An open-label prospective observational study of antiandrogen and non-antiandrogen early pharmacological approaches in females with mild-to-moderate COVID-19. The Pre-AndroCoV Female Trial

Background: While COVID-19 remains largely unclear and mortality continues to raise, early effective approaches prior to complications lack, as well as resea rches for characterization and therapeutical potential options in actual early COVID-19. Although females seem to be less affect ed than females, hyperandrogenic (HA) phenotype, like polycystic ovary syndrome (PCOS), idiopathic hirsutism, congenital adrenal hyperplasia (CAH) female androgenetic alopecia (AGA), or idiopathic HA may be at higher risk due to its inherent enhanced androgenic activity. The present study aimed to evaluate the effects of any early pharmac ological approach to females diagnosed with COVID -19 . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice. before seven days of symptoms, as well as investig ate whether HA is an additional risk factor in this population.

Females with symptoms for less than seven days confirmed for COVID-19 through positive real-time polymerase chain reaction (rtPCR-SARS-CoV-2) were classified and divided as non-HA, HA, and HA using spironolactone (HA-spiro) groups. Patients were questioned for baseline characteristics, 23 different diseases, 4 4 drug classes and vaccines, 2 8 different symptoms, and eight different parameters to measure COVID-19 related cli nical outcomes. Treatment was then provided, inc luding azithromycin 500mg/day for five days in all cases, associated with hydroxychloroquine 400mg/day for five days , nitazoxanide 500mg twice a day for six days, or ivermectin 0.2mg/kg/day por three days, and optionally spironolactone 100mg twice a day until cure. Patients were assessed for COVID -19 clinical course, clinical and viral duration, and disease progression.

In total, 270 females were enrolled, including 195, 67, and eight in non-HA, HA, and HA-spiro groups, respectively. Prevailing symptoms were anosmia (71.1%), ageusia (67.0%), headache (48.1%), myalgia (37.4%), dry cough (36.3%), nasal congestion or rhinorrhea (34.1%), fatigue (33.3%), weakness (29.5%), hyporexia (27.8%), thoracic pain (24.8%), diarrhea (2 4.1%) and dizziness ( 21.5%). Earliest symptoms (days) were dizziness (1.0 ± 0.2 day), abdominal pain (1.1 ± 0.3); conjunctival hyperemia (1.1 ± 0.5), nasal congestion or rhinorrhea (1.2 ± 0.5), headache (1.2 ± 0.5), dry cough (1.2 ± 0.5), myalgia (1.2 ± 0.4), nauseas (1.3 ± 0.5) and weakness (1.3 ± 0.5). Time-to-treat, positive rtPCR, and duration of symptoms with a nd without anosmia and ageusia were significantly lower in HA-spiro than non-HA, HA, and overall non -users. Time-to-treat was similar while all duration of symptoms and positive rtPCR -SARS-CoV-2 were significantly shorter in non-HA than HA. Spironolactone users were more likely to be asymptomatic than non-users during COVID-19. Fewer non-HA than HA females were affected by anosmia, ageusia, dry cough , fatigue, w eakness and hyporexia. A geusia, weakness and myalgia lasted shorter in non-HA than HA. None of the patients needed hospitalization or any other COVID-19 complication.

A sensitive, early detection of COVID -19 followed by a pharmaceutical approach with different drug comb inations yielded irrefutable differences compared to sex-, age-, body mass index (BMI) -, and disease-matched non-treated controls in terms of clinical outcomes, ethically disallowing placebo-control randomized clinical trials in . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint the early stage of COVID-19 due to the marked improvements. HA females presented more severe and prolonged clinical manifestations, although none progressed to worse outcomes. Spironolactone mitigated the additional risks due to HA.

COVID-19 is a multi-systemic and multi-factorial syndrome caused by the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS -CoV-2). Its exact mechanisms of action are still largely unclear, and despite the massive number of infections and deaths, effective approaches before it becomes severe lack (1-4). One of the most likely reasons to explain why we have failed to detect effec tive approaches is that while we are sea rching for molecules with antiviral activity, we are detecting COVID-19 too late, wh en viral infectivity no longer plays a key role in the pathophysiology at that stage, which will naturally lead to lack of effic acy from these antiviral approaches. While we have focused the vast majority of the researches on patients after they acute lung injury and hospitalized patients , a relative shortage of res earches in actual earlier stages of COVID-19, in comparison to the relevance of trying to discover effective approaches for secondary prevention, i.e., preventions of COVID-19 complications after its detection. Meanwhile. number of research es that allege t o have researched in mild patients actually included hospitalized patients only, which is inherently c ontradictory (5). Because of the larg e pre-symptomatic period, asy mptomatic infected subjects , prolonged incubation and viral she dding period, and unrevealed means of transmission, viral spreading remains, despite all unprecedented public policies. Yet the most characteristic and specific s ymptoms have been extensively described, unspecific clinical manifestations, particularly in the first days of the disease . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint (since anosmia and ageusia tend to appear aft er three a five days only), highly heterogeneous clinical presentation, and lack of good predictors of those who will further develop acute lung injury are still challenges to detect COVID-19 during the period when therapies focusing on antiviral activity may still be effective (6-9). An additional challenge to be overcome it the persistence of policies focusing on the sine-quo-non presence of fever or shortness of breath to perform the diagnostic real- time Polymerase Chain Reaction (rtPCR) for SARS -CoV-2 (10,11). While these two symptoms should not be considered as signs for the presence of COVID-19, but for severe COVID-19 instead, we will fail to diagnose COVID -19 when complications are potentially avoidable. The reports on the literature claiming that fever is present in the majority of patients with COVID -19 are based on data collecte d from registers that require fever to diagnose COVID -19, which is per se a limitation for a more accurate description of COVID-19 manifestations. Reports based on diagnostic tools for COVID- 19 that do not require fever show that fever may be present in as low as 10% of infected patients only (12-14). Several different molecules demonstrated in vitro antiviral activity against SARS- CoV-2 and have been proposed as promising therapies for COVID-19 (13), among which the most attempted drug combinations included azithromycin in the majority of the cases, in association with hydroxychloroquine , ivermectin or nitazoxanide (13,14). However, since detection of COVID -19 is predominantly delayed due the mandatory presence of fever for its suspect, antiviral approaches will be less effective, since at this stage of the disease viral infectivity becomes less central . Accordingly, randomized clinical trials (RCT) on alleged early COVID-19 yielded conflicting results, although the majority have been exclusively performed in hospitalized patients (10-14). We hav e concluded that pharmacological therapies for truly early and mild COVID-19 ha s not been investigated thoroughly, which precludes from conclusive findings regarding the efficacy of antiviral approaches at this stage. To evaluate potential antiviral therap ies, it is critical to detect COVID -19 during the f irst days after its appearance, which is only feasible with more sensitive approaches its diagnosis. . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Although underreported, risks for acute lung injury, thrombosis and other clinical complications in COV ID-19 are also related to increased exposure, enhanced activity, and/or hypersensitivity to androgens (15-23). Overrepresentation of males in terms of complications related to COVID -19 are not fully justified by differences in age, body mass index (BMI), prevalence of comorbidities, i.e., there is an inherent risk related to male sex (15-18). This is likely explained, at least partially, by the transmembrane serine protease 2 (TMPRSS-2), a critical protein for the SARS-CoV-2 entry into the cells, that are largely and solel y regulated by an drogens. Among males , androgenetic alopecia (AGA) as an independent predictor of complications related to COVID-19, possibly due to a resultant of overexpression of a ndrogen recept ors (AR), due to enhanced dihydrotestosterone (DHT) l evels, activity, response, or a combination between these factors, that discloses AGA as a clinical phenotype expression. While females are un derrepresented am ong severe COVID -19 patients, risk factors including menopause, aging, uncompensated type 2 dia betes mellitus (T2DM), hypertension and obesity seem to enhance the risk of severe COVID-19 in females more than in males. In addition, in an analogically similar manner than AGA males, females with any expression of hyperandrogenis m (HA), including polycy stic ovary syndrome (PCOS), idiopathic hirsutism, congenital adrenal hyperplasia (CAH) due to 21alpha- hydroxylase or 11beta-hydroxylase deficiency, female AGA, or idiopathic HA, have sufficient mechanistical plausibility to support the hypothesis that this population may be at higher risk compared to non-HA females. In this regard, the use of antiandrogens ha s demonstrated promising results, as already observed for both males and females, at least when used chronically (19-22). This reinforces the role of the role of antiandrogen approaches as an additional path to improve outcomes in COVID-19. Nonetheless, similarly to the use of antiviral therapies, antiandrogens should be tested during the first stage, as it affects viral infectivity. There were suffici ent theo retical, mechanisti cal, observational and epidemiological observations to intuitively hypothesize that if the lack of sensitivity is to detectCOVID-19 is addressed and therefore diagnosed during the first stage, preferably before seven day s of symp toms, antiviral pharma cological a ttempts could be then . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint effective. At this point, it is uncertain whether this approach would be affective, which is our objective. Together, the evaluation of sex diff erences, as well as differenc es between phenotypes within each sex, would also disclose additional information for promising approaches for specific populations. The objectives of the present study are to perform a thorough and comprehensive clinical characterization of patients with COVID-19 detected through a highly sensitive case-detection basis, and to explore the clinical responses and outcomes to a var iety of drug com binations. In addition, we aimed to detec t sex -specific and androge nic phenotype-specific clinical man ifestations and outcomes. This is an open-label prospective observational st udy performed alongside with our currently ongoing RCT (ClinicalTrials.gov Identifier: NCT044 46429. Available at clinicaltrials.gov (https://clinicaltrials.gov/ct2/show/NCT04446429?term=NCT04446429&draw=2&rank =1). The present study has received approval from the Institutional Review Board (IRB) of the Ethics Committee of the National Board of Ethics Committee of the Ministry of Health, Brazil (CEP/CONEP: Parecer 4.173.074 / CAAE: 34110420.2.0000.0008).

Subject selection This spe cific study is a n open label prospective observational study of the characterization and clinical outcomes of females with COVID-19 in response to specific therapeutic combinations. In order to detect cases durin g the earliest stages of COVID- 19, we employed a highly sensitive case-detection criteria for suspect for COVID-19. We changed from the mandatory presence of severity or specific signs (shortness of breath, fever, anosmia, ageusia) to the occurrence of absolutely any atypical symptom or changes in patters of chronic symptoms, even when not supposedly related to COVI D-19. Suspected females underwent rtPCR-SARS-CoV-2 (Abb ott RealTime SARS -CoV-2 . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Assay, Abbott, USA; or Cobas SARS-CoV-2, Roche, Switzerland), and those confirmed for SARS-CoV-2 were included. Inclusion criteria inc luded: 1. Confirmed COVID -19 through posi tive rtPCR - SARS-CoV-2; 2. 18 years old and above; 3. Le ss than seven days of beginning of symptoms; 4. Less than 72 hours of COVID -19 confirmation (in case COVID-19 had already been confirmed); 4. Non use or use for less than 24 hours of any potential antiviral drug; and 5. No previous use of glucocorticoid in the past seven days. Design and methods Parameters evaluated by the present study are depicted in Table 1. Females included in the study were actively questioned for baseline and medical characteristics, including 23 different diseases, 44 drug classes and vaccines, 28 different symptoms, in addition to the search for HA, which includes: 1. PCOS confirmed by two o f th ree Rotterdam Criteria; 2. Previously diagnosed CAH; 3. Known hyperandrogenism, clinical- or biochemically; 4. Hirsutism; and 5. Female AG A. Females were then divided according to t he presence or absence of HA (non-HA group) , and in case of hyperandrogenism, use of spironolactone 100mg/day or above (HA-Spiro group) or not (HA group). Clinical presentations were clustered into one of the following 1. Anosmia - Ageusia dominance; 2. Dengue -like symptomatology / clinical presentation; 3. Upper respiratory tract infection (URTI) URTI -like symptomatology / clinical presentation; 4. Gastrointestinal (GI) in fection-like symp tomatology / clinical presentation; 5. Mixed between clusters; 6. Unspecific presentation; or 7. Asymptomatic. To fill criteria for each cluster, it has been required for: 1. Anosmia -Ageusia dominance: at l east two of nas al congesti on or rhinorrhea , dry cough, self-reported perception of “sinusitis”, or self-reported perception of “sore throat”; 2. Dengue-like clinical presentation: at least three of myalgia, arthralgia, upper back pain, conjunctival hyperemia or pre-orbital pain; 3. URTI -like clinic al presentation: at l east two of nasal congestion or rhinorrhea , dry cough, self-reported perception of “sinusitis”, or self-reported perception of “sore throat”; . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint 4. GI infection-like clinical presentation: at least two of dia rrhea, nauseas, vomiting, or abdominal pain; 5. Mixed between clusters: when there are symptoms to fill criteria for at least two clusters 6. Unspecific presentation: when there are only unspecific or insufficient symptoms to fill criteria for any cluster; or 7. Asymptomatic. After characterization, drug combination including azithromycin 500mg/day for five days, with hydroxychloroquine 400mg/day for five days, nitazoxanide 500mg twice a day for six days, or ivermectin 0.2mg/kg/day for three days was then provided. The choice between hydroxychloroquine, nitazoxanide, and/or ivermectin was based on an almost-random manner, i.e., random ly, except when clinical judgement considere d otherwise. In addition, spironolactone, vitamin D, vitamin C, zinc, apibaxan, rivaroxaban, enoxaparin, and glucocorticoids could have been prescribed, also according to medical judgement. Patients were then evaluated for : 1. Time-to-appearance and duration of each symptom (number of days) ; 2. Time until full re mission of symptoms , not including anosmia and a geusia (number of days); 3. Time until full remis sion of symptoms, including anosmia and ageusia ( number of days); 4. Duration of positive rtPCR-SARS- CoV-2 (in number of days, where rtPCR was performed every seven days) ; 5. Level of clinical improvement in Days -7 to -4, -3 to -1, 0, 1, 2, 3, 7, 14, 21, 30 and 60 days, where 0 corresponds to the worst day of symptoms (scored according to the number and severity of symptoms) and 100 means asymptomatic or entirely recovered; 6. Ability to perform everyday activities in Days 0, 3, 7, 14 and 30 (0 = no loss of capacity and 100 = complete inability to perform any self -care or ever yday activity) ; 7. WHO C OVID Ordin al Outcomes Scale; 8. Brescia COVID-19 score; 9. Disease progression outcomes, including hospitalization, mechanical ventilation, and death. Full raw data is available at a data repository (https://osf.io/cm4f8/). Statistical analysis . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Nonparametric ANOVA (Kruskal -Wallis) was performed fo r all parameters, regardless of the distribution normality, and post-hoc adjusted Dunn’s test was performed for subgroup analyses, whenever p <0.05. All statistical tests were performed us ing XLSTAT (Microsoft, USA). Table 1. Parameters evaluated for the present prospective observational study. Aspect Parameter Baseline characteristics Age (y/o) Weight (kg) Height (m) BMI (kg/m2) Married (yes/no) and households (yes/no) Disease and treatment timing Time-to-treat (interval between beginning of symptoms and beginning of specific treatment) (days) Duration of positive rtPCR SARS-CoV-2 (days) Duration of symptoms (not including anosmia and ageusia) (days) Duration of symptoms (including anosmia and ageusia) (days) Medical history Existing diseases Hypertension Myocardial infarction Stroke Chronic heart failure Lipid disorders Type 2 diabetes mellitus (T2DM) Pre-diabetes Dysglicaemia (T2DM + pre-diabetes) Obesity Asthma Chronic Obstructive Pulmonary Disease (COPD) Chronic Kidney Disease (CKD) Liver fibrosis or cirrhosis Major depression Anxiety or anxiety-related disorders Attention deficiency and hyperactive disorders (ADHD) Insomnia Hypothyroidism Autoimmune disorders (any) Current or previous non-breast non-thyroid cancer Current or previous breast cancer Current or previous thyroid cancer Menopause Endometriosis Other diseases (any) Current medications . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Cardiovascular drugs Beta-blocker Angiotensin converter inhibitors (ACEi) (-pril) Angiotensin-2 receptor blockers (ARB) (-tan) Loop diuretics (furosemide) Thiazide diuretics (hydrochlorotiazide (HCTZ), indapamide) Calcium channel blockers (CCB) (-dipine) K-sparing diuretics (spironolactone) Statins (pitava-, rosuva-, atorva-, prava-, simvastatin) Other lipid-lowering agents (fibrates, ezetimibe, PCSK9 inhibitors) Aspirin Clopidogrel Warfarin Xa factor inhibitors (apibaxan, rivaroxaban) Direct thrombin inhibitors (dabagatran) Heparins Diabetes, obesity, and metabolic-related drugs Biguanides (metformin) Glucagon-like peptide 1 (GLP1) receptor analogues (GLP- 1Ra) (lira-, sema-, dulaglutide; exenatide) Sodium-glucose co-transporter 2 (SGLT2) inhibitors (SGLT2i) (empa-, dapa-, canagliflozin) Di-peptyl peptidase 4 (DPP4) inhibitors (DPP4i) (vilda-, sita- , saxa-, linagliptin) Sulfonylureas (glipizide, glimepiride, glicazide) Glitazone Acarbose Insulin Orlistat Hormone and related therapies Levothyroxine (with or without liothyronine) Oral contraceptives Hormonal replacement therapy (HRT) for menopause Other hormonal regimes Aromatase inhibitors (anastrozole; letrozole) Selective estrogen receptor modulators (SERMs) Central-acting drugs Hypnotics (zolpidem, zopiclone, eszopiclone, ramelteon) Selective serotonin reuptaker inhibitors (SSRIs) (fluoxetine, (des)venlafaxine, sertraline, (es)citalopram, vortioxetine, fluvoxamine) Other antidepressants and humor stabilizers (bupropion, topiramate, trazodone, ami- ou nortriptiline, topiramate, oxcarbamazepine) Benzodiazepines (Lora-, broma-, dia-, clonazepam; alpra-, midazolam) Atypical antipsychotics (olanzapine, quetiapine, risperidone, clozapine, aripiprazole) Central nervous system (CNS) stimulants (methylfenidate, lisdexamfetamine, modafinil) Other drugs Supplements Omega-3 (> 3g/day) Vitamina D (> 1,000iu/day) Zinc (> 15mg/day) Vitamin C (> 500mg/day) . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Multivitamins Vaccine BCG (lifetime) Influenza (in 2020) Pneumococcal 13 or 23 (since 2017) Lifestyle Current smoking (> 2 packs/week and > 10 pack-year) Regular physical activity (> 150min/week, moderate-to- vigorous, > 3 METs, for > 1y) Clinical characterization - Presence (yes/no) Time to appearance (days) Duration (days) Unspecific symptoms Fever “Feverish” Headache Shortness of breath Anosmia Ageusia / hypergeusia / dysgeusia Dizziness Weakness Fatigue Hyporexia / anorexia Thoracic pain Lower back pain Dry eyes / dry mouth / skin lesions Breast pain Upper respiratory tract infection-like symptoms Nasal congestion or rhinorrhea Dry cough “Sinusitis” (self-reported perception) “Sore throat” (self-reported perception) Dengue-like symptoms Myalgia Arthralgia Upper back pain Conjunctival hyperemia Pre-orbital pain Gastrointestinal (GI) infection-like symptoms Diarrhea Nauseas Vomiting Abdominal or pelvic pain Clinical clustering Anosmia-Ageusia dominance; or Dengue-like symptomatology / clinical presentation; or URTI-like symptomatology / clinical presentation; or GI infection-like symptomatology / clinical presentation; or Mixed; or Unspecific; or Asymptomatic Treatment (Azithromycin 500mg/day for 05 days +) Hydroxychloroquine 400mg/day for 05 days, or Nitazoxanide 500mg BID for 06 days, or Ivermectin 0.2mg/kg/day for 03 days +/- Spironolactone 100mg BID for 15 days, or Dutasteride 0.5mg/day until cure . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Additional treatments (added according to clinical judgement) Xa factor inhibitors Warfarin Enoxaparin Acetylsalicylic acid (ASA) Glucocorticoids (methylprednisolone, dexamethasone, prednisone, prednisolone) Bromhexine N-acetylcysteine Colchicine Vitamin C (additional dose, if already under use) Zinc (additional dose, if already under use) Vitamin D (additional dose, if already under use) Outcomes (Day 0 = beginning of treatment) WHO COVID Ordinal Outcomes (Stages 1-5) Day 0 Day 7 Day 14 Day 30 Day 60 Loss of ability to everyday activities (0-100; 0 = no loss; 100 = complete inability) Day 0 Day 3 Day 7 Day 14 Day 30 Symptom severity (0-100;0 = worst day of symptoms; 100 = no symptoms or fully recovered) Day -7 to -4 Day -3 to -1 Day 0 Day 1 Day 2 Day 3 Day 7 Day 14 Day 21 Day 30 Day 60 Thoracic CT scan (% of lungs affected) Day 0 Day 7 Day 14 Day 30 Disease progression outcomes Brescia COVID-19 Respiratory Severity Scale (0-4) Hospitalization Intensive Care Unit (ICU) Mechanical ventilation Noradrenaline/dopamine Death

. CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Tables 6 to 15 detail characteristics and parameters of overall females and for each group (non-HA, HA and HA-spiro), and overall and pairwise comparisons. Tables 2 to 5 depict baseline and medical chara cteristics, Tables 6 to 8 describe COVID-19 presentation, Tables 9 and 10 show the proposed pharmacological interventi ons for COVID-19, and Tables 11 to 15 depict COVID-19 clinical outcomes. Patients’ characterization In total, 270 females confirmed for COVID-19 were included. Of these, 195, 67, and eight were from the non-HA, HA, and HA-spiro groups. The dropout rate for clinical characterization and disease outcomes was zero. Baseline characteristics are described in Table 2. HA females were significantly younger, shorter, and heavier than non-HA. Table 2. Baseline characteristics. Baseline characteristics Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolacto ne Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Age (y/o) (p-value) 40.5 ± 11.5 42.5 ± 11.6 (p<0.0001 vs HA) (n/s(p=0.12) vs spiro) 35.2 ± 9.5 (n/s(p=0.54) vs spiro) 37.0 ± 7.1 (n/s(p=0.32) vs non-spiro) 40.6 ± 11.5 < 0.0001 Height (m) (p-value) 1.63 ± 0.06 1.64 ± 0.05 (p<0.0001 vs HA) (n/s(p=0.11) vs spiro) 1.60 ± 0.05 (n/s(p=0.62) vs spiro) 1.61 ± 0.04 (n/s(p=0.29) vs non-spiro) 1.63 ± 0.06 < 0.0001 BMI (kg/m2) (p-value) 24.4 ± 5.1 24.0 ± 4.8 (p=0.0008 vs HA) (n/s(p=0.28) vs spiro) 25.7 ± 5.9 (n/s(p=0.90) vs spiro) 25.4 ± 4.3 (n/s(p=0.40) vs non-spiro) 24.5 ± 5.2 0.003 Married (yes/no) 161 (59.6%) 105 (53.8%) 50 (74.6%) 6 (75%) 155 (59.2%) n/a Households (yes/no) 40 (14.8%) 34 (17.4%) 5 (7.5%) 1 (12.5%) 39 (14.9%) n/a HA = hyperandrogenic n/s = non-significant; n/a = non-applicable The major and prevailing diseases were present in similarly present in all groups (Table 3), while chronic k idney disease (CKD) was present in one patient , and liver fibrosis and cirrhosis, and current cancer were absent. Although HA had greater BMI than non-HA females, prevalence of obesity was similar between groups. . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Table 3. Comorbidities. Comorbidities Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Hypertension 38 (14.1%) 24 (12.3%) 12 (17.9%) 2 (25.0%) 36 (13.8%) 0.68 (n/s) Myocardial infarction 1 (0.4%) 1 (0.5%) 0 0 1 (0.4%) 0.99 (n/s) Stroke 1 (0.4%) 1 (0.5%) 0 0 1 (0.4%) 0.99 (n/s) Chronic Heart Failure (CHD) 0 0 0 0 0 n/a Lipid disorders 63 (23.3%) 43 (22.1%) 17 (25.4%) 3 (37.5%) 60 (22.9%) 0.72 (n/s) Type 2 diabetes mellitus (T2DM) 20 (7.4%) 10 (5.1%) 8 (11.9%) 0 18 (6.9%) 0.48 (n/s) Pre-diabetes 21 (7.8%) 11 (5.6%) 9 (13.4%) 2 (25.0%) 20 (7.6%) 0.62 (n/s) Dysglycemia (T2DM + pre- diabetes) 41 (15.3%) 21 (10.8%) 17 (25.4%) 1 (12.5%) 38 (14.5%) 0.79 (n/s) Obesity 52 (19.3%) 32 (16.4%) 18 (26.8%) 2 (25.0%) 50 (19.1%) 0.42 (n/s) Asthma 16 (5.9%) 12 (6.2%) 4 (6.0%) 0 16 (6.1%) 0.94 (n/s) Major depression 15 (5.6%) 11 (5.6%) 4 (6.0%) 0 15 (5.7%) 0.96 (n/s) Anxiety-related disorders 41 (15.2%) 22 (11.3%) 16 (23.9%) 3 (37.5%) 38 (14.5%) 0.17 (n/s) Attention deficiency and hyperactive disorders (ADHD) 14 (5.2%) 18 (9.4%) 7 (10.4%) 1 (12.5%) 13 (5.0%) 0.63 (n/s) Insomnia 15 (5.6%) 8 (4.1%) 6 (9.0%) 1 (12.5%) 14 (5.3%) 0.79 (n/s) Hypothyroidism 54 (20.0%) 38 (19.5%) 14 (20.9%) 2 (25.0%) 52 (19.8%) 0.96 (n/s) Autoimmune disorders 10 (3.7%) 9(4.6%) 1 (1.5%) 0 10 (3.8%) 0.38 (n/s) Menopause 65 (24.1%) 56 (28.7%)* 8 (11.9%) 1 (12.5%) 64 (24.4%) 0.10 (n/s) Endometriosis 29 (10.4%) 17 (8.7%) 11 (16.4%) 1 (12.5%) 28 (10.7%) 0.64 (n/s) Breast cancer 4 (1.5%) 3 (1.5%) 1 (1.5%) 0 4 (1.5%) 0.99 (n/s) HA = hyperandrogenic; n/s = non-significant; n/a = non-applicable *p=0.04 vs HA females Table 4 depicts the med ications used on a chronic and regular basis . Metformin and oral contraceptives were used by significant ly larger n umber of HA than non-HA females. None of the other drugs for metabolic, cardiovascular, psychiatric or hormonal diseases disclosed differences between groups. Warfarin, direct thrombin inhibitors and heparin were not used by any patient. The percentage of participants with vaccines for . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint BCG, influenza and pneumococca l were similar between groups , as well as practice of regular physical (Table 5). Table 4. Medications used. Current medications Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Beta-blocker 3 (1.1%) 3 (1.5%) 0 0 3 (1.1%) > 0.9 (n/s) Angiotensin converter inhibitors (ACEi) 8 (3.0%) 5 (2.6%) 2 (3.0%) 1 (12.5%) 7 (2.7%) > 0.9 (n/s) Angiotensin-2 receptor blockers (ARB) 32 (11.8%) 19 (9.7%) 12 (17.9%) 1 (12.5%) 31 (11.8%) 0.61 (n/s) Loop diuretics 3 (1.1%) 2 (1.0%) 0 1 (12.5%) 2 (0.8%) n/a Thiazide diuretics 10 (3.7%) 8 (4.1%) 2(3.0%) 0 10 (3.8%) > 0.9 (n/s) Calcium channel blockers (CCB) 16 (5.9%) 10 (5.1%) 6 (8.9%) 0 16 (6.1%) 0.86 (n/s) Statins 51 (18.9%) 33 (16.9%) 17 (25.4%) 1 (12.5%) 50 (19.1%) 0.56 (n/s) Aspirin 1 (0.4%) 1 (0.5%) 0 0 1 (0.4%) n/a Clopidogrel 1 (0.4%) 1 (0.5%) 0 0 1 (0.4%) n/a Xa factor inhibitors 3 (1.1%) 2 (1.0%) 0 1 (12.5%) 2 (0.8%) n/a Metformin 51 (18.9%) 26 (13.3%)* 22 (32.8%) 3(37.5%) 48 (18.3%) 0.038 Glucagon-like peptide 1 (GLP1) receptor analogues (GLP-1Ra) 20 (7.4%) 9 (4.6%) 10 (14.9%) 1 (12.5%) 19 (7.2%) 0.44 (n/s) Sodium-glucose co- transporter 2 (SGLT2) inhibitors (SGLT2i) 30 (11.1%) 16 (8.2%) 12 (17.9%) 2 (25.0%) 28 (10.7%) 0.39 (n/s) Di-peptyl peptidase 4 (DPP4) inhibitors (DPP4i) 6 (2.2%) 4 (2.0%) 1 (1.5%) 1 (12.5%) 5 (1.9%) 0.98 (n/s) Sulfonylureas 0 0 0 0 0 n/a Pioglitazone 0 0 0 0 0 n/a Insulin 1 (0.4%) 1 (0.5%) 0 0 1 (0.4%) n/a Orlistat 12 (4.4%) 2 (1.0%) 10 (14.9%) 0 12 (4.6%) 0.23 (n/s) Levothyroxine 54 (20.0%) 38 (19.5%) 14 (20.9%) 2 (25.0%) 52 (19.8%) 0.84 (n/s) Liothyronine 3 (1.1%) 3 (1.5%) 0 0 3 (1.1%) 0.77 (n/s) Estradiol (E; no P) 7 (2.6%) 6 (3.1%) 1 (1.5%) 0 7 (2.7%) 0.45 (n/s) Progesterone (P; no E) 0 0 0 0 0 n/a . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Combined therapy (E+P) (for menopause) 29 (10.7%) 24 (12.3%) 5 (7.5%) 0 29 (11.1%) 0.73 (n/s) Oral contraceptives 36 (13.3%) 17 (8.7%)* 19 (28.4%) 0 36 (13.7%) 0.045 Other hormonal regimens 0 0 0 0 0 n/a Aromatase inhibitors (anastrozole; letrozole) or Selective estrogen receptor modulators (SERMs) 3 (1.1%) 1 (0.5%) 2 (3.0%) 0 3 (1.1%) 0.85 (n/s) Hypnotics 19 (7.0%) 10 (5.1%) 8 (11.9%) 1 (12.5%) 18 (6.9%) 0.68 (n/s) Selective serotonin reuptaker inhibitors (SSRIs) 43 (15.9%) 27 (13.8%) 13 (19.4%) 3 (37.5%) 40 (15.3%) 0.77 (n/s) Other antidepressants and humor stabilizers 27 (10.0%) 15 (7.7%) 11 (16.4%) 1 (12.5%) 26 (10.0%) 0.57 (n/s) Benzodiazepines 3 (1.1%) 2 (1.0%) 1 (1.4%) 0 3 (1.1%) 0.99 (n/s) Atypical antipsychotics 7 (2.6%) 4 (2.0%) 3 (4.5%) 0 7 (2.7%) 0.91 (n/s) Central nervous system (CNS) stimulants 18 (6.7%) 8 (4.1%) 8 (11.9%) 2 (25.0%) 16 (6.1%) 0.42 (n/s) Finasteride 4 (1.5%) 0 4 (6.0%) 0 4 (1.5%) 0.76 (n/s) Oral minoxidil 10 (3.7%) 4 (2.0%) 6 (8.9%) 0 10 (3.8%) 0.69 (n/s) Omega 3 3 (1.1%) 2 (1.0%) 1 (1.5%) 0 3 (1.1%) 0.99 (n/s) Vitamin D 33 (11.1%) 21 (10.8%) 9 (13.4%) 3 (37.5%) 30 (11.4%) 0.44 (n/s) Zinc 20 (7.4%) 12 (6.1%) 6 (8.9%) 2 (25.0%) 18 (6.9%) 0.78 (n/s) Biotin 16 (5.9%) 9 (4.6%) 7 (10.4%) 0 16 (6.1%) 0.61 (n/s) Vitamin C 28 (10.4%) 18 (9.2%) 8 (11.9%) 2 (25.0%) 26 (9.9%) 0.63 (n/s) HA = hyperandrogenic; n/s = non-significant; n/a = non-applicable *p=0.04 vs HA females Table 5. Vaccines and lifestyle. Vaccine and lifestyle Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Vaccine - BCG 270 (100%) 195 (100%) 67 (100%) 8 (100%) 262 (100%) 1.0 (n/s) Vaccine – Influenza (2020) 46 (17.0%) 34 (17.4%) 11 (15.5%) 7 (13.5%) 43 (16.4%) 0.95 (n/s) Vaccine – Pneumococcal (since 2017) 37 (13.7%) 27 (13.8%) 10 (14.1%) 6 (11.5%) 35 (13.5%) 0.97 (n/s) Current smoking 2 (0.7%) 2 (1.0%) 0 0 2 (0.8%) 0.98 (n/s) Regular physical activity 76 (28.1%) 54 (27.7%) 20 (28.2%) 2 (25.0%) 71 (27.1%) 0.99 (n/s) HA = hyperandrogenic; n/s = non-significant . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint COVID-19 clinical presentation Table 6 descr ibes COVID-19 clusters of clinical presentation . URT I-like syndrome was statistically more prevalent in HA than non -HA, while anosmia -ageusia predominance, dengu e fever -like, GI infection -like, mi xed and unspecific symptomatology were similar between groups. Spironolactone users were more likely to be asymptomatic than non-users during COVID-19. Table 6. Clinical clustering Clinical clustering Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Anosmia-Ageusia dominance (p-value) 43 (15.9%) 34 (17.4%) (n/s (p=0.50) vs HA) (n/s (p=0.83) vs spiro) 8 (11.9%) (n/s(p=0.98) vs spiro) 1 (12.5%) (n/s (p=0.86) vs no-spiro) 42 (16.0%) 0.79 (n/s) Dengue fever-like (p-value) 77 (28.5%) 65 (33.3%) (n/s (p=0.67) vs HA) (n/s (p=0.32) vs spiro) 20 (29.8%) (n/s(p=0.43) vs spiro) 1 (12.5%) (n/s (p=0.34) vs no-spiro) 85 (32.4%) 0.58 (n/s) URTI-like (p-value) 87 (32.2%) 53 (27.2%) (p=0.012) vs HA) (n/s (p=0.92) vs spiro) 32 (47.8%) (n/s(p=0.29) vs spiro) 2 (25.0%) (n/s (p=0.72) vs no-spiro) 85 (32.4%) 0.04 GI infection-like (p-value) 33 (12.2%) 30 (15.4%) (n/s (p=0.97) vs HA) (n/s (p=0.85) vs spiro) 11 (16.4%) (n/s(p=0.89) vs spiro) 1 (12.5%) (n/s (p=0.88) vs no-spiro) 41 (15.6%) 0.98 (n/s) Mixed (p-value) 35 (12.6%) 22 (11.3%) (n/s (p=0.72) vs HA) (n/s (p=0.62) vs spiro) 12 (17.9%) (n/s(p=0.75) vs spiro) 0 (n/s (p=0.53) vs no-spiro) 34 (13.0%) 0.59 (n/s) Unspecific (p-value) 65 (24.1%) 49 (25.1%) (n/s (p=0.88) vs HA) (n/s (p=0.23) vs spiro) 16 (23.9%) (n/s(p=0.27) vs spiro) 0 (n/s (p=0.23) vs no-spiro) 65 (24.8%) 0.48 (n/s) Asymptomatic (p-value) 27 (10.0%) 21 (10.8%) ((n/s (p=0.34) vs HA) (n/s (p=0.06) vs spiro) 2 (3.0%) (p=0.03) vs spiro) 4 (50.0%) (p=0.047) vs no-spiro) 23 (8.8%) 0.089 (n/s) URTI = Upper respiratory tract infection; GI = Gastrointestinal; HA = hyperandrogenic; n/s = non- significant Table 7 describes the percentage of females presenting each symptom, its average duration, and time -to-appearance. Prevailing symptoms include anosmia ( 71.1%) and ageusia (67.0%) and headache (48.1%). Symptoms present in 20% to 40% of COVID-19 females include myalgia (37.4%), dry cough (36.3%), fever or “feverish” (34.1%), nasal congestion or rhinorrhea (34.1%), fatigue (3 3.3%), weakness (29.5%), hyporexia (27.8%), thoracic pain ( 24.8%), diarrhea (2 4.1%) and dizziness ( 21.5%). Symptoms . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint present in fewer than 20% o f patients incl ude “sore throat” ( 15.9%), pre-orbital pain (12.6%), arthralgia (10.4%), conjunctival hyperemia (8.1%), nauseas (8.1%), abdominal pain (7.8%), upper back pain (7.7%), “sinusitis” (6.7%), shortness of breath (5.9%), lower back pain (5.2%), and pre-orbital pain (3.5%), dry eyes (2.2%) and dry mouth (1.1%). COVID-19 has shown the fol lowing sequence of symptoms, in a chro nological order: 1. Earlier symptoms (< 2.0 days from the first symptom; in days): dizziness (1.0 ± 0.2), abdominal pain (1.1 ± 0.3); conjunctival hyperemia (1.1 ± 0.5), nasal congestion or rhinorrhea (1.2 ± 0.5), headache (1.2 ± 0.5), dry cough (1.2 ± 0.5), myalgia (1.2 ± 0.4), nauseas (1.3 ± 0.5), weakness (1.3 ± 0.5), “sinusitis” (1.3 ± 0.5), hiporexia (1.3 ± 0. 5), fatigue (1.4 ± 0.7), pre-orbital pain (1.4 ± 0.6), “sore throat” (1.4 ± 0.7), “feverish” (1.5 ± 0.6, fever (1.5 ± 0.6), arthralgia (1.5 ± 0.6) and diarrhea (1.5 ± 0.8); and 2. L ater symptoms (> 2.0 days since first symptom) include upper back pain (2.0 ± 0.8); lower back pain (2.1 ± 0.7); thoracic pain (2.5 ± 1.0); anosmia (3.4 ± 1.3), ageusia (3.4 ± 1.3), shortness of breath (3.8 ± 1.2). Duration of symptoms during COVID-19 ranged from those with less than three days, including dizziness (1.9 ± 1.0), abdominal pain (1.8 ± 0.7), nauseas (1.7 ± 0.9), fever (2.3 ± 0.8); “feverish” (2.4 ± 1.2), shortness of breath (2.5 ± 1.6), weakness (2.7 ± 1.4) and pre-orbital pain (2.8 ± 1.3); of inter mediate duration (3 to 5 days) , including arthralgia (3.0 ± 2.4), diarrhea (3.2 ± 1.5), myalgia (3.3 ± 1.4), lower back pain (3.3 ± 1.5), nasal congestion or rhinorrhea (3.6 ± 1.8), upper back pain (4.3 ± 2.1), thoracic pain (4.7 ± 2.5), hyporexia (4.8 ± 2.8), and conjunctival hyperemia (4.9 ± 2.6); and those with prolonged duration (> 5 days), including dry cough (5.4 ± 3.5) “sore throat” (5.5 ± 1.9), headache (5.7 ± 3.5), “sinusitis” (6.0 ± 2.6), dry eyes (6.2 ± 2.1), ageusia (7.0 ± 5.7), dry mouth (7.0 ± 0.0), fatigue (7.2 ± 4.9) and anosmia (7.9 ± 6.2). Similar time-to-appearance and duration : fever, “feverish”, nasal congestion or rhinorrhea, headache, shortness of breath, “sinusitis”, “sore throat”, dizziness, thoracic pain, lower back pain, diarrhea, nauseas, abdominal pain, pre-orbital pain, and dry eyes and mouth. Although conjunctival hyperemia was almost five times more present in HA than non-HA females (in terms of percentage), it did not reach statistical significance. . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Anosmia and ageusia were almost absent in spironolactone users. Fewer non-HA than HA females were affected by anosmia, ageusia, dry cough, fatigue, w eakness and hyporexia. Ageusia, weakness and myalgia lasted shorter in non-HA than HA. Time-to-appearance was shorter in non-HA for upper back pain. Table 7. Clinical manifestations in COVID-19: presence (%), time-to-appearance and duration. Clinical manifestations Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non- HA + HA) (n = 262) p-value (overall) Fever Presence (%) (p-value) 31 (11.5%) 19 (9.7%) (n/s (p=0.32) vs HA) (n/s (p=0.64) vs spiro) 12 (17.9%) (n/s(p=0.41) vs spiro) 0 (n/s(p=0.57) vs no-spiro) 31 (11.8%) 0.52 Time to appearance (days) (p-value) 1.5 ± 0.6 1.7 ± 0.6 (n/s (p=0.19) vs HA) (n/a vs spiro) 1.3 ± 0.5 (n/a vs spiro) - (n/a) 1.5 ± 0.6 n/a Duration (days) (p-value) 2.3 ± 0.8 2.4 ± 0.7 (n/s (p=0.48) vs HA) (n/a vs spiro) 2.2 ± 1.0 (n/a vs spiro) - (n/a) 2.2 ± 0.8 n/a “Feverish” Presence (%) (p-value) 61 (22.6%) 39 (20.0%) (n/s (p=0.12) vs HA) (n/s (p=0.38) vs spiro) 22 (32.8%) (n/s(p=0.13) vs spiro) 0 ((n/s(p=0.26) vs no-spiro) 61 (23.3%) 0.16 Time to appearance (days) (p-value) 1.5 ± 0.6 1.4 ± 0.5 (n/s (p=0.87) vs HA) (n/a vs spiro) 1.4 ± 0.7 (n/a vs spiro) 0 (n/a vs no-spiro) 1.4 ± 0.6 n/a Duration (days) (p-value) 2.4 ± 1.2 2.4 ± 1.0 (n/s (p=0.76) vs HA) (n/a vs spiro) 2.4 ± 1.4 (n/a vs spiro) 0 (n/a vs no-spiro) 2.4 ± 1.2 n/a Nasal congestion or rhinorrhea Presence (%) (p-value) 92 (34.1%) 63 (32.3%) (n/s (p=0.33) vs HA) (n/s (p=0.73) vs spiro) 27 (40.3%) (n/s(p=0.48) vs spiro) 2 (25.0%) ((n/s(p=0.65) vs no-spiro) 90 (34.5%) 0.56 Time to appearance (days) (p-value) 1.2 ± 0.5 1.2 ± 0.5 (n/s (p=0.69) vs HA) (n/a vs spiro) 1.2 ± 0.4 ((n/a vs spiro) 1.0 ± 0.0 (1;1) (n/a vs no- spiro) 1.2 ± 0.5 n/a Duration (days) (p-value) 3.6 ± 1.8 3.8 ± 1.8 (n/s (p=0.72) vs HA) (n/a vs spiro) 3.3 ± 1.5 (n/a vs spiro) 1.5 ± 0.5 (1;2) (n/a vs no-spiro) 3.6 ± 1.8 n/a Headache Presence (%) (p-value) 130 (48.1%) 89 (46.6%) (n/s (p=0.12) vs HA) (n/s (p=0.33) vs spiro) 39 (58.2%) (n/s(p=0.13) vs spiro) 2 (25.0%) ((n/s(p=0.15) vs no-spiro) 128 (48.8%) 0.16 Time to appearance (days) (p-value) 1.2 ± 0.5 1.1 ± 0.3 (n/s (p=0.10) vs HA) (n/a vs spiro) 1.4 ± 0.7 (n/a vs spiro) 1.5 ± 0.5 (1;2) (n/a vs no-spiro) 1.2 ± 0.5 n/a Duration (days) 5.7 ± 3.5 5.6 ± 3.5 6.0 ± 3.6 2.0 ± 1.0 (1;3) 5.8 ± 3.5 n/a . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint (p-value) (n/s (p=0.63) vs HA) (n/a vs spiro) (n/a vs spiro) (n/a vs no-spiro) Shortness of breath Presence (%) (p-value) 16 (5.9%) 10 (5.1%) (n/s (p=0.72) vs HA) (n/s (p=0.64) vs spiro) 6 (8.9%) (n/s(p=0.68) vs spiro) 0 ((n/s(p=0.79) vs no-spiro) 16 (6.1%) 0.85 Time to appearance (days) (p-value) 3.8 ± 1.2 3.5 ± 1.4 (n/s (p=0.28) vs HA) (n/a vs no-spiro) 4.2 ± 0.7 (n/a vs spiro) 0 (n/a vs no-spiro) 3.7 ± 1.2 n/a Duration (days) (p-value) 2.5 ± 1.6 2.4 ± 1.9 (n/s (p=0.72) vs HA) (n/a vs spiro) 2.7 ± 0.7 (n/a vs spiro) 0 (n/a vs no-spiro) 2.5 ± 1.6 n/a Anosmia Presence (%) (p-value) 192 (71.1%) 133 (68.2%) (p = 0.025 vs HA) (p = 0.0008 vs spiro) 58 (86.6%) (p = 0.0007 vs AGA-5ARi) 1 (12.5%) (p = 0.0036 vs no-5ARi) 191 (72.9%) 0.001 Time to appearance (days) (p-value) 3.4 ± 1.3 3.2 ± 1.2 (p=0.0009 vs HA) (n/a vs spiro) 3.9 ± 1.2 (n/a vs spiro) 3 (n/a vs no-spiro) 3.4 ± 1.3 < 0.0001 Duration (days) (p-value) 7.9 ± 6.2 7.8 ± 6.2 (n/s(p=0.63) vs HA) (n/a vs spiro) 8.1 ± 6.0 (n/a vs spiro) 3 (n/a vs no-spiro) 7.9 ± 6.2 < 0.0001 Ageusia Presence (%) (p-value) 181 (67.0%) 124 (63.6%) (p=0.015 vs HA) (p=0.014 vs spiro) 56 (83.6%) (p=0.001 vs spiro) 1 (12.5%) (p=0.007 vs no-5ARi) 180 (69.7%) 0.0013 Time to appearance (days) (p-value) 3.4 ± 1.3 3.3 ± 1.4 (p=0.011 vs HA) (n/a vs spiro) 3.8 ± 1.2 (n/a vs spiro) 1 (n/a vs no-spiro) 3.5 ± 1.3 n/a Duration (days) (p-value) 7.0 ± 5.7 6.7 ± 5.6 (p=0.045 vs HA) (n/a vs spiro) 7.9 ± 5.7.6 (n/a vs spiro) 3 (n/a vs no-spiro) 7.1 ± 5.7 n/a Dry cough Presence (%) (p-value) 98 (36.3%) 63 (32.3%) (p=0.039 vs HA) (n/s(p=0.73). vs spiro) 33 (49.2%) (n/s (p=0.26) vs spiro) 2 (25.0%) (n/s (p=0.56) vs no-spiro) 96 (36.6%) 0.10 (n/s) Time to appearance (days) (p-value) 1.2 ± 0.5 1.2 ± 0.5 (n/s (p=0.99) vs HA) (n/a vs spiro) 1.2 ± 0.5 (n/a vs spiro) 1.0 ± 0.0 (1;1) (n/a vs no-spiro) 1.2 ± 0.5 n/a Duration (days) (p-value) 5.4 ± 3.5 5.5 ± 3.6 (n/s (p=0.80) vs HA) (n/a vs spiro) 5.5 ± 3.2 (n/a vs spiro) 2.5 ± 0.5 (2;3) (n/a vs no-spiro) 5.5 ± 3.5 n/a “Sinusitis” Presence (%) (p-value) 18 (6.7%) 18 (9.2%) (n/s(p=0.83) vs HA) (n/s(p=0.66) vs spiro) 0 (n/s(p=0.73) vs spiro) 0 (n/s(p=0.67) vs no-spiro) 18 (6.9%) 0.89 (n/s) Time to appearance (days) (p-value) 1.3 ± 0.5 1.3 ± 0.5 (n/s(p=0.82) vs HA) (n/a vs spiro) 0 (n/a vs spiro) 0 (n/a vs no-spiro) 1.3 ± 0.5 n/a Duration (days) (p-value) 6.0 ± 2.6 6.0 ± 2.8 (n/s(p=0.68) vs HA) (n/a vs spiro) 0 (n/a vs spiro) 0 (n/a vs no-spiro) 6.0 ± 2.6 n/a “Sore throat” Presence (%) (p-value) 43 (15.9%) 25 (12.8%) (n/s (p=0.086) vs HA) (n/s(p=0.54) vs spiro) 18 (26.9%) (n/s(p=0.22) vs spiro) 0 (n/s(p=0.43) vs no-spiro) 43 (16.4%) 0.17 (n/s) Time to appearance (days) 1.4 ± 0.7 1.4 ± 0.6 (n/s (p=0.87) vs HA) 1.4 ± 1.2 (n/a vs spiro) 0 (n/a vs no-spiro) 1.4 ± 0.7 n/a . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint (p-value) (n/a vs spiro) Duration (days) (p-value) 5.6 ± 2.0 5.9 ± 2.1 ((n/s (p=0.41) vs HA) (n/a vs spiro) 5.3 ± 1.7 (n/a vs spiro) 0 (n/a vs no-spiro) 5.6 ± 2.0 n/a Dizziness Presence (%) (p-value) 58 (21.5%) 40 (20.5%) (n/s (p=0.44) vs HA) (n/s(p=0.33) vs spiro) 18 (26.9%) (n/s(p=0.22) vs spiro) 0 (n/s(p=0.29) vs no-spiro) 58 (22.1%) 0.42 (n/s) Time to appearance (days) (p-value) 1.0 ± 0.2 1.0 ± 0.0 (n/s (p=0.50) vs HA) (n/a vs spiro) 1.1 ± 0.3 (n/a vs spiro) 0 (n/a vs no-spiro) 1.0± 0.2 n/a Duration (days) (p-value) 1.9 ± 1.0 1.8 ± 0.9 (n/s (p=0.86) vs HA) (n/a vs spiro) 1.9 ± 1.1 (n/a vs spiro) 0 (n/a vs no-spiro) 1.9 ± 1.0 n/a Fatigue Presence (%) (p-value) 90 (33.3%) 54 (27.7%) (p=00015 vs HA). ((n/s(p=0.18) vs spiro) 36 (53.7%) (p=0.013 vs spiro) 0 ((n/s(p=0.098) vs no-spiro) 90 (34.3%) 0.0016 Time to appearance (days) (p-value) 1.4 ± 0.7 1.4 ± 0.6 ((n/s(p=0.74) vs HA) (n/a vs spiro) 1.5 ± 0.7 (n/a vs spiro) 0 (n/a vs no-spiro) 1.4 ± 0.7 n/a Duration (days) (p-value) 7.2 ± 4.9 6.8 ± 4.5 ((n/s(p=0.35) vs HA) (n/a vs spiro) 7.7 ± 5.4 (n/a vs spiro) 0 (n/a vs no-spiro) 7.2 ± 4.9 n/a Weakness Presence (%) (p-value) 74 (29.5%) 47 (24.1%) (p=0.048 vs HA) (n/s(p=0.25) vs vs spiro) 27 (40.3%) (n/s(p=0.063) vs vs spiro) 0 (n/s(p=0.17) vs vs no- spiro) 74 (28.2%) 0.056 (n/s) Time to appearance (days) (p-value) 1.3 ± 0.5 1.3 ± 0.5 (n/s(p=0.93) vs HA) (n/s(p=0.056) vs spiro) 1.3 ± 0.5 (n/a vs spiro) 0 (n/a vs no-spiro) 1.3 ± 0.5 n/a Duration (days) (p-value) 2.7 ± 1.4 2.4 ± 1.1 (p=0.031 vs HA) (n/a vs spiro) 3.2 ± 1.7 (n/a vs spiro) 0 (n/a vs no-spiro) 2.7 ± 1.4 n/a Myalgia Presence (%) (p-value) 101 (37.4%) 73 (37.4%) (n/s (0.59) vs HA) (n/s (p=0.073) vs spiro) 28 (41.8%) (n/s (p=0.054) vs spiro) 0 (n/s (p=0.063) vs no-spiro) 101 (38.4%) 0.15 (n/s) Time to appearance (days) (p-value) 1.2 ± 0.4 1.2 ± 0.4 (n/s (0.25) vs HA) (n/a vs spiro) 1.3 ± 0.5 (n/a vs spiro) 0 (n/a vs vs no-spiro) 1.2 ± 0.4 n/a Duration (days) (p-value) 3.3 ± 1.4 3.1 ± 1.4 (p = 0.037 vs HA) (n/a vs spiro) 3.7 ± 1.4 (n/a vs spiro) 0 (n/a vs no-spiro) 3.3 ± 1.4 n/a Arthralgia Presence (%) (p-value) 28 (10.4%) 22 (11.3%) (n/s (p = 0.78) vs HA) (n/s (p = 0.59) vs spiro) 6 (8.9%) (n/s (p = 0.68) vs spiro) 0 (n/s (p = 0.61) vs no-spiro) 28 (10.7%) 0.84 (n/s) Time to appearance (days) (p-value) 1.5 ± 0.6 1.4 ± 0.6 (n/s (p = 0.74) vs HA) (n/a vs spiro) 1.5 ± 0.5 (n/a vs spiro) 0 (n/a vs no-spiro) 1.5 ± 0.6 n/a Duration (days) (p-value) 3.0 ± 2.4 3.1 ± 2.4 (n/s (p = 0.67) vs HA) (n/a vs spiro) 2.7 ± 2.0 (n/a vs spiro) 0 (n/a vs no-spiro) 3.0 ± 2.4 n/a Hyporexia Presence (%) 75 (27.8%) 45 (23.1%) (p=0.008 vs HA) 30 (44.8%) (p=0.039 vs spiro) 0 (n/s(p=0.17) vs no-spiro) 75 (28.6%) 0.011 . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint (p-value) (n/s(p=0.27) vs spiro) Time to appearance (days) (p-value) 1.3 ± 0.5 1.2 ± 0.5 (n/s(p=0.17) vs HA) (n/a vs spiro) 1.4 ± 0.6 (n/a vs spiro) 0 (n/a vs no-spiro) 1.3 ± 0.5 n/a Duration (days) (p-value) 4.8 ± 2.8 4.9 ± 3.1 (n/s(p=0.99) vs HA) (n/a vs spiro) 4.6 ± 2.2 (n/a vs spiro) 0 (n/a vs no-spiro) 4.8 ± 2.8 n/a Thoracic pain Presence (%) (p-value) 67 (24.8%) 43 (22.0%) (n/s(p=0.093) vs HA) (n/s(p=0.099) vs spiro) 24 (35.8%) (n/s(p=0.22) vs spiro) 0 (n/s(p=0.24) vs no-spiro) 67 (25.6%) 0.11 (n/s) Time to appearance (days) (p-value) 2.5 ± 1.0 2.6 ± 0.9 (n/s(p=0.61) vs HA) (n/a vs spiro) 2.5 ± 1.1 (n/a vs spiro) 0 (n/a vs no-spiro) 2.6 ± 1.0 n/a Duration (days) (p-value) 4.7 ± 2.5 4.5 ± 2.4 (n/s(p=0.31) vs HA) (n/a vs spiro) 5.1 ± 2.7 (n/a vs spiro) 0 (n/a vs no-spiro) 4.7 ± 2.5 n/a Upper back pain Presence (%) (p-value) 28 (10.4%) 15 (7.7%) (n/s(p=0.15) vs HA) (n/s (p=0.71) vs spiro) 13 (19.4%) (n/s (p=0.37 vs spiro) 0 (n/s (p=0.61 vs no-spiro) 28 (10.7%) 0.31 (n/s) Time to appearance (days) (p-value) 2.0 ± 0.8 1.6 ± 0.5 (p=0.006 vs HA) (n/a vs spiro) 2.5 ± 0.7 (n/a vs spiro) 0 (n/a vs no-spiro) 2.0 ± 0.8 n/a Duration (days) (p-value) 4.3 ± 2.1 4.3 ± 1.4 (n/s (p=0.33) vs HA) (n/a vs spiro) 4.2 ± 2.7 (n/a vs spiro) 0 (n/a vs no-spiro) 4.3 ± 2.1 n/a Lower back pain Presence (%) (p-value) 14 (5.2%) 10 (5.1%) (n/s (p=0.99) vs HA) (n/s(p=0.81) vs spiro) 4 (6.0%) (n/s (p=0.78) vs spiro) 0 (n/s (p=0.80) vs no-spiro) 14 (5.3%) 0.96 (n/s) Time to appearance (days) (p-value) 2.1 ± 0.7 1.8 ± 0.6 (n/a vs HA) (n/a vs spiro) 2.7 ± 0.4 (n/a vs spiro) 0 (n/a vs no-spiro) 2 ± 0.8 n/a Duration (days) (p-value) 3.3 ± 1.5 2.9 ± 1.2 (n/a vs HA) (n/a vs spiro) 4.2 ± 1.8 (n/a vs spiro) 0 (n/a vs no-spiro) 4.3 ± 2.1 n/a Diarrhea Presence (%) (p-value) 65 (24.1%) 41 (21.0%) (n/s (p=0.15) vs HA) (ns (p=0.85) vs spiro) 22 (32.8%) (n/s (p=0.72) vs spiro) 2 (25.0%) (n/s (p=0.96) vs no-spiro) 63 (24.0%) 0.35 (n/s) Time to appearance (days) (p-value) 1.5 ± 0.8 1.4 ± 0.7 (n/s (p=0.58) vs HA) (n/a vs spiro) 1.5 ± 0.8 (n/a vs spiro) 1.5 ± 0.5 (1;2) (n/a vs no-spiro) 1.5 ± 0.8 n/a Duration (days) (p-value) 3.2 ± 1.5 3.1 ± 1.3 (n/s (p=0.41) vs HA) (n/a vs spiro) 3.5 ± 1.7 (n/a vs spiro) 1.5 ± 0.5 (1;2) (n/a vs no-spiro) 3.2 ± 1.5 n/a Nauseas Presence (%) (p-value) 22 (8.1%) 13 (6.7%) (n/s (p=0.65) vs HA) (n/s (p=0.38) vs spiro) 7 (10.4%) (n/s (p=0.50) vs spiro) 2 (25.0%) (n/s (p=0.40) vs no-spiro) 20 (7.6%) 0.63 (n/s) Time to appearance (days) (p-value) 1.3 ± 0.5 1.2 ± 0.6 (n/s (p=0.69) vs HA) (n/a vs spiro) 1.3 ± 0.4 (n/a vs spiro) 1.5 ± 0.5 (1;2) (n/a vs no-spiro) 1.2 ± 0.5 n/a Duration (days) (p-value) 1.7 ± 0.9 1.6 ± 0.7 (n/s (p=0.53) vs HA) (n/a vs spiro) 2.0 ± 1.1 (n/a vs spiro) 1.0 ± 0.0 (1;1) (n/a vs no-spiro) 1.7 ± 0.9 n/a Vomiting . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Presence (%) (p-value) 2 (0.7%) 1 (n/a vs HA) (n/a vs spiro) 1 (1.5%) (n/a vs spiro) 0 (n/a vs no-spiro) 2 (0.8%) 1.0 (n/s) Time to appearance (days) (p-value) 2.0 ± 1.0 1 (n/a vs HA) (n/a vs spiro) 3 (n/a vs spiro) 0 (n/a vs no-spiro) (1;3) n/a Duration (days) (p-value) 2.0 ± 1.0 3 (n/a vs HA) (n/a vs spiro) 1 (n/a vs spiro) 0 (n/a vs no-spiro) (1;3) n/a Abdominal pain Presence (%) (p-value) 21 (7.8%) 13 (6.7%) (n/s (p=0.52) vs HA) (n/s (p=0.75) vs spiro) 8 (11.9%) (n/s(p=0.58) vs spiro) 0 (n/s (p=0.70) vs no-spiro) 21 (8.0%) 0.75 Time to appearance (days) (p-value) 1.1 ± 0.3 1.1 ± 0.3 (n/s (p=1.0) vs HA) (n/a vs spiro) 1.1 ± 0.3 (n/a vs spiro) 0 (n/a vs no-spiro) 1.1 ± 0.3 n/a Duration (days) (p-value) 1.8 ± 0.7 1.7 ± 0.6 (n/s (p=0.23) vs HA) (n/a vs spiro) 2.1 ± 0.8 (n/a vs spiro) 0 (n/a vs no-spiro) 1.8 ± 0.7 n/a Conjunctival hyperemia Presence (%) (p-value) 22 (8.1%) 9 (4.6%) (n/s(p=0.071) vs HA) (n/s (p=0.82) vs spiro) 13 (19.4%) (n/s (p=0.37) vs spiro) 0 (n/s (p=0.69) vs no-spiro) 22 (8.4%) 0.18 Time to appearance (days) (p-value) 1.1 ± 0.5 1.2 ± 0.6 (n/s (p=0.88) vs HA) (n/a vs spiro) 1.1 ± 0.3 (n/a vs spiro) 0 (n/a vs no-spiro) 1.1 ±0.5 n/a Duration (days) (p-value) 4.9 ± 2.6 4.2 ± 0.4 (n/s (p=0.54) vs HA) (n/a vs spiro) 5.5 ± 3.3 (n/a vs spiro) 0 (n/a vs vs no-spiro) 5.0 ± 2.6 n/a Pre-orbital pain Presence (%) (p-value) 34 (12.6%) 23 (11.8%) (n/s (p=0.57) vs HA) (n/s (p=0.57) vs spiro) 11 (16.4%) (n/s (p=0.78) vs spiro) 0 (n/s (p=0.53) vs no-spiro) 34 (13.0%) 0.70 (n/s) Time to appearance (days) (p-value) 1.4 ± 0.6 1.3 ± 0.6 (n/s (p=0.27) vs HA) (n/a vs spiro) 1.5 ± 0.5 (n/a vs spiro) 0 (n/a vs no-spiro) 1.4 ± 0.6 n/a Duration (days) (p-value) 2.8 ± 1.3 2.4 ± 1.3 (p=0.016) vs HA) (n/a vs spiro) 3.5 ± 1.1 (n/a vs spiro) 0 (n/a vs no-spiro) 2.8 ± 1.3 n/a Dry eyes Presence (%) (p-value) 6 (2.2%) 5 (2.6%) (n/s (p=0.99) vs HA) (n/s vs spiro) 1 (1.5%) (n/s vs spiro) 0 (n/s vs no-spiro) 6 (2.3%) 0.98 (n/s) Time to appearance (days) (p-value) 1.4 ± 0.8 1.0 ± 0.0 (n/a vs HA) (n/a vs spiro) 3 (n/a vs spiro) 0 (n/a vs no-spiro) 1.4 ± 0.8 n/a Duration (days) (p-value) 6.2 ± 2.1 5.2 ± 1.1 (n/a vs HA) (n/a vs spiro) 10 (n/a vs spiro) 0 (n/a vs no-spiro) 6.2 ± 2.1 n/a Dry mouth Presence (%) (p-value) 3 (1.1%) 3 (1.5%) (n/s (p=0.96) vs HA) (n/s vs spiro) 0 (n/s vs spiro) 0 (n/s vs no-spiro) 3 (1.1%) 0.99 (n/s) Time to appearance (days) 1.0 ± 0.0 1.0 ± 0.0 (n/a vs HA) (n/a vs spiro) 0 (n/a vs spiro) 0 (n/a vs no-spiro) 1.0 ± 0.0 n/a . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint (p-value) Duration (days) (p-value) 7.0 ± 0.0 7.0 ± 0.0 (n/a vs HA) (n/a vs spiro) 0 (n/a vs spiro) 0 (n/a vs no-spiro) 7.0 ± 0.0 n/a HA = hyperandrogenic; n/s = non-significant n/a = non-applicable As shown in Table 8, among proposed therapeutical options for COVID -19, besides azithromycin, which was given to all patients, nitazoxanide, hydroxychloroquine and ivermectin were p rescribed in statistically equal proportions. There wer e also no differences between the percentage of additional drugs prescribed between groups (Table 9). Table 8. Drugs prescribed for COVID-19. Medications (+ azithromycin) Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Hydroxichloroquine 91 (33.7%) 68 (34.8%) 20 (29.8%) 3 (37.5%) 88 (33.6%) 0.81 (n/s) Nitazoxanide 129 (47.8%) 93 (47.7%) 31 (43.3%) 5 (62.5%) 124 (47.3%) 0.75 (n/s) Ivermectin 64 (23.7%) 45 (23.1%) 18 (26.9%) 1 (12.5%) 63 (24.0%) 0.71 (n/s) Spironolactone 171 (63.3%**) 124 (63.6%) 47 (70.1%) 0 171 (65.3%) 0.42 (n/s) HA = hyperandrogenic; n/s = non-significant; n/a = non-applicable **Not including spironolactone users (n = 8) Table 9. Additional drugs and supplements used to treat COVID-19. Additional drugs or supplements Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Xa factor inhibitors 30 (11.1%) 23 (11.8%) 7 (10.4%) 0 30 (11.4%) 0.85 (n/s) Enoxaparin 23 (8.5%) 15 (7.7%) 8 (11.9%) 0 23 (8.8%) 0.80 (n/s) Glucocorticoids 34 (12.6%) 24 (12.3%) 10 (14.9%) 0 34 (13.0%) 0.78 (n/s) Vitamin C 57 (21.1%) 40 (20.5%) 15 (22.4%) 2 (25.0%) 55 (21.0%) 0.99 (n/s) Zinc 68 (25.2%) 48 (24.6%) 17 (25.4%) 3 (37.5%) 65 (24.8%) 0.98 (n/s) Vitamin D 221 (81.8%) 160 (82.0%) 56 (83.6%) 5 (62.5%) 216 (82.4%) 0.98 (n/s) HA = hyperandrogenic; n/s = non-significant; n/a = non-applicable . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint COVID-19 course and outcomes Table 10 shows the average time-to-treat and the clinical and biochemical duration of COVID-19 manifestations. Time-to-treat, positive rtPCR, and duration of symptoms with and without anosmia and ageusia were all significantly lower in HA-spiro when compared to non-HA, HA, and overall non-users, even when only HA-spiro females with symptoms were included for the analysis. Conversely, while time-to-treat was similar, all other three outcomes were significantly shorter in non-HA compared to HA. Table 11 details the clinical course of COVID-19. Although the peak of symptoms is apparent ly higher in the HA group, it d id not rea ch statistical significance when compared to non-HA due to the wide variability. Conversely, the recovery process seems to be slower in this group, as evidenced by the largest diffe rences between HA and non- HA, and between HA a nd HA-spiro in Days 2 to 7 , which coincides with the early response to proposed pharmacological treatments. HA-spiro group had fewer symptoms than non-spironolactone users in Days -3 to -1 and Day 7, as well as than HA females in Days -3 to -1 and Days 2 to 7. Table 10. COVID-19 clinical outcomes. Clinical clustering Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Time-to-treat (p-value) 3.0 ± 1.8 3.0 ± 1.9 (n/s (p=0.11) vs HA) (p<0.007 vs spiro) 3.4 ± 1.4 (p=0.0001 vs spiro) 0.5 ± 1.1 (Median = 0; 95%CI = 0.7) (p=0.0003 vs no-spiro) (p = 0.0028 vs no-spiro after adjustment for symptomatic patients only) 3.1 ± 1.8 0.0004 Duration of positive rtPCR (days) (p-value) 13.9 ± 6.1 13.5 ± 5.7 (p=0.033 vs vs HA) (p=0.025 vs spiro) 15.7 ± 6.7 (p=0.0005 vs spiro)) 8.7 ± 3.0 (p=0.015 vs no-spiro) (p = 0.015 vs no-spiro after adjustment for symptomatic patients only) 14.0 ± 6.1 0.0057 Remission not including anosmia (days) (without (p-value) 5.8 ± 4.7 5.5 ± 4.7 (p=0.0003 vs vs HA) (p=0.0008 vs spiro) (p = 0.015 vs spiro after adjustment for symptomatic patients only) 7.2 ± 4.5 (p<0.0001 vs spiro)) (p=0.0002 vs spiro after adjustment for only symptomatic patients) 1.4 ± 1.4 (Median = 1; 95%CI = 1.0) (p=0.0003 vs no-spiro) (p = 0.0046 vs vs no-spiro after adjustment for symptomatic patients only) 6.0 ± 4.7 < 0.0001 (p < 0.0001 after adjustment for only symptomatic patients) Remission including anosmia (days) (without (p-value) 9.6 ± 7.3 9.0 ± 7.1 (p=0.0001 vs HA) (p=0.0004 vs spiro) (p=0.0046 vs spiro after adjustment for symptomatic patients only) 12.4 ± 7.0 (p<0.0001 vs spiro)) (p<0.0001 vs spiro after adjustment for only symptomatic patients) 1.7 ± 2.0 (Median = 1; 95%CI = 0.3) (p=0.0001 vs no-spiro) (p = 0.0013 vs no-spiro after adjustment for symptomatic patients only) 9.9 ± 7.3 < 0.0001 (<0.0001 after adjustment for only symptomatic patients) . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint HA = hyperandrogenic; n/s = non-significant; n/a = non-applicable Table 11. COVID-19 course. Clinical evolution Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Days -7 to -4 (p-value) 80.2 ± 36.3 (Median = 100; 95%CI = 4.3) (75.9% asymptomatic) 78.1 ± 37.9 (Median = 100; 95%CI = 5.3) [144 (73.8%) asymptomatic] (n/s(p=0.41) vs HA) (n/s (p=0.21) vs spiro) 84.1 ± 32.6 (Median = 100; 95%CI = 7.8) [53 (79.1%) asymptomatic] (n/s (p=0.34) vs spiro) 100.0 ± 0 (Median = 100; 95%CI = 0) [8 (100%) asymptomatic] (n/s(p=0.23) vs no-spiro) 79.6 ± 36.7 (Median = 100; 95%CI = 4.4) [197 (75.2%) asymptomatic] 0.35 (n/s) Days -3 to -1 (p-value) 34.5 ± 40.0 (Median = 10; 95%CI = 4.8) (24.1% asymptomatic) 36.2 ± 40.8 (Median = 10; 95%CI = 5.7) [49 (24.0%) asymptomatic] (n/s (p=0.41) vs AGA vs HA) (p=0.007 vs spiro) 23.4 ± 31.0 (Median = 10; 95%CI = 7.4) [9 (13.4%) asymptomatic] (p=0.003 vs spiro) 87.5 ± 33.1 (Median = 100; 95%CI = 22.9) [7 (87.5%) asymptomatic (p=0.005 vs no-spiro) 32.9 ± 39.0 (Median = 10; 95%CI = 4.7) [58 (22.1%) asymptomatic] 0.014 Day 0 (p-value) 32.0 ± 35.0 (Median = 20; 95%CI = 4.2) (12.6% asymptomatic) 35.0 ± 36.6 (Median = 20; 95%CI = 5.1) [28 (14.3%) asymptomatic] ((p=0.045) vs AGA vs HA) (n/s (p=0.29) vs spiro) 20.1 ± 23.1 (Median = 15; 95%CI = 5.5) [2 (3.0%) asymptomatic] (n/s (p=0.16) vs spiro) 57.5 ± 46.3 (Median = 80; 95%CI = 32.1[4 (50.0%) asymptomatic] (n/s (p=0.25) vs no-spiro) 31.2 ± 34.3 (Median = 20; 95%CI = 4.2) [30 (11.4%) asymptomatic] 0.072 (n/s) Day 1 (p-value) 55.9 ± 33.6 (Median = 60; 95%CI = 4.0) (12.6% asymptomatic) 59.9 ± 33.0 (Median = 70; 95%CI = 4.6) [28 (14.3%) asymptomatic] (p=0.0002 vs HA) (/n/s (p<0.30) vs spiro) 43.0 ± 30.3 (Median = 40; 95%CI = 7.3) [2 (3.0%) asymptomatic] (n/s (p=0.12) )vs spiro) 65.0 ± 44.5 (Median = 97.5; 95%CI = 30.8) [4 (50.0%) asymptomatic] (n/s (p=0.23) vs no-spiro) 55.6 ± 33.1 (Median = 60; 95%CI = 4.0) [30 (11.4%) asymptomatic] 0.0008 Day 2 (p-value) 75.1 ± 27.0 (Median = 90; 95%CI = 3.2) (20.4% asymptomatic) 78.0 ± 26.0 (Median = 90; 95%CI = 3.6) [46 (23.6%) asymptomatic] (p<0.0001 vs HA) (p=0.0051) vs spiro) 65.4 ± 26.8 (Median = 70; 95%CI = 6.4) [4 (6.0%) asymptomatic] (p=0.0007 vs spiro) 83.75 ± 32.7 (Median = 100; 95%CI = 22.6) [5 (62.5%) asymptomatic] (p=0.044 vs no-spiro) 74.8 ± 26.8 (Median = 82.5; 95%CI = 3.2) [50 (19.1%) asymptomatic] < 0.0001 Day 3 (p-value) 85.7 ± 22.5 (Median = 95; 95%CI = 2.7) (36.3% asymptomatic) 87.0 ± 22.8 (Median = 95; 95%CI = 3.2) [81 (41.5%) asymptomatic] (p=0.0007 vs HA) (p<0.0001 vs spiro) 80.3 ± 21.6 (Median = 90; 95%CI = 5.2) [9 (13.4%) asymptomatic] (p=0.0001 vs spiro) 100.0 ± 0 (Median = 100; 95%CI = 0) [8 (100%) asymptomatic] (p=0.0016 vs no-spiro) 85.3 ± 22.7 (Median = 90; 95%CI = 2.7) [90 (34.3%) asymptomatic] < 0.0001 Day 7 (p-value) 94.0 ± 14.3 (Median = 100; 95%CI = 1.7) (66.3% asymptomatic) 94.1 ± 15.5 (Median = 100; 95%CI = 2.2) [137 (70.2%) asymptomatic] (p=0.023 vs HA) (n/s (p=0.15) vs spiro) 93.0 ± 11.3 (Median = 100; 95%CI = 2.7) [34 (50.7%) asymptomatic] (p=0.023 vs spiro) 100.0 ± 0 (Median = 100; 95%CI = 0) [8 (100%) asymptomatic] (n/s (p=0.094) vs no-spiro) 93.8 ± 14.5 (Median = 100; 95%CI = 1.8) [171 (65.3%) asymptomatic] 0.018 Day 14 (p-value) 98.5 ± 7.6 (Median = 100; 95%CI = 0.9) (90.0% asymptomatic) 98.5 ± 8.4 (Median = 100; 95%CI = 1.2) [178 (91.3%) asymptomatic] (n/s(p=0.45) vs HA) (n/s (p=0.67) vs spiro) 98.3 ± 5.2 (Median = 100; 95%CI = 1.2) [57 (85.1%) asymptomatic] (n/s (p=0.49) vs spiro) 100.0 ± 0 (Median = 100; 95%CI = 0) [8 (100%) asymptomatic] (n/s (p=0.62) vs no-spiro) 98.4 ± 7.7 (Median = 100; 95%CI = 0.9) [235 (89.7%) asymptomatic] 0.67 (n/s) Day 21 (p-value) 99.4 ± 6.3 (Median = 100; 95%CI = 0.7) (97.4% asymptomatic) 99.4 ± 7.2 (Median = 100; 95%CI = 1.0) [191 (97.9%) asymptomatic] (n/s vs HA) (n/s vs spiro) 99.5 ± 2.7 (Median = 100; 95%CI = 0.7) [64 (95.2%) asymptomatic] (n/s vs spiro) 100.0 ± 0 (Median = 100; 95%CI = 0) [8 (100%) asymptomatic] (n/s vs no-spiro) 99.4 ± 6.3 (Median = 100; 95%CI = 0.9) [235 (89.7%) asymptomatic] 1.0 (n/s) Day 30 (p-value) 99.8 ± 3.0 (Median = 100; 95%CI = 0.4) (98.5%) asymptomatic) 99.7 ± 3.6 (Median = 100; 95%CI = 0.5) [191 (97.9%) asymptomatic] (n/s vs HA) (n/s vs spiro) 100.0 ± 0 (Median = 100; 95%CI = 0.0) [67 (100%) asymptomatic] (n/s vs spiro) 100.0 ± 0 (Median = 100; 95%CI = 0) [8 (100%) asymptomatic] (n/s vs no-spiro) 99.8 ± 3.1 (Median = 100; 95%CI = 0.8) [255 (98.5%) asymptomatic] 1.0 (n/s) . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Day 60 (p-value) 100.0 ± 0 (Median = 100; 95%CI = 0) (100% asymptomatic) 100.0 ± 0 (Median = 100; 95%CI = 0) [195 (100%) asymptomatic] (n/s vs HA) (n/s vs spiro) 100.0 ± 0 (Median = 100; 95%CI = 0.0) [67 (100%) asymptomatic] (n/s vs spiro) 100.0 ± 0 (Median = 100; 95%CI = 0) [8 (100%) asymptomatic] (n/s vs no-spiro) 100.0 ± 0.0 (Median = 100; 95%CI = 0.0) [262 (100%) asymptomatic] 1.0 (n/s) HA = hyperandrogenic; n/s = non-significant; n/a = non-applicable; CI = confidence interval As in Table 1 2, which depicts the World Health Organization (WH O) COVID Ordinal Outcomes in the present population, HA females had significantly higher scores than non-HA and HA -spiro groups in Days 0 and 7, and be came similar after Day 14. None of the patients presents as Scores 3 to 5 throughout the disease. The loss of capacity to perform everyday activities is described in Table 13. None of the spironolactone users had any level of loss of everyday activities. Conversely, HA were significantly more compromised when compared to non-HA and HA-spiro females (p = 0.004 and 0.028, respectively), by the time that proposed treatments started (Day 0). This was mitigated in the following days (Days 3, 7, 14 and 30). Except for spiro-HA (not affected at any time), Day 3 had significant improvement compared to Day 0. Table 12. WHO COVID-19 Ordinal Outcomes. WHO COVID Ordinal Outcomes Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Day 0 (p-value) (p=0.028 vs HA) (n/s((p=0.15) vs spiro) (p=0.028 vs spiro) (n/s((p=0.098) vs no-spiro) 0.002 Stage 1 180 (66.7%) 137 (70.3%) 35 (52.2%) 8 (100%) 172 (65.6%) Stage 2 90 (33.3%) 58 (29.7%) 32 (47.8%) 0 90 (34.4%) Stages 3-5 0 0 0 0 0 Day 7 (p-value) (p=0.021 vs HA) (n/s vs spiro) (p=0.015 vs spiro) (n/s vs no-spiro) 0.021 Stage 1 255 (94.4%) 186 (95.4%) 61 (91.0%) 8 (100%) 247 (94.3%) Stage 2 15 (5.6%) 9 (4.6.%) 6 (9.0%) 0 15 (5.7%) Stages 3-5 0 0 0 0 0 vs Day 0 (p-value) < 0.0001 < 0.0001 0.0001 1.0 (n/s) < 0.0001 Day 14 (p-value) (n/s vs HA) (n/s vs spiro) (n/s vs spiro) (n/s vs no-spiro) 0.84 Stage 1 268 (99.3%) 195 (100%) 65 (97.0%) 8 (100%) 260 (99.2%) Stage 2 2 (0.7%) 0 2 (3.0%) 0 2 (0.8%) . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Stages 3-5 0 0 0 0 0 vs Day 0 (p-value) < 0.0001 < 0.0001 < 0.0001 1.0 (n/s) < 0.0001 vs Day 7 (p-value) 0.45 (n/s) 0.43 (n/s) 0.55 (n/s) 1.0 (n/s) 0.28 (n/s) Days 30 and 60 (p-value) (n/s vs HA) (n/s vs spiro) (n/s vs spiro) (n/s vs no-spiro) Stage 1 270 (100%) 195 (100%) 67 (100%) 8 (100%) 262 (100%) n/a Stage 2 0 0 0 0 0 n/a Stages 3-5 0 0 0 0 0 n/a HA = hyperandrogenic; n/s = non-significant; n/a = non-applicable; n/s = non-significant Table 13. Loss of ability to perform everyday activities due to COVID-19. Loss of ability of everyday activities (%) Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Day 0 (p-value) 9.2 ± 15.2 (Median = 0; 95%CI = 1.8) (63.7% full functional capacity) 7.3 ± 13.2 (Median = 0; 95%CI = 1.8) [137 (70.2%) full functional capacity] (p=0.004) vs HA) (n/s (p=0.15) vs spiro) 16.0 ± 18.9 (Median = 0; 95%CI = 4.5) [35 (52.2%) full functional (p=0.028 vs spiro) 0 (Median = 0; 95%CI = 0) [8 (100%) full functional capacity] (n/s (p=0.098) vs no-spiro) 9.5 ± 15.3 (Median = 0; 95%CI = 1.8) [172 (65.6%) full functional capacity] 0.005 Day 3 (p-value) 3.2 ± 9.1 (Median = 0; 95%CI = 1.1) (83.0% full functional capacity) 2.4 ± 7.4 (Median = 0; 95%CI = 1.0) [172 (88.2%) full functional capacity] (n/s (p=0.17) vs HA)) (n/s (p=0.57) vs spiro) 6.0 ± 12.8 (Median = 0; 95%CI = 3.1) [52 (77.6%) full functional capacity] (n/s (p=0.30) vs spiro) 0 (Median = 0; 95%CI = 0) [8 (100%) full functional capacity] (n/s (p=0.48) vs no-spiro) 3.2 ± 9.0 (Median = 0; 95%CI = 1.1) [124 (85.5%) full functional capacity] 0.30 (n/s) vs Day 0 (p-value) 0.0001 0.0016 0.005 1.0 (n/s) 0.0001 Day 7 (p-value) 1.1 ± 4.7 (Median = 0; 95%CI = 0.6) (91.5% full functional capacity) 0.8 ± 4.0 (Median = 0; 95%CI = 0.6) [186 (95.4%) full functional capacity] (n/s (p=0.95) vs HA) (n/s (p=0.82) vs spiro) 1.9 ± 6.4 (Median = 0; 95%CI = 1.5) [61 (91.0%) full functional capacity] (n/s (p=0.68) vs spiro) 0 (Median = 0; 95%CI = 0) [8 (100%) full functional capacity] (n/s (p=0.78) vs no-spiro) 1.0 ± 4.7 (Median = 0; 95%CI = 0.6) [247 (91.5%) full functional capacity] 0.83 (n/s) vs Day 0 (p-value) < 0.0001 < 0.0001 < 0.0001 1.0 (n/s) < 0.0001 vs Day 3 (p-value) 0.082 (n/s) 0.21 (n/s) 0.17 1.0 (n/s) 0.055 (n/s) Day 14 (p-value) 0.1 ± 0.8 (Median = 0; 95%CI = 0.1) (96.3% full functional capacity) 0 (Median = 0; 95%CI = 0) [195 (100%) full functional capacity] (n/s vs HA) (n/s vs spiro) 0.3 ± 1.7 (Median = 0; 95%CI = 0.4) [65 (97.0%) full functional capacity] (n/s vs spiro) 0 (Median = 0; 95%CI = 0) [8 (100%) full functional capacity] (n/s vs no-spiro) 0.1 ± 0.8 (Median = 0; 95%CI = 0.1) [260 (96.3%) full functional capacity] 0.99 (n/s) vs Day 0 (p-value) < 0.0001 < 0.0001 < 0.0001 1.0 (n/s) < 0.0001 vs Day 7 0.93 (n/s) 0.99 (n/s) 0.67 (n/s) 1.0 (n/s) 0.95 (n/s) . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint (p-value) Day 30 (p-value) 0 (Median = 0; 95%CI = 0) (100% full functional capacity) 0 (Median = 0; 95%CI = 0) [195 (100%) full functional capacity] (n/s vs HÁ) (n/s vs spiro) 0 (Median = 0; 95%CI = 0) [67 (100%) full functional capacity] (n/s vs spiro) 0 (Median = 0; 95%CI = 0) [8 (100%) full functional capacity] (n/s vs no-spiro) 0 (Median = 0; 95%CI = 0) [262 (100%) full functional capacity] (n/s vs HÁ) (n/s vs spiro) 1.0 (n/s) HA = hyperandrogenic; n/s = non-significant; CI = confidence interval Table 14 estimates the percentage of lung affected in COVID-19 in Days 0, 7, 14 and 30 through chest computerized to mography (CT) scan . Because none of HA-spiro reported severe symptoms, none of th ese patients underwent CT. Non-HA had significantly less lungs affected than HA females, except for Day 30. None of these two groups had significant improvement Table 15 summarizes the scores of COVID-19 progress ion and severity , and complications related to the disease. None of the 270 patients enrolled in the present study presented any complication for COVID-19 throughout the treatment. Table 14. Chest CT scan. Chest CT scan (% of lungs affected) Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) Non-spiro (non-HA + HA) (n = 262) p-value (overall) Day 0 (p-value) 14.7 ± 14.8 (Median = 10; 95%CI = 1.8) [32 (34.4%) normal lungs of 93 patients that performed chest CT scan in day 0] 11.8 ± 13.3 (Median = 10; 95%CI = 1.9) [25 (41.0%) normal lungs of 61 patients that performed chest CT scan in day 0] (p=0.015 vs HA) (n/a vs spiro) 20.3 ± 15.7 (Median = 25; 95%CI = 3.8) [7 (21.89%) normal lungs of 32 patients that performed chest CT scan in day 0] (n/a vs spiro) n/a (no patients performed chest CT scan) (= overall) n/a Day 7 (p-value) 15.0 ± 14.6 (Median = 10; 95%CI = 1.7) [20 (31.25%) normal lungs of 64 patients that performed chest CT scan in day 7] 11.9 ± 12.9 (Median = 10; 95%CI = 1.8) [15 (38.4%) normal lungs of 39 patients that performed chest CT scan in day 7] (p=0.053 vs HA) (n/a vs spiro) 19.8 ± 15.8 (Median = 10; 95%CI = 3.8) [5 (20.0%) normal lungs of 25 patients that performed chest CT scan in day 7] (n/a vs spiro) n/a (= overall) n/a vs Day 0 (p-value) 0.85 (n/s) 0.89 (n/s) 0.92 (n/s) n/a Day 14 (p-value) 11.2 ± 11.7 (Median = 10; 95%CI = 1.4) [23 (39.0%) normal lungs of 59 patients that performed chest CT scan in day 14] 7.3 ± 9.0 (Median = 0; 95%CI = 1.3) [19 (51.3%) normal lungs of 37 patients that performed chest CT scan in day 14] (p=0.003 vs HA) (n/a vs spiro) 17.7 ± 12.8 (Median = 17.5; 95%CI = 3.0) [4 (18.1%) normal lungs of 22 patients that performed chest CT scan in day 14] (n/a vs spiro) n/a (no patients performed chest CT scan) (= overall) n/a . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint vs Day 0 (p-value) 0.25 (n/s) 0.18 (n/s) 0.62 (n/s) n/a vs Day 7 (p-value) 0.21 (n/s) 0.17 (n/s) 0.75 (n/s) n/a Day 30 (p-value) 10.9 ± 10.3 (Median = 17.5; 95%CI = 1.4) [6 (37.5%) normal lungs of 16 patients that performed chest CT scan in day 30] 9.6 ± 10.0 (Median = 10; 95%CI = 1.1) [5 (41.7%) normal lungs of 12 patients that performed chest CT scan in day 30] (n/s (p=0.32) vs HA) (n/a vs spiro) 15.0 ± 10.6 (Median = 17.5; 95%CI = 2.5) [1 (25.0%) normal lungs of 4 patients that performed chest CT scan in day 30] (n/a vs spiro) n/a (no patients performed chest CT scan) (= overall) n/a vs Day 0 (p-value) 0.49 (n/s) 0.78 (n/s) 0.44 (n/s) 1.0 (n/s) vs Day 7 (p-value) 0.44 (n/s) 0.72 (n/s) 0.53 (n/s) 1.0 (n/s) vs Day 14 (p-value) 0.95 (n/s) 0.52 (n/s) 0.67 (n/s) 1.0 (n/s) HA = hyperandrogenic; n/s = non-significant; CI = confidence interval Table 15. COVID-19 progression outcomes. Disease progression outcomes Overall females (n = 270) Non-HA females (n = 195) HA females (n = 67) Spironolactone Users (spiro) (n = 8) p-value (overall) Brescia COVID-19 Respiratory Severity Scale (0-4) 0 0 0 0 1.00 (n/s) Hospitalization 0 0 0 0 1.00 (n/s) Mechanical ventilation 0 0 0 0 1.00 (n/s) Noradrenaline/dopamine 0 0 0 0 1.00 (n/s) Death 0 0 0 0 1.00 (n/s) HA = hyperandrogenic; n/s = non-significant; CI = confidence interval

The importance of an early diagnosis of COVID-19 The present prospective observational study aimed to characterize patients upon the diagnosis of COVID-19. Oppositely to the prevailing recommendations to suspect for COVID-19 only in the presence of fever of s hortness of breath, we detected COVID-19 guided by any symptom , related or supposedly unrelated to the dise ase. Another key modification was the active questions, since patients tend to avoid reporting symptoms . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint that they do not believe to be due to COVID-19. The majority of patients had at least one complain that would not be reported if not actively questioned. Researches have largely focu sed approaches for stages two and three, w hile assumed that none of proposed therapies for earlier stages would be effective (‘wait and see’). Culturally, unlike bacterial in fections, specific antiviral treatment s tend to be avoided. Except for chronic and re lapsing viruses , antivirals have been st udied at a n irregular basis, and their clinical use have long been avoided. Moreover, drugs for other diseases that might disclose direct or indirect antiviral activity have been underestimated. More important than the culture of non-viral treatment that surrounds the medical field, is that the lack of specific symptoms and the usual presence of uncommon clinical presentations, th at likely precluded the majority of the patients from a n actual early diagnosis. The change in the paradigm for detection of CO VID-19 towards more sensitive and active search for C OVID-19 may have driven the better outcomes observed in the present study. We consider imperative that policies for COVID-19 diagnosis change The patient with COVID-19: characterization of the patients and the disease Baseline characteri stics were signi ficantly different bet ween non -HA and HA probably because HA females tend to be shorter and present higher BMI. The differences between age is understandable when one considers that the phenotypical ex pression of HA tends to occur earlier in lifetime. Furthermore, recognition of female HA (except for dramatic cases, as beard females) has only occurred in the last decades, and was underdiagnosed until short time ago. Differences in BMI between HA and non-HA was rela tively lower than differences in age, and obesity was similarly present be tween groups. Conversely, even being significantly younger, HA females presented more severe manifestations of COVID-19 than non-HA . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint While none of the prevailing chronic diseases presented any difference betw een groups, metformin and oral contraceptives were more used in HA than non-HA females, possibly due to the fact the PCOS is the most common underlying condition that leads to HA. In a similar manner of what has been observed in males, symptoms in females could be easily clustered (23-26). At least two s ymptoms of each cluster were together within a same patient in more than 80% of patients. The clusters of a nosmia-ageusia predominance, dengue fever-like, URTI-like, and GI infection-like clinical presentation have been detected, in addition to those with unspecific symptoms, or with more than one cluster. The remarkable URTI-like clinical symptoms among HA females may be due to enhanced infectivity that has been shown to occur under hyperandrogenic states. The similar number of patients treated with hydroxychloroquine, nitazoxanide and ivermectin and the lack of major differences at least in terms of COVID-19 complications shows that the choice for a specifi c medication to be added to azithromycin is not imperative. Also, although preliminary hypotheses and data demonstrating the important role of azithromycin as an adjuvant therapy against SARS-CoV-2, whether this drug plays an add-on benefit effect is uncertain, and remains unanswered, since we did not compare with versus without its use. Correspondingly, additional drugs and supplements were given according to an individualized clinical evaluation, and were used in similar percentages between groups. The use of antico agulant, specially Xa factor in hibitors and enoxaparin, were based on the risk for thr ombosis and development of other complications, since COVID -19 is a pro-thrombotic state per se. Vitamin D, and at slightly lesser extent, zinc and vitamin C, were prescr ibed to the majority of treated patien ts. Hence, whether these additi onal prescriptions helped to improve outcomes cannot be concluded from the present study. COVID-19 outcomes All early COVID -19 related outcomes, including du ration of positive rtPCR - SAR-CoV-2 and symptoms with or without anosmia and ageusia were markedly lower in HA-spiro when compared to non -HA, HA, and overall non -users. However, time-to- . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint treatment was equally lower in HA-spiro females. Hence, whether the shorter and milder clinical presentation in COVID-19 among spironolactone users is due to earlier beginning of treatment is unknown, although COVID-19 disease course remained better in the HA- spiro group after adjustments for non-asymptomatic patients only. Unlike HA-spiro, since time-to-treat and treatment options were similar between non-HA and HA females, obesity prevalence was similar, and HA were younger than non-HA, prolonged clinical and viral duration among HA females may only be explained by the hyperandrogenism. Although spironolactone users had more asym ptomatic presentation than other groups, when HA-spiro patients presented symptoms, these symptoms presented similar time-to-appearance and duration than n on-HA females. The great majority of the symptoms occurred in similar percentages and presented similar durations for all groups. Increased fatigue, dry cou gh and anosmia, and increased and prolonged weakness, myalgia and ageusia found in HA females should also be explained by hyperandrogenism, in an analogical manner than male A GA (5,6,14,16,17), due to en hanced expression of TMPRSS-2. Differences in disease course were more evident between Days 0 and 7 in HA females, as this population presented sl ower speed of improvement compared to other groups, as evidence d by t hree methods to quantify this speed (Ta bles 11 t o 13), which means that HA females were shown to improve, but took l onger than other groups . Despite the slower symptom regression in HA, all groups showed increased recovery rate after initiation of any of the proposed therapies. The importance of measuring loss of capacity to perform everyday activities lays on the fact that infected patients remain isolated for approximately 14 days, and needs to be able to perform the basic hygiene and self-care activities during this period. Although HA females were more compromised than non -HA and HA -spiro females, even this group was only slightly affected, and lasted for less than three days. Unlike rapid clinical recovery, r adiology did not show any improvement during the first 30 days after treatment initiatio n. However, there is a major bias that those that . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint underwent more than one chest CT were those that persisted with complains or were already compromised. Hence, radiological improvement may be less prolonged that what has been shown in the present study. Criteria employed to measure COVID-19 severity, ilike WHO COVID Scale and Brescia Respiratory Scale, revealed to be inaccurate in the present analysis, since none of the patients progressed to respiratory complications, hospitalization, and other outcomes. Instead, more sensitive scales were better to detect differences. Oppositely to non-spironolactone HA females, HA that use spironolactone had markedly better clinical outcomes related to COVID-19. Additionally, the small number of patients with COVID -19 taking spironolactone may reflect its potential pre ventive effect, despite the lack of any other data to support this hypothesis. Spironolactone has been proposed to exert multiple be neficial roles in COVID -19, incl uding increase of circulating-to-attached angiotensin converting enzyme 2 ( ACE2) ratio, increased angiotensin receptor ( AT) 2-to-AT1 ratio, decr eased TMPRSS2 ac tivity, as well as antiviral and anti -inflammatory activities (20-22), and i ts promising roles have been reinforced by the present analysis. Post-COVID syndrome? The importance of unveiling potential effective treatments for COVID-19 not only lays on preventing hospitalization, mechanical ventilation, and death, but also preventing long-term, post-cure persisting symptoms, termed as “post-COVID syndrome”, that has shown to be present in at least 20 to 30% of patients, and includes autoimmune, mental, psychiatric, muscular, respi ratory and gastro intestinal disorders, and persisted fatigue, resembling chronic fatigue syndrome (CFS), unrel ated to any specific biochemical marker to date (27,28). Hence, when ever a pharmacological intervention study is conducted, researchers should not only consider acute, but also chronic complications as targets to evaluate efficacy. The effect of spe cific treatment s on COVID-19: compar ison with existing literature . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint When duration and severity of clinical symptoms, disease duration, and related complications are compared with the ex tensive reports on the literature, differences are undisputable, particularly for clinical outcomes (10-12). At least in the present population of females, differences were sufficiently overwhelming to justify the lack of need for placebo-control trials, from an ethical perspective. However, open label RCTs and other prospective observational studies are mandatory to confirm the clear fin dings of the present study, when compared to the expected outcomes for COVID-19 in females.

As a prospective observational study conducted prior to a double-blind placebo- control RCT , aiming to better determine the selection process and parameters to be evaluated, as well as define the most plausible pharmacol ogical approach between hydroxychloroquine, nitazoxanide, ivermectin, or none. As per the study deisgn, the lack of a placebo group initially hampers from conclusive findings. However , this has been overcome by the evident differences when compared to the well-established COVID-19 course and outcomes, which disallows us from perform ing the RCT as a ful l placebo- control. The replication o f a highly sensitive case-detection guidance that include the occurrence of absolu tely any symptom as being suspected for COVID -19 may find barriers that may preclude from a successful approach, including: 1. Lack of general and medical education regarding the unspecific pattern of COVID-19 clinical presentation; 2. Self-judgement of not being affected by COVID-19; 3. Inability to correlate non-obvious symptoms with COVID-19, losing the window of opportunity; and 4. Cultural focus on the severe patient, neg lecting those are possibly preventable from CO VID-19 complication. Although this may not necessarily occur with all patients, the larger number of patients treated early, the better clinical outcomes should be. Final discussion . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint Since COVID -19 is an ex tensive, diffuse, and largely unclarified disease, a thorough medical evaluation that e ncompasses questions for all organs and systems should be performed at a regular basis, in clinical practice, not only restr icted to researching purposes. The early detec tion associated with any drug combination among azithromycin, hydroxychloroquine, nitazoxanide and ivermectin yielded notable improvements in terms of course of COVID-19, when compared to literature. Among females, presenting HA may be an additional risk for COVID-19 severity, as they presented to be slightly more affected and had a prolonged recovery process when compared to non -HA and spironolactone users. The chronic use of spironolact one in a daily dose 100mg or more in HA females not only seemed to mi tigate, but also reduce risks to below those found in non-HA. This has shown to have similar relative risks when compared to the presence of AGA and use of dutasteride among males, respectively.

A sensitive, early detection of COVID-19 followed by a pharmaceutical approach with a drug combination between azithrom ycin in all cases, associated with hydroxychloroquine, nitazoxanide or ivermectin demonstrated unequivocal differences when compared to the extensively described natural diseas e course when pharmacological treatments potentially effective for COVID-19 is not provided , particularly if detected until three days after the beginning of symptoms. When compared to non-hyperandrogenic (non-HA), HA females had a more severe and prolonged clinical presentation, even though risk of development of COVID -19 related complications remained unmeaningful among all groups. Spironol actone mitigated any additional risk due to HA. Funding statements . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint The fundi ng of prese nt study was fully suppo rted by Corpometria Institut e (Brasilia, DF, Brazil) and Applied Biology Inc (Irvine, CA, USA). Conflict of interest statement Authors declare no conflict of interest with any of the pharmacological interventions proposed by the present study.

1. Guan W, Ni Z, Hu Y, et al. Clinical characteristics of coronavirus disease 2019 in China. N Engl J Med 2020; Feb 28. 2. Hajifathalian K, Kumar S, Newberry C, et al. Obesity is associated with worse outcomes in COVID-19: Analysis of Early Data From New York City [published online ahead of print, 2020 May 29]. Obesity (Silver Spring) . 2020;10.1002/oby.22923. 3. Kalligeros M, Shehadeh F, Mylona EK, et al. Association of Obesity with Disease Severity among Patients with COVID-19 [published online ahead of print, 2020 Apr 30]. Obesity (Silver Spring). 2020;10.1002/oby.22859. 4. Palaiodimos L, Kokkinidis DG, Li W, et al. Severe obesity, increasing age and male sex are independently associated with worse in -hospital outcomes, and higher in-hospital mortality, in a cohort of patients with COVID-19 in the Bronx, New York [published online ahead of print, 2020 May 16]. Metabolism. 2020;108:154262. 5. Cavalcanti AB, Zampieri FG, Rosa RG, Azevedo LCP, Veiga VC, Avezum A, Damiani LP, Marcadenti A, Kawano-Dourado L, Lisboa T, Junqueira DLM, de Barros E Silva PGM, Tramujas L, Abreu -Silva EO, Laranjeira LN, Soares AT, Echenique LS, Pereira AJ, Freitas FGR, G ebara OCE, Dantas VCS, Furtado RHM, Milan EP, Golin NA, Cardoso FF, Maia IS, Hoffmann Filho CR, Kormann APM, Amazonas RB, Bocchi de Oliveira MF, Serpa-Neto A, Falavigna M, Lopes . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint RD, Machado FR, Berwanger O; Coalition Covid -19 Brazil I Investigators. Hydroxychloroquine with or without Azithromycin in Mild -to-Moderate Covid- 19. N Engl J Me d. 2020 Jul 23:NEJMoa2019014. doi: 10.1056/NEJMoa2019014. 6. Zhou F, Yu T, Du R, et al. Clinical course and risk factors for mortality of adult inpatients with C OVID-19 in Wuhan, China: a retr ospective cohort stud y [published online ahead of print, 2020 Mar 11] [published correction appears in L ancet. 2020 Mar 12;:]. Lancet. 2020;S0140-6736(20)30566-3. 7. Lauer SA, Grantz KH, Bi Q, et al. The Incubation Period of Coronavirus Disease 2019 ( COVID-19) From Publicly Repo rted Confirmed Cases: Estimation and Application [published online ahead of print, 2020 Mar 10]. Ann Intern Med . 2020;10.7326/M20-0504. 8. Wu C, Chen X, Cai Y, et al. Risk Factors Associated With Acute Respi ratory Distress Syndrome and Death in Patients With Coronavirus Disease 2019 Pneumonia in Wuhan, China [published online ahead of print , 2020 Mar 13]. JAMA Intern M ed. 2020;10.1001/jamainternmed.2020.0994. 9. Wu C, Chen X, Cai Y, et al. Risk Factors Associat ed With Acute Respirator y Distress Syndrome and Death in Patients With Coronavirus Disease 2019 P neumonia in Wuhan, China [published on line ahead of print, 2020 Mar 13]. JAMA Intern Med . 2020;10.1001/jamainternmed.2020.0994. 10. https://www.who.int/teams/health-care-readiness-clinical-unit/covid-19/data- platform (Last accessed September 29th, 2020) 11. https://www.cdc.gov/coronavirus/2019-ncov/hcp/clinical-guidance- management-patients.html (Last accessed September 29th, 2020) 12. Stokes EK, Zambrano LD, And erson KN, et al. Coronavirus Disease 2019 Cas e Surveillance - United States, January 22-May 30, 2020. MMWR Morb Mortal Wkly Rep 2020; 69:759. 13. Cadegiani, F.A. Repurposing existing drugs for COVID -19: an e ndocrinology perspective. BMC E ndocr Disord 20, 149 (2020). https://doi.org/10.1186/s12902 -020- 00626-0 14. Pascarella G, Strumia A, Piliego C, Bruno F, Del Buono R, Costa F, Scarlata S, Agrò FE. COVID-19 diagnosis and management: a comprehensive review. J Intern Med. 2020 Aug;288(2):192-206. doi: 10.1111/joim.13091. . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint 15. Goren A, Vano -Galvan S, Wambier CG, et al. A preliminary observation: male pattern hair loss among hospitalized COVID-19 patients in Spain - A potential clue to the role of androgens in COVID-19 severity [published online ahead of print, 2020 Apr 16]. J Cosmet Dermatol. 2020;10.1111/jocd.13443. 16. Goren A, McCoy J, Wambier CG, et al. What does androgenetic alopecia have to do with COVID-19? An insight into a potential new therapy [published online ahead of print, 2020 Apr 1]. Dermatol Ther. 2020;e13365. 17. McCoy J, Wambier CG, Herrera S, Vaño-Galván S, Gioia F, Comeche B, Ron R, Serrano-Villar S, Iwasiow RM, Tayeb MA, Cadegiani FA, Mesinkovska NA, Shapiro J, Sinclair R, Goren A. Androgen Receptor Gen etic Variant Predicts CO VID-19 Disease Severity: A Prospective Longitudinal Study of Hospitalized COVID-19 Male Patients. J Eur Acad Dermatol Venereol. 2020 Sep 25. 18. Kragholm K, Andersen MP, Gerds TA, et al. Association between male sex and outcomes of Coronavirus Disease 2019 (Covid-19) - a Danish nationwide, register-based study. Clin Infect Dis 2020. 19. Goren A, Wambier CG, Herrera S, McCoy J, Vaño-Galván S, Gioia F, Comeche B, Ron R, Serrano-Villar S, Ramos PM, Cadegiani FA, Kovacevic M, Tosti A, Shapiro J, Sinclair R. Anti -androgens may protect against severe COVID -19 outcomes: results from a prospective cohort study of 77 hospitalized men. J E ur Acad Dermatol Venereol. 2020 Sep 25. doi: 10.1111/jdv.16953. 20. Cadegiani FA. Can spironolactone be u sed to prevent COVID -19-induced acute respiratory distress syndrome in patients with hypertension? Am J Physiol Endocrinol Metab. 2020 May 1;318(5):E587-E588. doi: 10.1152/ajpendo.00136.2020. 21. Cadegiani FA, Wambier CG, Goren A. Spironolactone: An Anti -androgenic and Anti-hypertensive Drug That May Provide Protection Against the Novel Coronavirus (SARS-CoV-2) Induced Acute Respiratory Distress Syndrome (ARDS) in COVID-19. Front Med (Lausanne). 2020 Jul 28;7:453. doi: 10.3389/fmed.2020.00453. 22. Cadegiani FA, Go ren A, Wambier CG. Spironolactone may provide protection from SARS-CoV-2: Targeting androgens, angiotensin converting enzyme 2 (ACE2), and renin-angiotensin-aldosterone system (R AAS). Med Hypotheses. 2020 J ul 16;143:110112. doi: 10.1016/j.mehy.2020.110112. 23. https://www.uptodate.com/contents/coronavirus-disease-2019-covid-19-clinical- features?topicRef=126981&source=see_link (Last accessed September 29th, 2020) . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint 24. Kim GU, Kim MJ, Ra SH, Lee J, Bae S, Jung J, Kim SH. Clinical characteristics of asymptomatic and symptomatic patients with mild COVID-19. Clin Microbiol Infect. 2020 Jul;26(7):948.e1-948.e3. doi: 10.1016/j.cmi.2020.04.040. 25. Lechien JR, Chiesa -Estomba CM, De Si ati DR et al. Olfactory and gustatory dysfunctions as a clinical presentation of mild -to-moderate forms of the cor onavirus disease (COVID-19): a multicenter European study. Eur Arch Otorhinolaryngol. 2020 Aug;277(8):2251-2261. doi: 10.1007/s00405-020-05965-1. 26. Tong JY, Wong A, Zhu D, Fastenberg JH, Tham T. The Prevalence of Olfactory and Gustatory Dysfunction in COVID -19 Patien ts: A Systemat ic Revi ew and Meta - analysis. Otolaryngol Head Neck Surg. 2020 Jul;163(1):3 -11. d oi: 10.1177/0194599820926473. 27. Novak P. Post COVID -19 syndrome associated with orthostatic cerebral hypoperfusion syndrome, small fiber neuropathy and benefit of immunotherapy: a case report. eNeurologicalSci. 2020 Dec;21:100276. doi: 10.1016/j.ensci.2020.100276. 28. Clauw DJ, Häuser W, Coh en SP, Fitzcharles MA. Considering the potential for an increase in chronic pain after the COVID -19 pandemic. Pain. 2020 Aug;161(8):1694-1697. doi: 10.1097/j.pain.0000000000001950. . CC-BY-ND 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted October 6, 2020. ; https://doi.org/10.1101/2020.10.05.20206870doi: medRxiv preprint