch, Department of Public Health, University of Copenhagen, Øster Farimagsgade 5, \nCopenhagen, 1014 Denmark. \n2Department of Public and Occupational Health, Amsterdam Public Health Research     Institute, \nAmsterdam UMC, University of Amsterdam, Amsterdam, The Netherlands. \n3Division of Endocrinology, Diabetes, and Metabolism, Department of Medicine, Johns Hopkins \nUniversity, Baltimore, United States. \n4Department of Infectious Diseases, Copenhagen University Hospital - Amager and Hvidovre, \nHvidovre, Denmark. \n5Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of \nCopenhagen, Copenhagen, Denmark. \n6Department of Public Health Sciences, Stockholm University, Stockholm, Sweden.   \n7Centre for Health Equity Studies (CHESS), Stockholm University/Karolinska Institutet, \nStockholm Sweden. \n8Section of Biostatistics, Department of Public Health, University of Copenhagen, Øster \nFarimagsgade 5, Copenhagen, 1014, Denmark. \n \n \n*Corresponding author                                                                                                                            \nGeorge F. Mkoma (george.mkoma@sund.ku.dk)  \n \n \n \n \n \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \nNOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.\n\n \n \nAbstract \nBackground  \nEthnic minorities living in high-income countries have been disproportionately affected by \nCOVID-19 in terms of infection rates and hospitalisations; however, less is known about long \nCOVID in this population. Our aim was to examine the risk of long COVID and associated \nsymptoms among ethnic minorities. \nMethods and Findings  \nA Danish nationwide register-based cohort study of individuals diagnosed with COVID-19 aged \n≥18 years (n=2 334 271) between January 2020 and August 2022. We calculated the risk of long \nCOVID diagnosis and long COVID symptoms among ethnic minorities compared with native \nDanes using multivariable Cox proportional hazard regression and logistic regression, \nrespectively. \nEthnic minorities from North Africa (adjusted hazard ratio [aHR] 1.41; 95% CI 1.12–1.79), \nMiddle East (aHR 1.38; 95% CI 1.24–1.55), Eastern Europe (aHR 1.35; 95% CI 1.22–1.49), and \nAsia (aHR 1.23; 95% CI 1.09–1.40) had significantly greater risk of long COVID diagnosis than \nnative Danes in both unadjusted and adjusted models. In the analysis by largest countries of \norigin, the greater risks of long COVID diagnosis were found in Iraqis (aHR 1.56; 95% CI 1.30–\n1.88), Turks (aHR 1.42; 95% CI 1.24–1.63), and Somalis (aHR 1.42; 95% CI 1.07–1.91) after \nadjustment for confounders. Significant factor associated with an increased risk of long COVID \ndiagnosis was COVID-19 hospitalisation. Furthermore, the odds of reporting cardiopulmonary \nsymptoms (including dyspnoea, cough, and chest pain) and any long COVID symptoms were \nhigher among North African, Middle Eastern, Eastern European, and Asian than among native \nDanes in both unadjusted and adjusted models.  \nConclusions  \nBelonging to an ethnic minority group was significantly associated with an increased risk of long \nCOVID indicating the need to better understand long COVID drivers and address care and \ntreatment strategies in this population.  \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n1 \n \nIntroduction \nGlobally, millions of people have now been infected with SARS-COV-2, the virus causing \ncoronavirus disease 2019 (COVID-19) [1]. Despite increased risk of hospitalisation and death \nin the first weeks of SARS-COV-2 infection, many COVID-19 survivors experience a range \nof symptoms including fatigue, cardiopulmonary symptoms (dyspnoea, cough, and chest \npain), and neurological symptoms (headache, depression, and memory loss) persisting \nbeyond weeks or months after the acute phase of COVID-19 infection; the condition known \nas long COVID as per National Institute for Health and Care Excellence (NICE) guidelines \n[2–5]. Long COVID or post-acute sequelae of COVID-19 is an emerging epidemic that is \nanticipated to affect the quality of life of many COVID-19 survivors [6,7]. Hence, \nunderstanding the demographic profile of long COVID sufferers is of use for planning \nhealthcare services.  \nEthnic minorities living in high-income countries have been disproportionately affected by \nCOVID-19 in terms of infection rates, hospitalisations, and severe morbidity [8]. However, \nstudies on long COVID among ethnic minorities are few and their findings suggest that this \npopulation exhibits a greater risk of long COVID [9–14]. For example, compared with the \nmajority White populations in the United States and the United Kingdom, individuals who \nbelong to Black and Asian ethnicity were observed to have a higher chance of reporting long \nCOVID symptoms after acute COVID-19 infection [9–13]. In the Netherlands, the risk of \nlong COVID was found to be higher in patients with Surinamese, Moroccan, and Turkish \norigin than in those with Dutch origin [14]. Overall, the previous studies have several \nshortcomings, including studies were based on a single hospital setting or localised area, the \nstudies did not compare symptoms distribution before and after COVID-19 diagnosis, and \nmost of the studies were survey-based. In addition, comorbidities and socioeconomic factors \nsuch as income and education were not considered in some studies, albeit these factors may \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n2 \n \nimpact the likelihood of reporting long COVID symptoms. Notwithstanding the limitations \nfrom the previous research, evidence has emerged showing that older age, disease severity, \nuse of intensive care, comorbidities, and not receiving COVID-19 vaccine are associated with \nincreased risk of long COVID in the general population [2,3,15].  \n \nIn the present study, nationwide register data from individuals diagnosed with COVID-19 in \nDenmark were used. First, we hypothesised that ethnic minorities (defined by their region \nand country origin) have a higher risk of long COVID diagnosis compared to native Danes \ntaking into account comorbidities, socioeconomic factors, civil status, COVID-19 related \nhospitalisation, and vaccination status. Second, we examined whether the risk of fatigue, \nheadache, cardiopulmonary, or any of these long COVID symptoms differed between ethnic \nminorities and native Danes within 6 months before COVID-19 diagnosis, 0 to 4 weeks, and \n>4 weeks to 6 months after COVID-19 diagnosis. \nMethods \nSetting \nDenmark has a population of approximately 5.8 million people. Testing for SARS-COV-2 \ninfection by PCR was launched in March 2020. During March–May 2020, testing for SARS-\nCOV-2 by PCR was offered for individuals with mild to severe symptoms of respiratory tract \ninfection [16]. Universal testing for SARS-COV-2 infection by PCR was nationally \nimplemented from May 18, 2020. Additionally, vaccination against COVID-19 started on \nDecember 27, 2020 [16]. The Danish healthcare system is publicly financed by general taxes \nand access to test for SARS-COV-2 and vaccination against COVID-19 is free of charge for \nall residents [17]. \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n3 \n \nData sources and study population \nThis nationwide register-based cohort study utilised data from the Danish COVID-19 \nsurveillance database, the Danish National Patient Registry (DNPR), the Danish Vaccination \nRegister, and Statistics Denmark. The study population included all individuals residing in \nDenmark who had first-time tested positive for SARS-COV-2 (COVID-19 diagnosis) aged \n18 years or older from January 1, 2020 to August 31, 2022 [18]. The study population was \nlinked with the DNPR, which is a nationwide hospital register containing information on all \nprimary and secondary diagnoses among hospitalised patients [19]. The DNPR contributed \ndata on individuals who had COVID-19 as the primary reason for hospitalisation identified in \naccordance with 10th version of International Standard Classification of Diseases (ICD-10): \nICD-10 codes B34.2, B34.2A, B97.2, or B97.2A. Furthermore, the DNPR provided \ninformation on comorbidities and symptoms related to hospital contacts before and after \nCOVID-19 diagnosis. We retrieved data on first, second, and third dose of COVID-19 \nvaccine from the Danish Vaccination Register [20]. Statistics Denmark contributed \nindividual-level data on country of origin, date of immigration, highest attained education, \nfamily income, civil status, and date of death [21–23]. Linkage between the registers was \npossible due to the availability of unique personal identification number assigned to all \nDanish residents [23].  \nRegion and country of origin \nThe study population was categorised based on individual and parental region and country of \norigin [24]. The following eight groups were constructed according to their region of origin, \nwith these groups being the modified version of those used by the World Bank: (i) Denmark \n(ii) Northern Europe, (iii) Western Europe, (iv) Eastern Europe, (v) Asia, (vi) Middle East, \n(vii) North Africa, and (viii) Subsaharan Africa [25]. Participants from North America, South \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n4 \n \nAmerica, and Oceania were excluded in the study as their numbers were relatively small. \nFurthermore, we classified the study population based on the largest countries of origin \namong the population of ethnic minorities residing in Denmark. The largest countries of \norigin selected were Norway, Sweden, Afghanistan, Iraq, Iran, Somalia, Pakistan, and \nTurkey. Individuals originating outside Denmark and their descendants formed the ethnic \nminority population. Participants originating and/or born in Denmark (native Danes) \nconstituted the reference group.  \nOutcome \nThe study participants were followed up from the date of a positive test for SARS-COV-2 \ninfection until a long COVID diagnosis, death, emigration, or study end (August 31, 2022), \nwhichever came first. The primary outcome of interest was long COVID diagnosis defined as \ncomplications persisting beyond the acute COVID-19 infection that cannot be explained by \nalternative diagnosis [26]. Presence of a long COVID diagnosis was determined by both ICD-\n10 codes (B94.8 or B94.8A) and the actual date of diagnosis. In addition, we examined \nhospital contacts related to long COVID symptoms such as fatigue, headache, dyspnoea, \nchest pain, cough, and depression or anxiety as a secondary outcome. Symptoms were \nidentified by ICD-10 codes in relation to the date of hospital contact (Supplementary Table \nS1). Due to small outcome events on a single symptom by ethnic group, some symptoms \nwere assessed as a composite outcome. In the present study, the following groups of \nsymptoms were considered: fatigue, headache, cardiopulmonary symptoms (including \ndyspnoea, cough, and chest pain), and any of these selected long COVID symptoms. The \nanalysis was restricted to the population experiencing these groups of symptoms within 6 \nmonths before COVID-19 diagnosis, 0 to 4 weeks (acute phase of COVID-19 infection), and \n>4 weeks to 6 months after COVID-19 diagnosis.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n5 \n \nCovariates \nCovariates included in the analysis were age, sex, comorbidities, civil status, highest attained \neducation, family income, length of residency, COVID-19 hospitalisation (as a proxy for \ndisease severity), and vaccination against COVID-19. Age was analysed as a continuous \nvariable and subsequently categorised as 18–60 years and >60 years in further analyses. \nCOVID-19 hospitalisation was assessed as yes or no. Presence of comorbidities was \ndetermined by Charlson Comorbidity Index (CCI) based on discharge diagnosis within five \nyears prior to COVID-19 diagnosis (Supplementary Table S1). The CCI included 17 \nconditions with scores assigned according to their severity [27]. The CCI score was divided \ninto three groups: 0 (indicating no comorbidity), 1–2, and ≥3. Civil status was classified as \ncohabiting, living alone, or other. Education was grouped as low, medium, or high in \naccordance with the International Standard Classification of Education [28]. Income was \ncategorised as low, middle, or high tertiles according to the total income distribution among \npatients with COVID-19 in the specific calendar year. Length of residency was a time \ndifference in years between date of arrival in Denmark and date of COVID-19 diagnosis. \nStatistical analyses \nCategorical and continuous variables were summarised by frequencies and percentages and \nby medians and interquartile ranges, respectively. We used multivariable Cox proportional \nhazard regression models to investigate the association between region and country of origin \nand the risk of long COVID diagnosis. Age, sex, civil status, education, family income, and \nCCI were identified as confounders using directed acyclic graphs; hence, these covariates \nwere adjusted in the Cox models (Supplementary Figure S1). We refrained from adjusting for \nlength of residency, COVID-19 hospitalisation, and COVID-19 vaccination status as these \ncovariates were deemed to belong in the causal pathway for the risk of long COVID/reporting \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n6 \n \nlong COVID symptoms. The proportional hazard assumption was assessed by Schoenfeld \nresiduals. In addition, we performed subgroup analyses in which the risk of long COVID \ndiagnosis was compared between ethnic minorities and native Danes by age groups (18–60 \nyears vs. >60 years), by COVID-19 hospitalisation (no vs. yes), and by COVID-19 \nvaccination (yes vs. no). The native Danes aged 18–60 years, native Danes non-hospitalised, \nand native Danes vaccinated were the reference groups in the three subgroup analyses. \nFurthermore, we assessed the association between region and country of origin and hospital \ncontacts related to groups of symptoms by fitting multivariable logistic regression models \nusing the same set of covariates like in the Cox models. We compared hospital contacts \nrelated to groups of symptoms within 6 months before vs. 6 months after COVID-19 \ndiagnosis in each ethnic group. Subsequently, we analysed hospital contacts related to groups \nof symptoms comparing ethnic minorities and native Danes in three time periods: 6 months \nbefore COVID-19 diagnosis, 0 to 4 weeks, and >4 weeks to 6 months after COVID-19 \ndiagnosis. All hazard ratios (HR) and odds ratios (OR) with their corresponding 95% \nconfidence interval (CI) were presented as unadjusted and adjusted, with native Danes \nregarded as the reference population. All analyses were performed in R statistical software \n(version 4.2.2).   \n \n \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n7 \n \nResults \nParticipants characteristics \nBetween January 2020 and August 2022, 2 334 271 individuals were first-time diagnosed \nwith COVID-19, of whom 40 321 (1.7%) were hospitalised and 2 292 950 (98.3%) were non-\nhospitalised individuals (Figure 1). Of the diagnosed COVID-19 cases, 1 973 998 (84.6%) \nwere native Danes and 360 273 (15.4%) were ethnic minorities. Overall, 6479 native Danes \nand 755 ethnic minorities died within 6 months after COVID-19 diagnosis. The results on \nsociodemographic characteristics of the study participants showed that compared with native \nDanes, ethnic minorities, particularly those from Eastern Europe, Asia, Middle East, North \nAfrica, and Subsaharan Africa were younger at the time of COVID-19 diagnosis and were \nmore likely to have low level of education and more likely to have low family income (Table \n1). North African (4.6%), Middle Eastern (4.2%), Eastern European (2.7%), Asian (2.8%), \nand Subsaharan African (2.5%) were in general more likely than native Danes (1.5%) to be \nhospitalised for COVID-19. Additionally, ethnic minorities from North Africa (28.5%) and \nMiddle East (26.1%) as well as those from Pakistan (28.5%), Turkey (27.8%), Iraq (27.5%), \nIran (26.2%), and Afghanistan (24.7%) had a higher proportion of individuals with \ncomorbidities (CCI score of 1–2) than native Danes (20.3%) (Supplementary Table S2). \nRisk of long COVID diagnosis \nWe found that North African (HR 1.35; 95% CI 1.10–1.67), Middle Eastern (HR 1.31; 95% \nCI 1.18–1.44), Eastern European (HR 1.21; 95% CI 1.11–1.32), and Asian (HR 1.14; 95% CI \n1.03–1.28) had a higher risk of long COVID diagnosis than native Danes in unadjusted model \n(Figure 2). After adjustment for age, sex, civil status, education, family income, and \ncomorbidities, the risk of long COVID diagnosis remained significantly higher in North \nAfrican (adjusted hazard ratio [aHR] 1.41; 95% CI 1.12–1.79), Middle Eastern (aHR 1.38; \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n8 \n \n95% CI 1.24–1.55), Eastern European (aHR 1.35; 95% CI 1.22–1.49), and Asian (aHR 1.23; \n95% CI 1.09–1.40) than in native Danes. In the analysis by largest countries of origin, the \nresults were most evident in individuals originating from Iraq (aHR 1.56; 95% CI 1.30–1.88), \nTurkey (aHR 1.42; 95% CI 1.24–1.63), and Somalia (aHR 1.42; 95% CI 1.07–1.91) (Figure \n3).  \nWhen investigating factors associated with increased risk of long COVID diagnosis, we \nobserved that of all ethnic-age groups, the risk of long COVID diagnosis was highest among \nSubsaharan African aged >60 years (aHR 1.72; 95% CI 1.17–2.52) (Supplementary Table \nS3). Compared with native Danes non-hospitalised, COVID-19 hospitalisation was \nsignificantly associated with a higher risk of long COVID diagnosis among both native \nDanes and ethnic minorities before and after adjustment for confounders (Table 2). However, \nthe hazard ratios of long COVID diagnosis for ethnic minorities like North African (aHR \n3.98; 95% CI 2.75–5.75), Middle Eastern (aHR 4.43; 95% CI 3.71–5.29), Eastern European \n(aHR 4.49; 95% CI 3.84–5.23), Asian (aHR 3.44; 95% CI 2.79–4.23), and Subsaharan \nAfrican (aHR 4.30; 95% CI 3.05–6.07) were still higher than that of native Danes (aHR 2.82; \n95% CI 2.64–3.00) among individuals hospitalised for COVID-19. Further analysis showed \nthat individuals not receiving COVID-19 vaccine exhibited greater risk of long COVID \ndiagnosis than individuals vaccinated; and the association was found in native Danes only \n(aHR 1.47; 95% CI 1.33–1.63).  \nHospital contacts related to long COVID symptoms \nThe majority of ethnic minority groups and native Danes exhibited higher probabilities and \nhigher odds of hospital contacts related to fatigue, headache, cardiopulmonary symptoms, and \nany long COVID symptoms within 6 months after COVID-19 diagnosis as compared to 6 \nmonths before COVID-19 diagnosis in the adjusted estimates (Figure 4 and Supplementary \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n9 \n \nTable S4). However, compared with native Danes, differences in odds ratios of hospital \ncontacts related to cardiopulmonary symptoms and any long COVID symptoms were more \npronounced among North African, Middle Eastern, Eastern European, Asian, and Northern \nEuropean, especially beyond 4 weeks to 6 months after COVID-19 diagnosis in both \nunadjusted and adjusted estimates (Table 3). Although the Subsaharan African group did not \nshow significant difference from native Danes in the odds of hospital contacts related to any \nlong COVID symptoms, this group was observed to have higher odds of hospital contacts \nrelated to symptoms like fatigue, headache, and cardiopulmonary symptoms beyond 4 weeks \nto 6 months after COVID-19 diagnosis. Moreover, analysis by largest countries of origin \nrevealed that ethnic minority groups, especially Swedes, Afghans, Iraqis, Iranians, Somalis, \nPakistanis, and Turks had higher odds of hospital contacts related to any long COVID \nsymptoms than native Danes, particularly beyond 4 weeks to 6 months after COVID-19 \ndiagnosis in both unadjusted and adjusted models (Supplementary Tables S5 and S6). \nDiscussion \nThis Danish nationwide cohort study found that compared with native Danes, the risk of long \nCOVID diagnosis was higher in the majority of ethnic minorities, notably for North African, \nMiddle Eastern, Eastern European, and Asian in both unadjusted and adjusted models. Our \nfindings also confirm that the chances of reporting cardiopulmonary symptoms (including \ndyspnoea, cough, and chest pain) and any long COVID symptoms were higher among North \nAfrican, Middle Eastern, Eastern European, and Asian than among native Danes, especially \nbeyond 4 weeks to 6 months after COVID-19 diagnosis in both unadjusted and adjusted \nmodels. In the analysis by largest countries of origin, this study found that the risk of long \nCOVID diagnosis was higher in Iraqis, Turks, and Somalis than in native Danes after \nadjustment for all relevant covariates. While chances of reporting any long COVID \nsymptoms were higher in Swedes, Afghans, Iraqis, Iranians, Somalis, Pakistanis, and Turks \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n10 \n \nthan in native Danes, particularly beyond 4 weeks to 6 months after COVID-19 diagnosis in \nboth unadjusted and adjusted estimates. \nThe observed higher risk of long COVID among ethnic minority groups than native Danes \nmay be explained by several factors including older age, COVID-19 hospitalisation, use of \nintensive care, and comorbidities as previously reported [2,3]. Compared with native Danes, \nolder age (>60 years) was only observed as a contributing factor to the increased risk of long \nCOVID in the Subsaharan African group. Another factor that was found to contribute to the \nrisk of long COVID was COVID-19 hospitalisation (a marker of disease severity). Our \nestimates demonstrate that COVID-19 hospitalisation was associated with increased risk of \nlong COVID in both ethnic minorities and native Danes. Notwithstanding the increased risk \nof long COVID in native Danes hospitalised for COVID-19, the present study found that \nethnic minorities were more likely than native Danes to be hospitalised for COVID-19. In \naddition, differences in hazard ratios of long COVID diagnosis for North African, Middle \nEastern, Eastern European, Asian, and Subsaharan African compared with that of native \nDanes hospitalised for COVID-19 were substantial. Hence, this may signify a greater burden \nof long COVID in these ethnic minority groups. Furthermore, the use of intensive care may \npartly contribute to the increased risk of long COVID among ethnic minorities. A recent \nDanish study using data from patients hospitalised for COVID-19 has reported that ethnic \nminorities originating from non-Western countries had a higher chance than native Danes of \nuse of mechanical ventilation [29], which could be seen as another marker of COVID-19 \nseverity associated with high burden of long COVID found in ethnic minorities living in \nDenmark. Apart from markers of disease severity, the risk of long COVID may also be \ninfluenced by presence of comorbidities. Our estimates show that individuals originating \nfrom North Africa and Middle East as well as those from Pakistan, Turkey, Iraq, Iran, and \nAfghanistan were more likely than native Danes to have a high burden of comorbidities. \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n11 \n \nDespite considering a wide range of comorbidities using Charlson comorbidity index in our \nmodels, the risk of long COVID remained significantly higher in ethnic minority groups \nreported. Therefore, this work suggests that the high burden of long COVID in this \npopulation may also be rooted in the complex interplay between the COVID-19 variant, \nimmunological factors, and circulatory system [30]. In line with our findings, previous \nstudies in the United States, United Kingdom, and Netherlands have reported that individuals \nof African, Asian, and Turkish origin exhibited higher chances of reporting long COVID \nsymptoms than the native majority population [14]. Recent evidence also shows that \nindividuals not receiving COVID-19 vaccine have a higher risk of long COVID in the general \npopulation.15 Although the majority of ethnic minorities were less likely than native Danes to \nreceive COVID-19 vaccine, their risk of long COVID did not differ by COVID-19 \nvaccination, indicating other factors such as barriers to accessing healthcare, differences in \nhealthcare demand, and late contact with healthcare when having COVID-19 may have \ncontributed to their increased risk of long COVID. \nStrengths and limitations \nCompared with previous research, the present study has several strengths, including using a \nnationwide sample of individuals diagnosed with COVID-19 in Denmark, using an \nestablished ICD-10 based diagnosis of long COVID in Denmark, and incorporating a wide \nrange of comorbidities and sociodemographic factors. To our knowledge, no previous study \non long COVID in ethnic minorities that had included symptoms experienced before COVID-\n19 diagnosis. Overall, studies on long COVID among ethnic minorities are still lacking and \nthis study is among the few to contribute knowledge to the existing body of literature. \nHowever, there are some limitations. First, long COVID diagnosis in the registers was \nimplemented from April 2020, which entails lack of registration of long COVID cases \nbetween January and March 2020 [26]. Second, symptoms included in the study were in \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n12 \n \nconnection with hospital contact and identified by ICD-10 codes, which may have introduced \nsome selection and registration bias. This is because these symptoms may not be \nrepresentative of long COVID situation in Denmark as some may choose not to contact \nhospital if the symptoms are not interfering with their daily routines. Additionally, the \nregistration of symptoms might also be problematic as individuals are more likely to receive \nclinical-based diagnoses rather than symptoms-based diagnoses in hospital setting. For that \nreason, we could not perform analysis of a single symptom by largest countries of origin due \nto small sample size. Therefore, it is possible that individuals are experiencing long COVID \nsymptoms more than what it is reported. We were unable to account for health-seeking \nbehaviours and cultural norms associated with healthcare contact as such data were \nunavailable. It may be the case that some ethnic minority groups have a certain tendency of \ncontacting the hospital, which may have influenced our estimates.  \nOur results have implications for clinical work and research. First, these findings raise \nintriguing question regarding preparedness and resilience of healthcare system in highly \nanticipated burden of long COVID. This implies that the health care system that dealt with \nthe acute consequences of COVID-19 now also has to consider a new situation of long-term \nand indirect consequences of the pandemic which puts new demands on health care \nprofessionals and the society. Second, the higher risk of long COVID in ethnic minorities is a \nmajor concern for equity in health and addressing this health problem may require \nmultisectoral response, funding, care, and treatment approaches which are culturally \nacceptable to this population. Lastly, future research should also focus on understanding key \ndrivers of long COVID and impact of long COVID on sick-leave and labour market \nparticipation among ethnic minorities.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n13 \n \nEthics statement: This study was approved by the Danish Data Protection Agency, reference \nnumber 514-0670/21-3000. No further approval is required regarding registry-based research. \nFinancial disclosure: This work was supported by the Novo Nordisk Foundation \n(ID:0067528). The funders had no role in the study design, data analysis, or writing of the \nmanuscript.  \nData availability: Data that supports the findings of this work are stored at Statistics \nDenmark and are not publicly available. Data access may be granted upon approval from the \nrelevant data custodians.                                                                                           \nAcknowledgements: None \nCompeting interests: Dr. Benfield reports grants from Novo Nordisk Foundation, Lundbeck \nFoundation, Simonsen Foundation, GSK, Pfizer, Gilead, Kai Hansen Foundation and Erik \nand Susanna Olesen’s Charitable Fund; personal fees from GSK, Pfizer, Bavarian Nordic, \nBoehringer Ingelheim, Gilead, MSD, Pentabase ApS, Becton Dickinson, Janssen and Astra \nZeneca; outside the submitted work. All other authors declared no potential competing \ninterests. \nAuthors’ contributions: GFM, CA, TB, MR, AC, JHP, and MN conceived the study and \nanalytical strategy. GFM performed the literature review, analysed cohort data and prepared \nresults. GFM, CA, TB, MR, AC, JHP, and MN contributed to the interpretation of results. \nGFM wrote the first draft of the report. All authors commented on the report draft and \napproved the final text. \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n14 \n \nReferences  \n1. World Health Organization. WHO coronavirus (COVID-19) dashboard 2023. \nAvailable from: https://covid19.who.int. \n2. Crook H, Raza S, Nowell J, Young M, Edison P. Long covid-mechanisms, risk \nfactors, and management. BMJ. 2021;374:n1648.  \n3. Lopez-Leon S, Wegman-Ostrosky T, Perelman C, et al. More than 50 long-term \neffects of COVID-19: a systematic review and meta-analysis. Sci Rep. \n2021;11(1):16144.  \n4. Caspersen IH, Magnus P, Trogstad L. Excess risk and clusters of symptoms after \nCOVID-19 in a large Norwegian cohort. Eur J Epidemiol. 2022;37(5):539–548.  \n5. National Institute for Health and Care Excellence. COVID-19 rapid guideline: \nmanaging the long-term effects of COVID-19 NICE guideline; 2020. Available from:  \nhttps://www.nice.org.uk/guidance/ng188/resources/covid19-rapid-guideline-\nmanaging-the-longterm-effects-of-covid19-pdf-51035515742. \n6. Malik P, Patel K, Pinto C, et al. Post-acute COVID-19 syndrome (PCS) and health-\nrelated quality of life (HRQoL)-A systematic review and meta-analysis. J Med Virol. \n2022;94(1):253–262. \n7. Xie Y, Xu E, Bowe B, Al-Aly Z. Long-term cardiovascular outcomes of COVID-19. \nNat Med. 2022;28(3):583–590.  \n8. Hayward SE, Deal A, Cheng C, et al; ESCMID Study Group for Infections in \nTravellers and Migrants (ESGITM). Clinical outcomes and risk factors for COVID-19 \namong migrant populations in high-income countries: A systematic review. J Migr \nHealth. 2021;3:100041.  \n9. Khullar D, Zhang Y, Zang C, et al. Racial/Ethnic Disparities in Post-acute Sequelae \nof SARS-CoV-2 Infection in New York: an EHR-Based Cohort Study from the \nRECOVER Program. J Gen Intern Med. 2023;38(5):1127–1136.  \n10. Yomogida K, Zhu S, Rubino F, Figueroa W, Balanji N, Holman E. Post-Acute \nSequelae of SARS-CoV-2 Infection Among Adults Aged ≥18 Years - Long Beach, \nCalifornia, April 1-December 10, 2020. MMWR Morb Mortal Wkly Rep. \n2021;70(37):1274–1277.  \n11. Subramanian A, Nirantharakumar K, Hughes S, et al. Symptoms and risk factors for \nlong COVID in non-hospitalized adults. Nat Med. 2022;28(8):1706–1714.  \n12. Thompson EJ, Williams DM, Walker AJ, et al. Long COVID burden and risk factors \nin 10 UK longitudinal studies and electronic health records. Nat Commun. \n2022;13(1):3528.  \n13. Halpin SJ, McIvor C, Whyatt G, et al. Postdischarge symptoms and rehabilitation \nneeds in survivors of COVID-19 infection: A cross-sectional evaluation. J Med Virol. \n2021;93(2):1013-1022.  \n14. Chilunga FP, Appelman B, van Vugt M, et al. Differences in incidence, nature of \nsymptoms, and duration of long COVID among hospitalised migrant and non-migrant \npatients in the Netherlands: a retrospective cohort study. Lancet Reg Health Eur. \n2023;29:100630.  \n15. Brannock MD, Chew RF, Preiss AJ, et al. Long COVID risk and pre-COVID \nvaccination in an EHR-based cohort study from the RECOVER program. Nat \nCommun. 2023;14(1):2914.  \n16. Fogh K, Strange JE, Scharff BFSS, et al. Testing Denmark: a Danish Nationwide \nSurveillance Study of COVID-19. Microbiol Spectr. 2021;9(3):e0133021.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n15 \n \n17. The Danish Patient Safety Authority–EU Health Insurance. How the Danish \nHealthcare System Works. https://lifeindenmark.borger.dk/healthcare/the-danish-\nhealthcare-system/how-the-danish-healthcare-system-works (accessed July 28, 2023).  \n18. Statens Serum Institut. COVID-19 surveillance database. Available from:  \nhttps://en.ssi.dk/covid-19. \n19. Schmidt M, Schmidt SA, Sandegaard JL, Ehrenstein V, Pedersen L, Sørensen HT. \nThe Danish National Patient Registry: a review of content, data quality, and research \npotential. Clin Epidemiol. 2015;7:449–490.  \n20. Grove Krause T, Jakobsen S, Haarh M, Mølbak K. The Danish vaccination register. \nEuro Surveill. 2012;17(17):20155.  \n21. Statistics Denmark. Income Statistics Register. Available from: \nhttps://www.dst.dk/en/Statistik/dokumentation/documentationofstatistics/income-\nstatistics.  \n22. Statistics Denmark. Education Attainment Register. Available from: \nhttps://www.dst.dk/en/Statistik/dokumentation/documentationofstatistics/highest-\neducation-attained.  \n23. Schmidt M, Pedersen L, Sørensen HT. The Danish Civil Registration System as a tool \nin epidemiology. Eur J Epidemiol. 2014;29(8):541–549.  \n24. Statistics Denmark. Immigrants and Descendants 2017 (Statistical presentation). \nAvailable from: \nhttps://www.dst.dk/Site/Dst/SingleFiles/GetArchiveFile.aspx?fi=91448101625&fo\n=0&ext=kvaldel.  \n25. The World Bank. Countries and Economies. Washington: The World Bank. Available \nfrom: https://data.worldbank.org/country.  \n26. The Danish Health Data Agency (Sundhedsdatastyrelsen). Report to the National \nPatient Register in Connection With COVID-19. Available from: \nhttps://sundhedsdatastyrelsen.dk/-/media/sds/filer/rammer-og-\nretningslinjer/patientregistrering/lpr_indberetningsvejledninger/patientregistrering-\n_covid_19.pdf?la=da.  \n27. Sundararajan V, Henderson T, Perry C, Muggivan A, Quan H, Ghali WA. New ICD-\n10 version of the Charlson comorbidity index predicted in-hospital mortality. J Clin \nEpidemiol. 2004;57(12):1288–1294.  \n28. International Standard Classification of Education. Available from: \nhttps://ec.europa.eu/eurostat/statistics-\nexplained/index.php?title=International_Standard_Classification_of_Education_(ISC\nED)#ISCED_1997_.28fields.29_and_ISCED-F_2013.  \n29. Norredam M, Islamoska S, Petersen JH, Benfield T. COVID-19 mortality and use of \nintensive care among ethnic minorities - a national register-based Danish population \nstudy. Eur J Epidemiol. 2023:1–9.  \n30. Davis HE, McCorkell L, Vogel JM, Topol EJ. Long COVID: major findings, \nmechanisms and recommendations. Nat Rev Microbiol. 2023;21(3):133–146.  \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n16 \n \n                                                          \n \n \n \n \n \n \n \n                                                                                                                                            \n \n \n \n \n \n \n        \n \n \n \n \n \n                                                                                                                                                  \n \n \n \n \n \nFigure 1. Flowchart of the study population.  \n \n \n \nHospitalised               \n(n=40 321) \nEthnic minorities               \n(n=10 091) \nDanes               \n(n=30 230) \nIndividuals who had \ntested positive for SARS-\nCOV-2 between January \n2020 and August 2022                                                                                  \n(n=2 334 271) \n \nEthnic minorities              \n(n=413) \nDeath           \n(n=4332) \nDanes              \n(n=3919) \nNon-hospitalised \n(n=2 293 950) \nDanes               \n(n=1 943 768) \nEthnic minorities   \n(n=350 182) \nDeath             \n(n=2902) \nEthnic minorities             \n(n=342) \nDanes             \n(n=2560) \nDeath within 6 months \nafter positive test for \nSARS-COV-2 \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n17 \n \n \n \n \nFigure 2. Hazard ratios of long COVID diagnosis by region of origin.  The adjusted model composed age, sex, civil status, education, family income, and Charlson comorbidity index. HR=hazard ratio. CI=confidence interval. \n \n \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n18 \n \n \n \n \nFigure 3. Hazard ratios of long COVID diagnosis by largest countries of origin.  The adjusted model composed age, sex, civil status, education, family income, and Charlson comorbidity index. HR=hazard ratio. CI=confidence interval. \n \n \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n19 \n \n \n \nFigure 4. Adjusted probabilities of hospital contacts related to specific symptoms within 6 months after COVID -19 diagnosis compared with 6 months before \nCOVID-19 diagnosis by region of origin. Hospital contacts related to cardiopulmonary symptoms included dyspnoe a (difficulty in breathing), cough, and chest pain as a \ncomposite outcome. Hospital contacts related to any long COVID symptoms included fatigue, headache, dyspnoea (difficulty in b reathing), cough, chest pain, depression \nand/or anxiety as a composite outcome. The adjusted model composed age, sex, civil status, education, family income, and Charlson comorbidity index.  \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n20 \n \nTable 1. Individuals who had first-time tested positive for SARS-COV-2 between January 2020 and August 2022 by region of origin.  \n Denmark Northern \nEurope \nWestern Europe Eastern Europe Asia Middle East North Africa Subsaharan \nAfrica \nn 1 952 021 19 842 37300 125 517 62 192 59 138 8693 22 252 \nImmigrants NA 18 131 (91.4%) 34 910 (93.6%) 100 063 (79.7%) 49 690 (79.9%) 45 539 (76.7%) 5146 (59.2%) 18 274 (82.1%) \nDescendants NA 1711 (8.6%) 2390 (6.4%) 25 454 (20.3%) 12 502 (20.1%) 13 819 (23.3%) 3547 (40.8%) 3978 (17.9%) \nLength of residency, years NA 34 (14–39) 28 (10–39) 26 (13–35) 24 (15–35) 22 (9–30) 31 (23–37) 21 (12–26) \nAge, years 61 (43–75) 60 (38–75) 57 (38–75) 45 (33–59) 46 (34–60) 45 (32–57) 52 (36–65) 41 (31–55) \nSex \n   Female 1 026 373 (52.6%) 12 471 (62.8%) 17 288 (46.3%) 66 216 (52.7%) 36 541 (58.7%) 29 022 (48.9%) 4386 (50.4%) 11 524 (51.8%) \n   Male 925 648 (47.4%) 7371 (37.2%) 20 012 (53.7%) 59 301 (47.3%) 25 651 (41.3%) 30 116 (51.1%) 4307 (49.6%) 10 728 (48.2%) \nCivil status \n   Cohabiting  862 240 (44.2%) 7478 (37.7%) 14 968 (40.1%) 61 418 (48.9%) 36 289 (58.3%) 25 103 (42.3%) 4222 (48.6%) 7300 (32.8%) \n   Living alone 852 792 (43.7%) 10 369 (52.3%) 19 203 (51.5%) 53 768 (42.9%) 20 893 (33.6%) 28 750 (48.4%) 3251 (37.4%) 12 230 (55.0%) \n   Other 236 989 (12.1%) 1995 (10.0%) 3129 (8.4%) 10 331 (8.2%) 5010 (8.1%) 5505 (9.3%) 1220 (14.0%) 2722 (12.2%) \nEducation \n   Low 465 063 (23.8%) 2056 (10.4%) 3400 (9.1%) 29 781 (23.7%) 17 855 (28.7%) 25 713 (43.3%) 3063 (35.2%) 9419 (42.3%) \n   Medium 904 288 (46.3%) 6757 (34.0%) 11 004 (29.5%) 51 271 (40.8%) 21 218 (34.1%) 17 769 (29.9%) 3191 (36.7%) 7500 (33.7%) \n   High 571 393 (29.3%) 9804 (49.4%) 20 482 (54.9%) 37 218 (29.7%) 19 081 (30.7%) 10 972 (18.5%) 1907 (22.0%) 3282 (15.2%) \n   Missing 11 277 (0.6%) 1225 (6.2%) 2414 (6.5%) 7247 (5.8%) 4038 (6.5%) 4904 (8.3%) 532 (6.1%) 1951 (8.8%) \nFamily income \n   Low 353 452 (18.1%) 6433 (32.4%) 13 275 (35.6%) 54 973 (43.8%) 28 051 (45.1%) 39 598 (66.7%) 4830 (55.6%) 14 264 (64.1%) \n   Middle 585 134 (30.0%) 4637 (23.4%) 8378 (22.5%) 39 808 (31.7%) 18 003 (28.9%) 9153 (15.4%) 2253 (25.9%) 4137 (18.6%) \n   High 859 757 (44.0%) 7173 (36.2%) 12 851 (34.4%) 19 974 (15.9%) 11 789 (19.0%) 5028 (8.5%) 961 (11.0%) 1835 (8.2%) \n   Missing 153 678 (7.9%) 1599 (8.0%) 2796 (7.5%) 10 762 (8.6%) 4349 (7.0%) 5579 (9.4%) 649 (7.5%) 2016 (9.1%) \nCOVID-19 hospitalisation 30 230 (1.5%) 343 (1.7%) 543 (1.4%) 3423 (2.7%) 1798 (2.8%) 2551 (4.2%) 413 (4.6%) 575 (2.5%) \nIntensive care 12 014 (0.6%) 100 (0.5%) 167 (0.4%) 504 (0.4%) 226 (0.4%) 280 (0.5%) 57 (0.6%) 148 (0.7%) \nCOVID-19 vaccination \n   One dose 1 813 312 (92.9%) 17 465 (88.0%) 31 702 (85.0%) 80 249 (63.9%) 55 003 (88.4%) 40 781 (68.7%) 5176 (59.5%) 15 400 (69.2%) \n   Two doses 1 796 381 (92.0%) 17 104 (86.2%) 31 027 (83.1%) 76 751 (61.1%) 53 820 (86.5%) 38 722 (65.2%) 4924 (56.6%) 14 483 (65.1%) \n   Three doses 1 489 444 (76.3%) 13 044 (65.7%) 23 220 (62.3%) 38 161 (30.4%) 35 449 (57.0%) 17 321 (29.2%) 2560 (29.4%) 5897 (26.5%) \nCharlson comorbidity index* \n    0 1 550 412 (79.4%) 16 412 (82.7%) 31 994 (85.8%) 101 384 (80.8%) 50 558 (81.3%) 43 825 (73.8%) 6205 (71.4%) 18 041 (81.1%) \n    1–2 396 200 (20.3%) 3381 (17.0%) 5234 (14.0%) 23 972 (19.1%) 11 563 (18.6%) 15 470 (26.1%) 2475 (28.5%) 4154 (18.7%) \n    ≥3 5409 (0.3%) 49 (0.3%) 72 (0.2%) 161 (0.1%) 71 (0.1%) 63 (0.1%) 13 (0.1%) 57 (0.2%) \nData are in median (IQR) or n (%). *Charlson comorbidity index composed myocardial infarction, congestive heart failure, peri pheral vascular disease, cerebrovascular disease, chronic obstructive pulmonary disease, rheumatic disease, dementia, peptic ulcer \ndisease, hemiplegia, diabetes without complications, diabetes with complications, mild liver disease, moderate to severe live r disease, renal disease, malignancy, metastatic cancer, and acquired immunodeficiency sy ndrome (AIDS). NA=not applicable. \n \n \n \n \n \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n21 \n \n \nTable 2. Hazard ratios of long COVID diagnosis by hospitalisation for COVID -19.  \n \n COVID-19 \nhospitalisatio\nn \n \nn \nUnadjusted \nHR (95% CI) \nAdjusted \nHR (95% CI) \nDenmark  No 1985 1.00 (reference) 1.00 (reference) \nYes 1483 6.53 (6.14 to 6.95) 2.82 (2.64 to 3.00) \nNorthern Europe  No 25 0.79 (0.58 to 1.07) 0.90 (0.66 to 1.23) \nYes 22 7.58 (4.98 to 11.52) 3.44 (2.21 to 5.34) \nWestern Europe  No 22 0.59 (0.43 to 0.81) 0.73 (0.53 to 1.01) \nYes 23 6.15 (4.08 to 9.28) 2.57 (1.69 to 3.92) \nEastern Europe  No 169 1.05 (0.94 to 1.17) 1.15 (1.02 to 1.30) \nYes 204 10.67 (9.27 to 12.29) 4.49 (3.84 to 5.23) \nAsia  No 96 1.04 (0.90 to 1.20) 1.14 (0.98 to 1.33) \nYes 108 8.83 (7.29 to 10.70) 3.44 (2.79 to 4.23) \nMiddle East  No 151 1.14 (1.01 to 1.29) 1.21 (1.05 to 1.39) \nYes 161 10.16 (8.67 to 11.90) 4.43 (3.71 to 5.29) \nNorth Africa  No 29 1.25 (0.96 to 1.63) 1.27 (0.94 to 1.71) \nYes 33 8.31 (5.89 to 11.73) 3.98 (2.75 to 5.75) \nSubsaharan Africa  No 31 0.77 (0.58 to 1.01) 0.99 (0.73 to 1.33) \nYes 37 8.48 (6.13 to 11.73) 4.30 (3.05 to 6.07) \n \nThe adjusted model composed age, sex, civil status, education, family income, and Charlson comorbidity index. \nHR=hazard ratio. CI=confidence interval. \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n22 \n \nTable 3. Odds ratios of hospital contacts related to specific symptoms by region of origin.  \n 6 months before COVID-19 diagnosis 0 to 4 weeks after COVID-19 diagnosis >4 weeks to 6 months after COVID-19 diagnosis \n  \nn \nUnadjusted  \nOR (95% CI) \nAdjusted  \nOR (95% CI) \n  \nn \nUnadjusted  \nOR (95% CI) \nAdjusted  \nOR (95% CI) \n \nn \nUnadjusted  \nOR (95% CI) \nAdjusted  \nOR (95% CI) \nHospital contacts related to fatigue \nDenmark  2004  1.00 (reference) 1.00 (reference) 561 1.00 (reference) 1.00 (reference) 969 1.00 (reference) 1.00 (reference) \nNorthern Europe  19  0.87 (0.63 to 1.21) 0.94 (0.67 to 1.33) 10 1.13 (0.67 to 1.92) 1.06 (0.58 to 1.93) 15 1.74 (1.23 to 2.45) 1.87 (1.30 to 2.69) \nWestern Europe  36 0.91 (0.71 to 1.18) 0.99 (0.74 to 1.31) 16 1.63 (1.14 to 2.31) 1.82 (1.27 to 2.62) 13 0.62 (0.40 to 1.00) 0.81 (0.51 to 1.28) \nEastern Europe  137 1.00 (0.87 to 1.15) 1.15 (0.98 to 1.34) 43 1.25 (1.01 to 1.57) 1.37 (1.05 to 1.78) 96 1.41 (1.19 to 1.68) 1.42 (1.16 to 1.74) \nAsia  56 0.83 (0.67 to 1.02) 0.99 (0.79 to 1.25) 15 1.15 (0.83 to 1.59) 1.40 (0.97 to 2.00) 36 1.08 (0.83 to 1.42) 1.26 (0.84 to 1.69) \nMiddle East  119 1.33 (1.14 to 1.56) 1.48 (1.24 to 1.77) 33 1.48 (1.13 to 1.94) 1.68 (1.25 to 2.27) 61 1.48 (1.19 to 1.83) 1.44 (1.13 to 1.85) \nNorth Africa  19 1.44 (1.01 to 2.07) 1.53 (1.04 to 2.24) § § § 10 1.14 (0.63 to 2.07) 1.04 (0.54 to 2.01) \nSubsaharan Africa  28 1.07 (0.78 to 1.47) 0.96 (0.64 to 1.44) 15 1.75 (1.11 to 2.75) 2.32 (1.42 to 3.77) 19 1.91 (1.35 to 2.71) 2.00 (1.35 to 2.96) \nHospital contacts related to headache \nDenmark  2765 1.00 (reference) 1.00 (reference) 512 1.00 (reference) 1.00 (reference) 1630 1.00 (reference) 1.00 (reference) \nNorthern Europe  34 1.24 (0.94 to 1.64) 1.04 (0.75 to 1.44) 8 1.07 (0.55 to 2.07) 1.19 (0.61 to 2.30) 17 1.56 (1.12 to 2.16) 1.73 (1.23 to 2.42) \nWestern Europe  38 0.69 (0.51 to 0.92) 0.80 (0.58 to 1.10) 10 0.97 (0.56 to 1.68) 1.21 (0.70 to 2.10) 22 0.95 (0.68 to 1.32) 1.08 (0.75 to 1.56) \nEastern Europe  276 1.99 (1.80 to 2.19) 1.39 (1.24 to 1.56) 60 2.39 (1.95 to 2.93) 1.68 (1.34 to 2.11) 183 2.44 (2.16 to 2.75) 1.66 (1.44 to 1.91) \nAsia  117 1.57 (1.35 to 1.83) 1.18 (1.00 to 1.40) 26 1.70 (1.23 to 2.35) 1.08 (0.74 to 1.58) 78 2.04 (1.71 to 4.59) 1.46 (1.19 to 1.79) \nMiddle East  206 2.51 (2.24 to 2.82) 1.60 (1.40 to 1.83) 42 2.38 (1.84 to 3.09) 1.34 (0.99 to 1.80) 121 2.48 (2.12 to 2.89) 1.52 (1.27 to 1.81) \nNorth Africa  37 3.47 (2.75 to 4.37) 2.27 (1.75 to 2.95) 8 2.60 (1.43 to 4.72) 1.21 (0.57 to 2.56) 17 2.23 (1.51 to 3.28) 1.53 (1.02 to 2.32) \nSubsaharan Africa  48 2.56 (2.08 to 3.14) 2.04 (1.64 to 2.53) 11 2.44 (1.53 to 3.90) 1.07 (0.58 to 1.96) 39 2.75 (2.11 to 3.59) 1.67 (1.22 to 2.26) \nHospital contacts related to cardiopulmonary symptoms* \nDenmark  14 019 1.00 (reference) 1.00 (reference) 4117 1.00 (reference) 1.00 (reference) 9027 1.00 (reference) 1.00 (reference) \nNorthern Europe  117 0.82 (0.72 to 0.92) 0.80 (0.70 to 0.92) 34 0.60 (0.46 to 0.78) 0.66 (0.50 to 0.86) 102 1.16 (1.02 to 1.33) 1.33 (1.16 to 1.53) \nWestern Europe  217 0.89 (0.81 to 0.98) 1.01 (0.92 to 1.12) 68 1.30 (1.13 to 1.49) 1.57 (1.36 to 1.81) 131 0.82 (0.72 to 0.93) 0.97 (0.85 to 1.11) \nEastern Europe  983 1.09 (1.04 to 1.14) 1.09 (1.03 to 1.15) 447 1.70 (1.58 to 1.82) 1.87 (1.73 to 2.02) 836 1.57 (1.50 to 1.66) 1.52 (1.43 to 1.61) \nAsia  486 1.09 (1.02 to 1.17) 1.19 (1.11 to 1.28) 178 1.35 (1.21 to 1.50) 1.48 (1.32 to 1.67) 387 1.53 (1.42 to 1.64) 1.55 (1.44 to 1.68) \nMiddle East  720 1.42 (1.34 to 1.49) 1.29 (1.21 to 1.37) 287 2.07 (1.91 to 2.25) 2.02 (1.84 to 2.22) 539 1.87 (1.76 to 1.98) 1.66 (1.54 to 1.78) \nNorth Africa  108 1.17 (1.02 to 1.35) 1.07 (0.92 to 1.26) 46 1.65 (1.32 to 2.05) 1.51 (1.18 to 1.92) 103 2.15 (1.87 to 2.46) 1.93 (1.67 to 2.25) \nSubsaharan Africa  168 1.06 (0.95 to 1.19) 1.11 (0.97 to 1.26) 50 1.12 (0.91 to 1.37) 1.21 (0.97 to 1.52) 122 1.27 (1.11 to 1.46) 1.26 (1.08 to 1.46) \nHospital contacts related to any long COVID symptoms* \nDenmark  25 375 1.00 (reference) 1.00 (reference) 6506  1.00 (reference) 1.00 (reference) 17 516 1.00 (reference) 1.00 (reference) \nNorthern Europe  233 0.91 (0.83 to 1.00) 0.90 (0.81 to 1.00) 54  0.75 (0.62 to 0.90) 0.83 (0.68 to 1.01) 204  1.20 (1.09 to 1.33) 1.36 (1.23 to 1.51) \nWestern Europe  410  0.92 (0.85 to 1.00) 1.05 (0.98 to 1.14) 120  1.31 (1.16 to 1.49) 1.52 (1.32 to 1.71) 273 0.90 (0.82 to 0.98) 1.10 (1.01 to 1.20) \nEastern Europe  1925  1.18 (1.14 to 1.23) 1.07 (1.03 to 1.11) 649 1.63 (1.54 to 1.73) 1.64 (1.53 to 1.75) 1587 1.55 (1.49 to 1.61) 1.33 (1.28 to 1.39) \nAsia  889  1.11 (1.05 to 1.16) 1.07 (1.02 to 1.13) 265 1.36 (1.25 to 1.49) 1.34 (1.21 to 1.48) 722  1.53 (1.45 to 1.61) 1.34 (1.27 to 1.43) \nMiddle East  1382 1.54 (1.48 to 1.60) 1.20 (1.14 to 1.26) 428 1.97 (1.84 to 2.11) 1.72 (1.59 to 1.86) 1009 1.80 (1.72 to 1.88) 1.31 (1.24 to 1.39) \nNorth Africa  225  1.43 (1.30 to 1.59) 1.17 (1.04 to 1.30) 64 1.50 (1.24 to 1.81) 1.26 (1.03 to 1.56) 195 2.13 (1.93 to 2.36) 1.71 (1.53 to 1.91) \nSubsaharan Africa  320 1.14 (1.05 to 1.24) 1.03 (0.94 to 1.14) 95 1.26 (1.08 to 1.47) 1.21 (1.02 to 1.44) 245 1.32 (1.20 to 1.46) 1.11 (0.99 to 1.24) \n \nHospital contacts related to cardiopulmonary symptoms included dyspnoea (difficulty in breathing), cough, and chest pain as a composite outcome. Hospital contacts related to any long COVID symptoms included fatigue, headache, dyspnoea (difficulty in \nbreathing), cough, chest pain, depression and/or anxiety as a composite outcome. ··Estimated could not be displayed due to small numbers. The adjusted model com posed age, sex, civil status, education, family income, and Charlson comorbidity index. \nOR=odds ratio. CI=confidence interval. \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n23 \n \nSupplementary material \nSearch strategy for literature review \nBlock 1 \n“Post-Acute COVID-19 Syndrome” OR “Post-Acute COVID-19 Syndromes” OR “COVID-19 Syndrome, Post-\nAcute” OR “Long Haul COVID-19” OR “COVID-19, Long Haul” OR “Long Haul COVID 19” OR “Long Haul \nCOVID-19s” OR “Post Acute COVID-19 Syndrome” OR “Post Acute COVID 19 Syndrome” OR “Long \nCOVID” OR “Post-Acute Sequelae of SARS-CoV-2 Infection” OR “Post Acute Sequelae of SARS CoV 2 \nInfection” OR “Post COVID Conditions” OR “Post-COVID Conditions” OR “Post-COVID Condition” OR \n“Long-Haul COVID” OR “COVID, Long-Haul” OR “Long Haul COVID” OR “Long-Haul COVIDs” \n  \nAND \n  \n  \nBlock 2 \n \n“migrant*” OR “transient” OR “refugee*” OR “asylum seeker*” OR “displaced person*” OR “asylee” OR \n“immigrant*” OR “foreigner*” OR “emigrant*” OR “ethnic minority” OR “racial minority” OR “diaspora*” \nOR “human migration” OR “undocumented immigrant*” OR “undocumented worker*” OR “unauthorized \nimmigrant*” OR “Black American*” OR “African American*” OR “black” OR “Afro -American*” or \n“African*” OR “Hispanic” OR “Latino” or “Latinx” OR “Hispanic*” OR “Latin American*” OR “Asian” OR \n“Asian American” OR “Middle Eastern” OR “Southwestern Asian” OR “ethnicity” OR “ethnic group” OR \n“nationality” OR “race” OR “racial group*” OR “Continental Population Group” OR “BAME” OR “BIPOC” \nOR “native-born*” or “indigenous people*” or “native people*”  \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n24 \n \nTable S1. ICD-10 codes of symptoms and diagnoses. \nSymptom ICD-10 codes \nFatigue R53, R539, R539A, R539C, R539E, R539F, G933 \nHeadache R51–R519 \nDyspnoea (difficulty in breathing) R06, R060, R060A, R061, R063, R064, R065, R066, R068, R068A, \nR068B, R068C, R068D, R068E \nCough R05–R059 \nChest pain R07–R074 \nDepression/anxiety F32–F339, F411–F418, F430–F4300 \nDiagnosis  \nCOVID-19 B34.2, B34.2A, B97.2, B97.2A \nLong COVID B94.8, B94.8A \nMyocardial infarction I21–I21.9, I22, I25.2 \nCongestive heart failure I10–I10.9, I11, I11.0, I13.0, I13.2, I25.5, I42.0, I42.6, I42.7, I42.8, \nI42.9, I43, I50, I50.0, I50.1, I50.9 \nPeripheral vascular disease I70, 171, I73.1, I73.8, I73.9, I77.1, I79.0, I79.2 \nCerebrovascular disease G45, I60–I64, I67, I69 \nChronic obstructive pulmonary \ndisease \nJ43, J44–J44.9 \nRheumatic disease M05, M06, M12.3, M07.0–M07.3 \nDementia F00–F03, F05.1, G30, G31.1, G31.9 \nPeptic ulcer disease K25–K28 \nHemiplegia G11.4, G80, G81, G82, G83.0–G83.3, G83.8 \nDiabetes without complications E10, E11, E10.0, E10.0, E10.1, E11.0, E11.1, E12.0, E12.1, E13.0, \nE13.1, E14.0, E14.1 \nDiabetes with complications E10, E11, E10.2, E10.5, E10.7, E11.2, E11.7, E12.2, E12.7, E13.2, \nE13.7, E14.2, E14.7  \nMild liver disease B15–B19, K70, K70.0, K70.1 \nModerate to severe liver disease I85.0, I85.9, I98.2, I98.3, K70.3, K70.4, K70.9, K73, K74.0, K74.6, \nK75.4  \nRenal disease I12.0, I13.1, N03.2–N03.7, N05.2–N05.7, N11, N18, N19, N25.0, \nQ61.1, Z49, Z94.0, Z99.2 \nMalignancy C00–C09, C10–C41, C45–C58, C60–C76, C81–C86, C88–C97 \nMetastatic cancer C77–C80 \nAcquired immunodeficiency \nsyndrome \nB20–B24, R75, Z21 \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n25 \n \n \n \nFigure S1. Directed Acyclic Graphs for confounders assessment. Age, sex, civil status, comorbidities, \neducation, and income were identified as confounders. \n \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n26 \nTable S2. Individuals who had first-time tested positive for SARS-COV-2 between January 2020 and August 2022 by largest countries of origin.  \n Denmark Norway Sweden Afghanistan Iraq Iran Somalia Pakistan Turkey \nn 1 952 021 7200 7078 9273 15 600 9200 9077 12 462 35 460 \nImmigrants NA 6680 (92.8%) 6395 (90.4%) 8455 (91.2%) 12 571 (80.1%) 7667 (83.3%) 6274 (69.1%) 7031 (56.4%) 18 764 (52.9%) \nDescendants NA 520 (7.2%) 683 (9.6%) 818 (8.8%) 3119 (19.9%) 1533 (16.7%) 2803 (30.9%) 5431 (43.6%) 16 696 (47.1%) \nLength of residency, years NA 36 (17–41) 36 (16–38) 19 (16–21) 22 (19–26) 27 (14–34) 24 (20–27) 36 (29–39) 34 (26–38) \nAge, years 61 (43–75) 62 (39–76) 64 (41–76) 44 (32–58) 48 (32–60) 49 (36–60) 43 (28–56) 56 (42–71) 49 (36–62) \nSex          \n   Female 1 026 373 (52.6%) 4752 (66.0%) 4308 (60.8%) 4470 (48.2%) 7807 (49.8%) 4400 (47.8%) 4856 (53.4%) 6415 (51.5%) 18 358 (51.8%) \n   Male 925 648 (47.4%) 2448 (34.0%) 2770 (39.2%) 4803 (51.8%) 7793 (50.2%) 4800 (52.2%) 4221 (46.6%) 6047 (48.5%) 17 102 (48.2%) \nCivil status \n   Cohabiting  862 240 (44.2%) 2615 (36.3%) 2903 (41.0%) 4211 (45.4%) 6402 (40.8%) 4126 (44.8%) 2030 (22.4%) 7832 (62.8%) 19 954 (56.3%) \n   Living alone 852 792 (43.7%) 3824 (53.1%) 3464 (48.9%) 4610 (49.7%) 7730 (49.3%) 3832 (41.7%) 5894 (64.9%) 3672 (29.5%) 11 723 (33.0%) \n   Other 236 989 (12.1%) 761 (10.6%) 711 (10.1%) 452 (4.9%) 1558 (9.9%) 1242 (13.5%) 1153 (12.7%) 958 (7.7%) 3783 (10.7%) \nEducation \n   Low 465 063 (23.8%) 597 (8.3%) 663 (9.4%) 3363 (36.3%) 6399 (40.8%) 2365 (25.7%) 4684 (51.6%) 4426 (35.5%) 15 902 (44.8%) \n   Medium 904 288 (46.3%) 2464 (34.2%) 2406 (34.0%) 3183 (34.3%) 5242 (33.4%) 3060 (33.3%) 2724 (30.0%) 4086 (32.8%) 11 904 (33.6%) \n   High 571 393 (29.3%) 3769 (52.4%) 3522 (49.7%) 1581 (17.0%) 2728 (17.4%) 3309 (36.0%) 682 (7.5%) 3293 (26.4%) 5304 (15.0%) \n   Missing 11 277 (0.6%) 370 (5.1%) 487 (6.9%) 1146 (12.4%) 1321 (8.4%) 466 (5.0%) 987 (10.9%) 657 (5.3%) 2350 (6.6%) \nFamily income \n   Low 353 452 (18.1%) 2441 (33.9%) 1877 (26.5%) 5783 (62.4%) 9801 (62.5%) 4073 (44.3%) 6921 (76.3%) 6437 (51.7%) 16 345 (46.1%) \n   Middle 585 134 (30.0%) 1423 (19.8%) 1541 (21.8%) 1896 (20.4%) 2845 (18.1%) 2042 (22.2%) 1108 (12.2%) 3243 (26.0%) 11 055 (31.2%) \n   High 859 757 (44.0%) 2728 (37.9%) 3160 (44.6%) 839 (9.1%) 1571 (10.0%) 2279 (24.8%) 249 (2.7%) 1904 (15.3%) 5632 (15.9%) \n   Missing 153 678 (7.9%) 608 (8.4%) 500 (7.1%) 755 (8.1%) 1473 (9.4%) 806 (8.7%) 799 (8.8%) 878 (7.0%) 2428 (6.8%) \nCOVID-19 hospitalisation 30 230 (1.5%) 123 (1.7%) 142 (2.0%) 370 (3.9%) 831 (5.1%) 366 (3.9%) 306 (3.3%) 618 (4.8%) 1547 (4.3%) \nIntensive care 12 014 (0.6%) 35 (0.5%) 37 (0.5%) 28 (0.3%) 78 (0.5%) 36 (0.4%) 68 (0.7%) 74 (0.6%) 160 (0.4%) \nCOVID-19 vaccination \n   One dose 1 813 312 (92.9%) 6320 (87.8%) 6258 (88.4%) 7565 (81.5%) 11 037 (70.3%) 7859 (85.4%) 5713 (62.9%) 9957 (79.9%) 23 754 (67.0%) \n   Two doses 1 796 381 (92.0%) 6199 (86.1%) 6184 (87.4%) 7246 (78.1%) 10 482 (66.8%) 7689 (83.6%) 5248 (57.8%) 9586 (76.9%) 22 559 (63.6%) \n   Three doses 1 489 444 (76.3%) 4889 (67.9%) 4856 (68.6%) 3667 (39.5%) 4684 (29.8%) 5278 (57.3%) 1586 (17.5%) 4523 (36.3%) 10 760 (30.3%) \nCharlson comorbidity index* \n  0 1 550 412 (79.4%) 5971 (82.9%) 5789 (81.8%) 6974 (75.2%) 11 356 (72.4%) 6774 (73.6%) 7179 (79.1%)  8900 (71.4%) 25 568 (72.1%) \n  1–2 396 200 (20.3%) 1210 (16.8%) 1274 (18.0%) 2291 (24.7%) 4318 (27.5%) 2410 (26.2%) 1884 (20.8%) 3549 (28.5%) 9856 (27.8%) \n  ≥3 5409 (0.3%) 19 (0.3%) 15 (0.2%) 8 (0.1%) 16 (0.1%) 16 (0.2%) 14 (0.1%) 13 (0.1%) 36 (0.1%) \nData are in median (IQR) or n (%). *Charlson comorbidity index composed myocardial infarction, congestive heart failure, peri pheral vascular disease, cerebrovascular disease, chronic obstructive pulmonary disease, rheumatic disease, dementia, peptic ulcer \ndisease, hemiplegia, diabetes without complications, diabetes with complications, mild liver disease, moderate to severe live r disease, renal disease, malignancy, metastatic cancer, and acquired immunodeficiency syndrome (AIDS). NA=n ot applicable. \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n27 \n \nTable S3. Hazard ratios of long COVID diagnosis by age group.  \n \n  \nAge group \n \nn \nUnadjusted \nHR (95% CI) \nAdjusted \nHR (95% CI) \nDenmark  18–60 2126 1.00 (reference) 1.00 (reference) \n>60 1342 0.76 (0.72 to 0.80) 0.52 (0.49 to 0.55) \nNorthern Europe  18–60 32 1.08 (0.79 to 1.48) 1.44 (1.05 to 1.98) \n>60 15 0.50 (0.33 to 0.75) 0.34 (0.22 to 0.52) \nWestern Europe  18–60 26 0.59 (0.42 to 0.83) 0.74 (0.52 to 1.06) \n>60 19 0.62 (0.43 to 0.90) 0.47 (0.33 to 0.68) \nEastern Europe  18–60 293 1.04 (0.93 to 1.15) 1.19 (1.06 to 1.33) \n>60 80 1.15 (0.97 to 1.35) 0.85 (0.70 to 1.04) \nAsia  18–60 150 0.95 (0.83 to 1.10) 0.93 (0.79 to 1.08) \n>60 54 1.13 (0.92 to 1.38) 1.11 (0.90 to 1.37) \nMiddle East  18–60 243 1.10 (0.99 to 1.24) 1.16 (1.02 to 1.32) \n>60 69 1.31 (1.07 to 1.61) 1.04 (0.83 to 1.32) \nNorth Africa  18–60 47 1.38 (1.08 to 1.77) 1.38 (1.06 to 1.79) \n>60 15 0.83 (0.55 to 1.26) 0.60 (0.36 to 1.01) \nSubsaharan Africa  18–60 53 0.67 (0.52 to 0.86) 0.85 (0.64 to 1.12) \n>60 15 1.60 (1.11 to 2.31) 1.72 (1.17 to 2.52) \n \nThe adjusted model composed age, sex, civil status, education, family income, and Charlson comorbidity index. \nHR=hazard ratio. CI=confidence interval. \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n28 \n \nTable S4. Odds ratio of hospital contacts related to specific symptoms 6 months after COVID -19 \ndiagnosis compared with 6 months before COVID-19 diagnosis by region of origin. \n \n 6 months before \nCOVID-19 diagnosis \n6 months after COVID-19 diagnosis \n  \nOR (95% CI) \nUnadjusted \nOR (95% CI) \nAdjusted \nOR (95% CI) \nHospital contacts related to fatigue \nDenmark  1.00 (reference) 3.10 (2.96 to 3.25) 3.06 (2.91 to 3.21) \nNorthern Europe  1.00 (reference) 4.41 (2.84 to 6.84) 4.21 (2.64 to 6.71) \nWestern Europe  1.00 (reference) 2.93 (1.99 to 4.31) 3.31 (2.20 to 4.97) \nEastern Europe  1.00 (reference) 2.76 (2.27 to 3.36) 2.62 (2.10 to 3.27) \nAsia  1.00 (reference) 2.62 (1.94 to 3.54) 2.70 (1.94 to 3.76) \nMiddle East  1.00 (reference) 2.28 (1.81 to 2.87) 2.49 (1.93 to 3.21) \nNorth Africa  1.00 (reference) 1.07 (0.54 to 2.09) 0.98 (0.46 to 2.10) \nSubsaharan Africa  1.00 (reference) 2.95 (1.90 to 4.58) 4.89 (2.90 to 8.24) \nHospital contacts related to headache \nDenmark  1.00 (reference) 3.30 (3.16 to 3.45) 3.30 (3.14 to 3.46) \nNorthern Europe  1.00 (reference) 3.05 (2.04 to 4.55) 3.65 (2.34 to 5.70) \nWestern Europe  1.00 (reference) 3.91 (2.59 to 5.88) 3.50 (2.24 to 5.46) \nEastern Europe  1.00 (reference) 2.63 (2.29 to 3.03) 2.84 (2.43 to 3.33) \nAsia  1.00 (reference) 2.65 (2.13 to 3.29) 2.58 (2.02 to 3.29) \nMiddle East  1.00 (reference) 2.11 (1.78 to 2.51) 2.34 (1.93 to 2.84) \nNorth Africa  1.00 (reference) 1.36 (0.91 to 2.02) 1.60 (1.03 to 2.50) \nSubsaharan Africa  1.00 (reference) 1.88 (1.36 to 2.59) 2.00 (1.41 to 2.84) \nHospital contacts related to cardiopulmonary symptoms \nDenmark  1.00 (reference) 3.97 (3.91 to 4.04) 4.10 (4.03 to 4.17) \nNorthern Europe  1.00 (reference) 3.68 (3.10 to 4.37) 4.45 (3.68 to 5.37) \nWestern Europe  1.00 (reference) 3.40 (2.96 to 3.90) 3.55 (3.06 to 4.12) \nEastern Europe  1.00 (reference) 4.00 (3.76 to 4.26) 4.47 (4.17 to 4.80) \nAsia  1.00 (reference) 3.51 (3.21 to 3.83) 3.69 (3.35 to 4.07) \nMiddle East  1.00 (reference) 3.57 (3.31 to 3.83) 4.31 (3.96 to 4.68) \nNorth Africa  1.00 (reference) 3.99 (3.31 to 4.82) 4.68 (3.81 to 5.74) \nSubsaharan Africa  1.00 (reference) 2.68 (2.27 to 3.16) 3·38 (2.81 to 4.07) \nHospital contacts related to any long COVID symptoms \nDenmark  1.00 (reference) 4.14 (4.09 to 4.19) 4.30 (4.25 to 4.36) \nNorthern Europe  1.00 (reference) 3.78 (3.33 to 4.28) 4.36 (3.80 to 5.01) \nWestern Europe  1.00 (reference) 3.73 (3.37 to 4.14) 3.92 (3.51 to 4.38) \nEastern Europe  1.00 (reference) 3.71 (3.54 to 3.89) 4.15 (3.93 to 4.38) \nAsia  1.00 (reference) 3.60 (3.36 to 3.85) 3.71 (3.44 to 4.01) \nMiddle East  1.00 (reference) 3.24 (3.07 to 3.43) 3.83 (3.59 to 4.08) \nNorth Africa  1.00 (reference) 3.37 (2.93 to 3.86) 4.07 (3.49 to 4.73) \nSubsaharan Africa  1.00 (reference) 2.73 (2.41 to 3.09) 3.29 (2.86 to 3.77) \n \nHospital contacts related to cardiopulmonary symptoms included dyspnoea (difficulty in breathing), cough, and \nchest pain as a composite outcome. Hospital contacts related to any long COVID symptoms included fatigue, \nheadache, dyspnoea (difficulty in breathing), cough, chest pain, depression and/or anxiety as a composite \noutcome. The adjusted model composed age, sex, civil status, education, family income, and Charlson \ncomorbidity index. OR=odds ratio. CI=confidence interval. \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n29 \n \nTable S5. Odds ratio of hospital contacts related to any long COVID symptoms 6 months after COVID -19 \ndiagnosis compared with 6 months before COVID-19 diagnosis by largest countries of origin. \n \n 6 months before \nCOVID-19 diagnosis \n6 months after COVID-19 diagnosis \n  \nOR (95% CI) \nUnadjusted \nOR (95% CI) \nAdjusted \nOR (95% CI) \nDenmark  1.00 (reference) 4.14 (4.09 to 4.19) 4.30 (4.25 to 4.36) \nNorway  1.00 (reference) 2.59 (2.08 to 3.21) 2.69 (2.13 to 3.39) \nSweden  1.00 (reference) 5.22 (4.26 to 6.39) 5.89 (4.70 to 7.37) \nAfghanistan  1.00 (reference) 4.61 (3.95 to 5.37) 4.67 (3.87 to 5.63) \nIraq  1.00 (reference) 3.04 (2.73 to 3.38) 3.68 (3.25 to 4.17) \nIran  1.00 (reference) 4.52 (3.91 to 5.23) 5.16 (4.40 to 6.04) \nSomalia  1.00 (reference) 2.74 (2.31 to 3.26) 3.24 (2.67 to 3.93) \nPakistan  1.00 (reference) 3.18 (2.84 to 3.57) 3.38 (2.98 to 3.83) \nTurkey  1.00 (reference) 3.50 (3.25 to 3.77) 3.92 (3.61 to 4.25) \n \n \nHospital contacts related to any long COVID symptoms included fatigue, headache, dyspnoea (difficulty in \nbreathing), cough, chest pain, depression and/or anxiety as a composite outcome. The adjusted model composed \nage, sex, civil status, education, family income, and Charlson comorbid ity index. OR=odds ratio. CI=confidence \ninterval. \n \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint \n\n \n30 \n \n \nTable S6. Odds ratios of hospital contacts related to any long COVID symptoms by largest countries of origin.  \n \n \n 6 months before COVID-19 diagnosis 0 to 4 weeks after COVID-19 diagnosis >4 weeks to 6 months after COVID-19 diagnosis \n n (%) Unadjusted  \nOR (95% CI) \nAdjusted  \nOR (95% CI) \nn (%) Unadjusted  \nOR (95% CI) \nAdjusted  \nOR (95% CI) \nn (%) Unadjusted  \nOR (95% CI) \nAdjusted  \nOR (95% CI) \nHospital contacts related to any long COVID symptoms* \nDenmark  25 375 (1.3) 1.00 (reference) 1.00 (reference) 6506 (0.3) 1.00 (reference) 1.00 (reference) 17 516 (0.9) 1.00 (reference) 1.00 (reference) \nNorway  83 (1.1) 0.91 (0.79 to 1.05) 0.88 (0.75 to 1.03) 14 (0.2) 0.74 (0.54 to 1.01) 0.75 (0.54 to 1.04) 62 (0.8) 0.99 (0.83 to 1.17) 1.04 (0.86 to 1.25) \nSweden  81 (1.1) 0.87 (0.74 to 1.01) 0.81 (0.68 to 0.97) 23 (0.3) 0.73 (0.53 to 1.01) 0.76 (0.54 to 1.07) 81 (1.1) 1.57 (1.36 to 1.80) 1.74 (1.50 to 2.02) \nAfghanistan  178 (1.9) 1.25 (1.11 to 1.40) 1.04 (0.91 to 1.19) 52 (0.5) 1.73 (1.44 to 2.09) 1.60 (1.28 to 1.99) 170 (1.8) 1.91 (1.70 to 2.14) 1.31 (1.13 to 1.51) \nIraq  382 (2.4) 1.51 (1.40 to 1.62) 1.19 (1.09 to 1.30) 118 (0.7) 1.73 (1.51 to 1.98) 1.50 (1.29 to 1.75) 280 (1.7) 1.71 (1.57 to 1.87) 1.26 (1.14 to 1.39) \nIran  210 (2.2) 1.57 (1.41 to 1.74) 1.42 (1.27 to 1.59) 60 (0.6) 2.37 (2.01 to 2.79) 2.26 (1.89 to 2.70) 133 (1.4) 1.79 (1.58 to 2.01) 1.59 (1.40 to 1.81) \nSomalia  151 (1.6) 1.38 (1.22 to 1.55) 1.10 (0.96 to 1.26) 47 (0.5) 1.61 (1.30 to 2.00) 1.51 (1.20 to 1.90) 121 (1.3) 1.68 (1.46 to 1.92) 1.24 (1.06 to 1.44) \nPakistan  278 (2.2) 1.07 (0.98 to 1.16) 0.94 (0.86 to 1.04) 95 (0.7) 1.43 (1.24 to 1.65) 1.25 (1.07 to 1.47) 217 (1.7) 1.64 (1.51 to 1.79) 1.37 (1.25 to 1.50) \nTurkey  749 (2.1) 1.29 (1.23 to 1.37) 1.07 (1.01 to 1.14) 284 (0.8) 2.06 (1.89 to 2.24) 1.93 (1.75 to 2.12) 666 (1.8) 1.73 (1.63 to 1.83) 1.32 (1.24 to 1.42) \n \nHospital contacts related to any long COVID symptoms included fatigue, headache, dyspnoea (difficulty in breathing), cough, chest pain, depression and/or anxiety as a composite outcome. T he adjusted model composed age, sex, civil status, education, family \nincome, and Charlson comorbidity index. OR=odds ratio. CI=confidence interval. \n \n \n \n . CC-BY 4.0 International licenseIt is made available under a \n is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)\nThe copyright holder for this preprint this version posted August 24, 2023. ; https://doi.org/10.1101/2023.08.22.23294402doi: medRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}