key: cord-264709-p835wf4f
authors: Menezes, A. M. B.; Victora, C. G.; Hartwig, F. P.; Silveira, M. F.; Horta, B. L.; Barros, A. J. D.; Whermeister, F. C.; Mesenburg, M. A.; Pellanda, L. C.; Dellagostin, O. A.; Struchiner, C. J.; Burattini, M. N.; Barros, F. C.; Hallal, P. C.
title: High prevalence of symptoms among Brazilian subjects with antibodies against SARS-CoV-2: a nationwide household survey
date: 2020-08-12
journal: nan
DOI: 10.1101/2020.08.10.20171942
sha: 
doc_id: 264709
cord_uid: p835wf4f

Since the beginning of the pandemic of COVID-19, there has been a widespread assumption that most infected persons are asymptomatic. A frequently-cited early study from China suggested that 86% of all infections were undocumented, which was used as indirect evidence that patients were asymptomatic. Using data from the most recent wave of the EPICOVID19 study, a nationwide household-based survey including 133 cities from all states of Brazil, we estimated the proportion of people with and without antibodies for SARS-CoV-2 who were asymptomatic, which symptoms were most frequently reported, the number of symptoms reported and the association between symptomatology and socio-demographic characteristics. We were able to test 33,205 subjects using a rapid antibody test that was previously validated. Information on symptoms was collected before participants received the test result. Out of 849 (2.7%) participants who tested positive for SARS-CoV-2 antibodies, only 12.1% (95%CI 10.1-14.5) reported no symptoms since the start of the pandemic, compared to 42.2% (95%CI 41.7-42.8) among those who tested negative. The largest difference between the two groups was observed for changes in smell or taste (56.5% versus 9.1%, a 6.2-fold difference). Symptoms change in smell or taste, fever and myalgia were most likely to predict positive test results as suggested by recursive partitioning tree analysis. Among individuals without any of these three symptoms (74.2% of the sample), only 0.8% tested positive, compared to 18.3% of those with both fever and changes in smell or taste. Most subjects with antibodies against SARS-CoV-2 in Brazil are symptomatic, even though most present only mild symptoms.

or taste. Most subjects with antibodies against SARS-CoV-2 in Brazil are symptomatic, 50 even though most present only mild symptoms. 51

Since the beginning of the pandemic of COVID-19, there is a widespread notion that 54 most people infected by SARS-CoV-2 are asymptomatic, following an early article 55 from China stating that 86% of those infected did not report any symptoms. 1 More 56 recently, several clinical studies became available, showing that the prevalence of 57 asymptomatic infected individuals ranges from 4% to 75%. 2-6 These discrepancies 58 might be explained by the use of different lists of symptoms, different recall periods, as 59 well as different populations. Population-based studies are particularly relevant for 60 studying SARS-CoV-2 symptoms, because asymptomatic patients or those with mild 61 symptoms may be identified at home, rather than in health service-based studies. 62 63 Using data from the most recent wave of the EPICOVID19 study, a nationwide 64 household-based survey including 133 cities from all states of Brazil, 7 we estimate the 65 proportion of people with and without antibodies for SARS-CoV-2 who were 66 asymptomatic. We investigated which symptoms were most frequently reported, how 67 many symptoms were reported by each subject, and the associations between 68 symptoms and sociodemographic characteristics. We also performed conditional 69 inference tree analyses using binary recursive partitioning to identify which 70 combinations of symptoms were most likely to predict positive test results. populous municipality in each region was included in the sample. So far, the study has entailed three waves of data collection (May 14-21, June 4-7, and June 21-24). Here we 78 report on findings from the third wave of data collection which included a detailed 79 investigation of symptoms. 80 81 A multi-stage probabilistic sample was adopted, with 25 census tracts selected in each 82 one of the 133 sentinel cities, with probability proportionate to size. In each sampled 83 tract, 10 households were systematically selected, totaling 250 households per 84 municipality. All household residents were listed, and age and sex recorded on a list. 85

One individual was then randomly selected as the respondent for that household. Then, 86 a finger prick blood sample was obtained and a questionnaire applied. If the selected 87 subject did not accept to participate, a second resident was randomly chosen. In case of 88 another refusal, the interviewers moved to the next household to the right of the one 89 that had been originally selected. The total planned sample size was 33,250 individuals. 90

The WONDFO SARS-CoV-2 Antibody Test (Wondfo Biotech Co., Guangzhou, China) 92 was used for the detection of antibodies for SARS-CoV-2 93 (https://en.wondfo.com.cn/product/wondfo-sars-cov-2-antibody-test-lateral-flow-94 method-2/); this rapid point-of-care test is based on the principle of immune assay of 95 year completed/grade; recoded as primary or less; secondary; university or higher), self-115 reported skin color, and household assets. The official Brazilian classification of 116 ethnicity recognizes five groups, based on the question: "What is your race or color?" 117

The five response options are "white", "brown" ("pardo" in Portuguese), "black", 118 "yellow" and "indigenous". Interviewers were instructed to check the "yellow" option 119 when the respondent mentions being of Asian descent, and "indigenous" when any of 120 the multiple first nations were mentioned. We also performed conditional inference tree analyses using binary recursive 153 partitioning, accounting for multiple testing 13 . The objective of these analyses was to 154 identify which combinations of the 11 symptoms were most likely to predict positive 155 test results. 156 157 Analyses were performed using the software Stata version 14.1 (StataCorp, College 158

Station, TX, USA) and conditional inference tree analyses were performed using R 159 3.6.1 (https://www.r-project.org/). Data will become publicly available 30 days after 160 completion of the fieldwork at http://www.epicovid19brasil.org/. 161

Of the target sample size comprising 33,250 individuals, we were able to include 33,205 164 (99.9%) participants in the study (missing information for 45 subjects). To achieve this 165 number, a total of 59,724 houses were contacted, with 19.8% of refusals and 24.6% of 166 houses being empty at the time of the visit. Of the 31,869 participants included (after 167 excluding for missing on symptoms and previous COVI-19 diagnosis), 849 subjects 168

(2.7%) tested positive for SARS-CoV-2 antibodies. Test results were only disclosed 169 after the interview on symptoms had been completed. Table 1 shows the distribution of 170 the sample according to sociodemographic characteristics. 171 172 Each of the 11 symptoms investigated were significantly (P<0.01) more likely to be 173 reported by those testing positive as compared to those testing negative ( Table 2 ). The 174 most frequently reported symptoms among positive cases were headaches (58.0%), 175 change in smell or taste (56.5%), fever (52.1%), cough (47.7%) and myalgia (44.1%).

to SARS-CoV-2 antibodies. The largest ratios between positive and negative subjects 178 were observed for change in smell or taste (6.2-fold), fever (4.3-fold), shivering (3.3-179 fold) and myalgia (2.8-fold). The sensitivity and specificity for positive test results, for 180 each symptom, are presented in Supplementary Table. The two symptoms with 181 sensitivity above 50% and specificity above 85% were change in smell or taste, 182 followed by fever. In the first two waves of the EPICOVID19 nationwide survey, we identified that, 219 contrary to what is often reported, most subjects with antibodies were symptomatic. 220

However, symptoms had only been assessed for those with positive tests, and the 221 information was collected after the individual had learned about the test result. We 222 addressed the possibility of bias by asking all participants, regardless of the test result, 223

in the third wave. The question on symptoms covered the four-month period since the 224 first COVID-19 cases were reported in the country. The questionnaire was applied 225 before the test result was known, so that respondents were blind to their serological 226 status, and this allowed us to compare symptoms among those testing positive and those testing negative. Subjects with a previous diagnosis of COVID-19 and missing 228 information for symptoms (0.73% of the whole sample) were excluded from the 229 analyses in order to ensure that the respondents were not aware of their condition. 230

The above results from the third wave of the study confirmed a high prevalence of 232 symptoms using a 4-month recall period; only 12.1% positive subjects were 233 asymptomatic, compared to 42.2% of those without antibodies. Inclusion in our 234 analyses of individuals who tested negative was useful for identifying which symptoms 235 were most strongly associated with the presence of antibodies. For example, headaches 236

were the most common symptom affecting 58.0% of those positive, but were also 237 reported by 35.5% of those who tested negative, a prevalence ratio of only 1.6. In 238 contrast, changes in smell or taste affected 56.5% among those who tested positive and 239 9.1% in the negative ones, respectively. This symptom provided the best discrimination, 240 with a prevalence ratio of 6.2. Recent studies have shown that when SARS-CoV-2 241 enters the nasal and oral epithelium through the angiotensin-converting enzyme 2 242 (ACE2) and transmembrane serine protease 2 (TMPRSS2), it may cause damages to 243 olfactory and gustatory receptor cells resulting in anosmia or ageusia 14, 15 . 244

Overall, symptoms were more frequent among females than males, in subjects aged 30-246 29 years and in those with higher education. Children and adolescents were 247 substantially less likely to report symptoms than adults, which is compatible with the 248 lower infection-fatality rates observed in these age groups 16 . In contrast, prevalence of 249 symptoms fell with age from 30 to over 70 years, which does not reflect the age pattern 250 in infection-fatality and case-fatality 17 . The difference in reported symptoms between 251 women and men is also at odds with the higher case-fatality among males 18 .

Comparison of our findings on the prevalence of symptoms with the literature are 253 affected by the settings in which studies were done, by the phase of infection, the 254 duration of recall, and by the ways in which symptoms were recorded, as well as 255 whether or not the subjects were aware or suspicious of being infected. The prevalence 256 for asymptomatic subjects in the literature ranges from 4% to 75% 2-6, 19, 20 , whereas in 257 our study it was 12.1%. We identified five published reviews that provided pooled 258 prevalence estimates for symptoms 4, 5, 21-23 among individuals who tested positive in 259 health facilities. We found lower prevalence (52.1%) for fever (pooled prevalence 260 ranging from 78.4% to 92.8%) and cough (47.7% versus pooled prevalence ranging 261 from 58.3% to 72.2%). Our estimates for myalgia (44.1%) and difficulty breathing 262 (23.1%) were within the ranges reported in the studies (29.4% to 51.0%, and 20.6% to 263 45.6%, respectively). Lastly, prevalence of headache in our study (58%) was 264 considerably higher than in the reviews (8.0% to 14.0%). One may assume the 265 prevalence ranges of symptoms based on individuals who sought care in medical 266 facilities would tend to be higher than in our population-based survey, but this was not 267 the case, except for fever or cough. 268 269 Notably, change in smell or taste was not investigated in these review papers. We 270 searched the literature and change in smell or taste or anosmia/ ageusia was identified in 271 a multicenter European study with prevalence of 85.6% (anosmia) and 88.0% 272 (ageusia) 24 and a very low prevalence in a retrospective study in China (5.1% for 273 hyposmia and 5.6% for hypogeusia) 25 , whereas we found 56.5%. 274

Besides the aforementioned symptoms, some studies have hypothesized that the the gastrointestinal (GI) tract and play lead to GI manifestations 26 . The pooled 278 prevalence of GI symptoms has ranged in the literature from 7.4 to 12.5% for diarrhea 279 (against 25.6% in our study), and 4.6% to 10.2% for nausea and/or vomiting (compared 280 to 9.5% in our study) [26] [27] [28] . 281

It is likely that the information on symptoms from population-based studies, such as the 283 one from Spain 29 , would be comparable to our study; however, the recall time in that 284 study was two weeks, compared to up to four months in our survey. In this study, the 285 only symptom specifically reported was anosmia, that was present around 27% of 286 positive subjects, in the three waves. 287

The decision tree analyses were useful for identifying a subgroup of individuals who 289 presented both fever and change in smell or taste, among whom seroprevalence was 290 18.3%, compared to only 0.8% among subjects that did not present these two 291 symptoms, nor presented body aches. 292 293 It is clear from the literature that no single symptom correlates perfectly with SARS-294

CoV-2 infection, thus raising the possibility that the use of multiple symptoms might be 295 appropriate for screening purposes. However, the literature on this topic is still scarce. 296

A study using app-based self-reported data in the United States and in the United 297

Kingdom identified that change in smell or taste is the single symptom most strongly 298 correlated with infection and, using stepwise logistic regression, identified a prediction 299 model that also includes fatigue, persistent cough and loss of appetite 30 . We also 300 identified change in smell or taste as the single most predictive symptom, but the two 301 additional symptoms prioritized in the conditional inference tree analysis were fever and myalgia. Given that the symptoms are partially correlated to one another, it is possible 303 that models including different symptoms yield similar predictions, and would therefore 304 be of similar practical use. Another app-based study including mostly individuals in the 305 United Kingdom identified that, collectively, symptoms improve predicting prognosis 31 . 306

This indicates that symptoms may be used not only for screening, but also for patient 307 monitoring and planning health service needs. 308 309 Our study has limitations. Differentiation recall bias is a concern, particularly by using a 310 4-month recall period, but the alternative -as in the Spanish survey -was to ask for 311 symptoms in a shorter, more recent period and potentially misclassifying individuals 312 who had the disease in the past, and for whom antibodies remained detectable. In order 313 to evaluate the likelihood of differential recall bias, we excluded the 242 participants 314 who had a diagnosis of COVID-19 prior to the interview. Another limitation is the 315 growing evidence that antibody levels decrease rapidly over time, for example by 14% 316 in the same subjects in the Spanish study 29 , and in our own (unpublished) analyses 317 comparing the first and third waves of the survey in cities with high initial prevalence. 318

This would lead some individuals who had the disease to test negative, and yet report 319 symptoms that occurred at the time of the episode. This type of bias would reduce the 320 difference in reported symptoms among subjects who tested positive and negative. An 321 additional limitation is the growing evidence that antibody levels decrease rapidly over 322 time, for example by 14% in the same subjects in the Spanish study 28 , and in our own 323 (unpublished) analyses comparing the first and third waves of the survey in cities with 324 high initial prevalence. This would lead some individuals who had the disease to test 325 negative, and yet report symptoms that occurred at the time of the episode. This characteristic of the dynamics of the infection would reduce the difference in reported 327 symptoms among subjects who tested positive and negative. 328 329 Positive aspects of our study, on the other hand, included the population basis over an 330 area of 8.5 million square km, the large sample size, collection of symptoms in positive 331 and negative cases, and blinding of respondents as test results were only disclosed after 332 the clinical history was collected. 333

In summary, our analyses show that most individuals with antibodies against SARS-334

CoV-2 report having presented symptoms, even though in most cases these were mild. 335

Our findings can be used to implement surveillance systems in Brazil 

Substantial undocumented infection facilitates the 361 rapid dissemination of novel coronavirus (SARS-CoV-2)

Follow-up of asymptomatic patients with SARS

CoV-2 infection

Covid-19: identifying and isolating asymptomatic people helped 366 eliminate virus in Italian village

Clinicopathological characteristics of 8697 patients 368 with COVID-19 in China: a meta-analysis

Clinical characteristics of coronavirus disease 370 2019 (COVID-19) in China: A systematic review and meta-analysis

Asymptomatic patients as a source of

COVID-19 infections: A systematic review and meta-analysis

EPICOVID19 protocol: repeated 375 serological surveys on SARS-CoV-2 antibodies in Brazil

Test performance evaluation of SARS

CoV-2 serological assays

Cohort Profile Update: The 1982

Pelotas (Brazil) Birth Cohort Study

Sensitivity and specificity of a 383 rapid test for assessment of exposure to SARS-CoV-2 in a community-based setting in

Características étnico-raciais da população: 386 classificações e identidades: Instituto Brasileiro de Geografia e Estatística--IBGE; 2013. 387 12. Rutstein SO. The DHS wealth index: approaches for rural and urban areas: 388 Macro International Incorporated

Unbiased Recursive Partitioning: A Conditional 390

Olfactory and gustatory 393 dysfunctions in COVID-19 patients: A systematic review and meta-analysis. Asian Pac 394 J Allergy Immunol. 2020. 395 15. Butowt R, Bilinska K. SARS-CoV-2: Olfaction, Brain Infection, and the Urgent 396 Need for Clinical Samples Allowing Earlier Virus Detection

Estimating the infection fatality risk of

COVID-19 in

Estimates of the severity of coronavirus 402 disease 2019: a model-based analysis

The male excess in case-fatality rates 404 for COVID-19. A meta-analytic study of the age-related differences and consistency 405 over six countries

CoV-2 infections: a living systematic review and meta-analysis

A Systematic Review of Asymptomatic Infections 410 with COVID-19

Prevalence of comorbidities and its effects in 412 patients infected with SARS-CoV-2: a systematic review and meta-analysis

Clinical, laboratory and imaging features of COVID-19: A systematic review and meta-416 analysis

Olfactory and gustatory 420 dysfunctions as a clinical presentation of mild-to-moderate forms of the coronavirus 421 disease (COVID-19): a multicenter European study

Neurological Manifestations of Hospitalized 424 Patients with COVID-19 in Wuhan, China: a retrospective case series study

Novelty in the gut: a 427 systematic review and meta-analysis of the gastrointestinal manifestations of COVID-428 19

Gastrointestinal Symptoms and Fecal Viral Shedding in Patients With Coronavirus 431 Disease 2019: A Systematic Review and Meta-analysis

Gastrointestinal Manifestations of

Infection and Virus Load in Fecal Samples From a Hong Kong Cohort: 435 Systematic Review and Meta-analysis

Estudio nacional de 438 sero-epidemiologia de la infección por SARS-CoV-2 en España

Real-time tracking of self-reported 442 symptoms to predict potential COVID-19

Symptom clusters in Covid19: A 444 potential clinical prediction tool from the COVID Symptom study app