Digit ratios and their asymmetries as risk factors of developmental instability and hospitalization for COVID-19

Authors: A. Kasielska-TrojanJ. T. ManningM. JabłkowskiJ. Białkowska-WarzechaA. L. Hirschberg & B. Antoszewski  Scientific Reports volume 12, Article number: 4573 (2022) Cite this article Article Open Access Published: 


COVID-19 presents with mild symptoms in the majority of patients but in a minority it progresses to acute illness and hospitalization. Here we consider whether markers for prenatal sex hormones and postnatal stressors on developmental instability, i.e. digit ratios and their directional and unsigned asymmetries, are predictive of hospitalization. We focus on six ratios: 2D:3D; 2D:4D; 2D:5D; 3D:4D; 3D:5D; 4D:5D and compare hospitalized patient and control means for right, and left ratios, directional asymmetries (right–left) and unsigned asymmetries [|(right–left)|]. There were 54 patients and 100 controls. We found (i) patients differed in their digit ratios from controls (patients > controls) in all three ratios that included 5D (2D:5D, 3D:5D and 4D:5D) with small to medium effect sizes (d = 0.3 to 0.64), (ii) they did not differ in their directional asymmetries, and (iii) patients had greater |(right–left)| asymmetry than controls for 2D:4D (d = .74) , and all ratios that included 5D; 2D:5D (d = 0.66), 3D:5D (d = .79), 4D:5D (d = 0.47). The Composite Asymmetry of the two largest effects (2D:4D + 3D:5D) gave a patient and control difference with effect size d = 1.04. All patient versus control differences were independent of sex. We conclude that digit ratio patterns differ between patients and controls and this was most evident in ratios that included 5D. Large |(right–left)| asymmetries in the patients are likely to be a marker for postnatal stressors resulting in developmental perturbations and for potential severity of COVID-19.


Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes a respiratory and systemic illness (COVID-19) which may present as a severe pneumonia in 10–15% of patients. Severe disease can lead to acute respiratory distress and multi-organ failure often followed by intravascular coagulopathy1,2. Due to this variety and unknown severity and death risk factors, many studies and analyses have focused on identifying biomarkers of severe disease or poor outcomes in COVID-19 infections. Recent studies have shown that the clinical progress could be severe in cases of increased: neutrophil-lymphocytes ratio, C-reactive protein (CRP), troponin I, lactate dehydrogenase and that the troponin I, elder age and SO2 values are linked to in-hospital mortality. Across nations, there is variation in case fatality rates and in predictors of mortality3. For example, data from Belgium indicated severity was associated with older age, renal insufficiency, higher lactate dehydrogenase and thrombocytopenia and obesity4. Patterns of severity from Chinese studies included higher age, male sex, higher Body Mass Index, hypertension, lower T lymphocyte and B lymphocyte count, higher white blood cell count, higher D2 dimer, procalcitonin, CRP and aspartate aminotransferase. Among these variables age and weight appeared to be independent risk factors for disease severity5. Importantly, identifying these risk factors did not significantly change our understanding of the COVID-19 pandemic nor did it facilitate a reduction in mortality.

In many populations the severity of COVID-19 is sex dependent (males > females)6. The excess of male deaths has led to two opposing suggestions: (i) The androgen-driven COVID-19 pandemic theory7,8, and (ii) the male hypogonadism theory9. With regard to support for the former, viral entry to cells is androgen dependent, involving priming of the spike proteins and cleaving of angiotensin converting enzyme 2 (ACE2). Both processes are facilitated by trans-membrane protease, serine 2 (TMPRSS2)10. Androgen receptor activity is a requirement for the transcription of the TMPRSS2 gene, suggesting that testosterone facilitates SARS-CoV2 cell entry11. Thus the androgen-driven COVID-19 pandemic theory postulates that high mortality from SARS-CoV2 in men is related to hypergonadism. In contrast, proponents of the male hypogonadism theory point to theory-inconsistent relationships between testosterone and COVID-19 in males9. Thus, in men COVID-19 mortality rates increase with age but testosterone levels decrease12. The male hypogonadism theory gave a rationale for the analyses conducted by Manning and Fink9 who considered national values of digit ratios, in this case 2nd to 4th digit ratio, in relation to Covid-19 case fatality rates (CFR’s). The relative lengths of the second digit and fourth digit (digit ratio or 2D:4D) is sexually dimorphic (2D:4D males < 2D:4D females). It has been suggested that 2D:4D is a biomarker of prenatal sex steroids exposure, i.e. low 2D:4D correlates with high prenatal testosterone and low oestrogens, while high 2D:4D correlates with low foetal testosterone and high oestrogen. There is considerable support for the link between 2D:4D and prenatal sex steroids13 but for a contrasting view see McCormick & Carre, 202014. Manning and Fink found that nations with high CFR’s had high mean male 2D:4D9, thus supporting the hypogonadal theory (in support see Sahin, 2020 and for a critical view see Jones et al., 202015,16). With regard to right–left asymmetry of 2D:4D, i.e. directional asymmetry of 2D:4D, (Δ r–l 2D:4D) is also thought to be a negative correlate of high prenatal testosterone and low prenatal oestrogen9,17,18,19,20,21. In general unsigned asymmetries (such as that of single digit R–L asymmetries) may reflect developmental instability related to postnatal stressors including sex steroids and to correlates of low socio-economic status such as poor nutrition22,23.

Sex differences in digit ratios, with males < females, are present across a number of digits24,25,26,27. Here we focus on six ratios from digits 2 to 5, i.e. 2D:3D, 2D:4D, 2D:5D, 3D:4D, 3D:5D and 4D:5D (digit 1 is difficult to measure accurately). There is considerable evidence that prenatal sex steroids have an effect on 2D:4D. However, effect sizes for 2D:4D are likely to be linked to other ratios that share 2D or 4D. Right 2D:4D is stable during growth in children and adolescents supporting the contention that it retains information pertaining to prenatal sex steroids. However, other ratios, such as those that include 5D (and in particular 3D:5D), show sex differences (males < females) but are highly unstable during growth in children and adolescents. This instability is present in both hands but is expressed most intensely in the left hand24,27. The difference in stability across right and left hands suggests that R–L differences in ratios may contain important information which pertains to developmental instability rather than effects of prenatal sex steroids. Therefore, for each ratio we consider values from the right hand, left hand, Δ right–left (directional asymmetry) and |(right–left)|(FA). The purpose of this preliminary report was to focus on associations between digit ratios and severity of COVID-19, as evidenced by hospitalization of patients.

Following from the across-nation correlations between digit ratio and CFR’s we suggest that, in comparison to controls, patients hospitalized for COVID-19 will have: (i) high right and left hand digit ratios, and high Δ right–left directional asymmetry, indicating exposure to low prenatal testosterone and high prenatal oestrogen, and (ii) high |(right–left)| unsigned asymmetry (FA), indicating heightened levels of developmental instability arising from stressors such as pubertal sex steroids. With regard to these predictions we emphasize that there is potential for considerable inter-correlation between digit ratios. In this regard, 2D:4D has been shown to exhibit developmental stability while 3D:5D is particularly unstable during development24,27. Therefore, the patterns associated with 2D:4D and 3D:5D are least likely to be affected by inter-correlations between digit ratios. Thus, 2D:4D may contain information concerning prenatal influences and 3D:5D information concerning postnatal effects of developmental instability. Consequently, we suggest these two digit ratios should be the focus of greatest attention.


Participants were recruited from a Department of Infectious Diseases and Liver Diseases of a Medical University. All consecutive patients with diagnosed COVID-19 who were hospitalized in the Department due to the severe or high risk of severe COVID-19 were included. During a first wave of the Covid-19 pandemic (March–August 2020) there were 54 (28 men and 26 women) patients who met the study criteria (Inclusion Criteria: 1. admitted to hospital because of Covid-19, 2. positive PCR test, 3. conscious and able to give informed written consent for participation; Exclusion Criteria: 1. unconscious, unable to give written consent for participation, 2. Covid-19 positive patients hospitalized because of other than Covid health issues, pregnant women, children (< 18 years), patients after transplantations, during immunotherapy and with renal failure requiring dialyses). Of these, there were 51 for whom the right hand ratios could be measured, 52 for the left hand and 49 for whom R–L measurements were possible (one hand only was available for measurement for 5 patients due to a hand injury and/or finger contractures). The protocol of the study included a clinical questionnaire based on medical records (age, symptoms) severity of the disease (scale 0–4; 0 -no symptoms, 1—mild, 2—medium, 3—severe, 4—critical), length of hospitalization and oxygen therapy, days in intensive care unit, concomitant diseases, history of smoking and occupational exposure, and laboratory test results (white blood count, fibrinogen, d-dimers, platelets count, oxygen saturation, procalcitonin) and anthropometric measurements.

Controls, 47 women (mean age 51.3 ± 16.1 years) and 53 men (mean age 52.2 ± 14.4 years) were recruited from a Plastic Surgery Out-patient Clinic (approximately 80% of the sample) and among other volunteers after the first wave of COVID-19. We consider our sample to be representative of the general population. Thus, the Out-patient Clinic is state-funded, attendees are from a variety of backgrounds and ages, and they present with a variety of needs such as removal of scars, moles and eyelid disorders (ptosis, ectropion, entropion). We did not include women after breast cancer who come for breast reconstruction, patients with skin cancer, post-bariatric patients, any patient who has immunosuppression. Controls were included based on a negative history of COVID-19 (non-infected or non-symptomatic subjects). One woman reported injury of the 3rd finger of the left hand and was included in the study after exclusion of this finger measurement. All the participants were White (based on patients’ medical data and controls recruitment).

Ethical statement

The protocol was agreed by the Bioethical Committee of the Medical University of Lodz (RNN/152/20/KE). All methods were performed in accordance with the relevant guidelines and regulations. Written informed consent was obtained from all participants.

Hand images

With regard to the measurement of digit ratio, our preference would have been for direct measurement of fingers. However, it was difficult to measure digit length directly from the hands of the patients because many of them were very ill and measurers were hampered by personal protective equipment. Moreover, direct digit measurement requires a period of time during which the patient and measurer are in close proximity. This is to be avoided with an infectious viral agent. Indirect methods such as the use of photocopies or scanners, give a permanent record of digit lengths. Against this, it was felt that the use of photocopiers or scanners was not appropriate as repeated use of such machines may result in cross infection resulting from virus particles being left on surfaces. Moreover, in comparison to directly measured digits, indirect images yield lower 2D:4D ratios24,28,29 with magnitudes that may vary by sex and hand30. These effects may extend to asymmetries also, and the accuracy of asymmetries measured from photocopies has been questioned31. Therefore, it was decided to photograph the hands of the patients. Typically, the patient was sitting up in bed and he/she was instructed to place their hands horizontally with the palms uppermost, the digits straight and together. In order to minimize inconvenience to the patient it was decided not to use a tripod with the camera. Rather, the experimenter held the camera approximately 30 cm above the patient’s hand. This protocol was felt to be appropriate as it would minimise the amount of proximity necessary between experimenter and patient. Moreover, it gives a permanent image of the supine hand which did not involve potential distortions resulting from digit contact with glass surfaces. It is to be noted that the relative lengths of digits within a hand can be obtained in this way but R–L contrasts of absolute measures of digit length are likely to be unreliable as they will be influenced by small vertical differences in distance between hand and camera. Photographs were checked for definition at the tips of the digits and at the metacarpophalangeal crease at the base of the digits. A second photograph was taken if the first was not deemed to be of sufficient quality.


Eight measurements were taken from patients’ and controls’ hand photographs: second, third, fourth and fifth digits’ lengths (2D, 3D, 4D and 5D) (right (R) and left hand (L)). On the basis of the these parameters the following ratios were calculated: 2D:3D, 2D:4D, 2D:5D, 3D:4D, 3D:5D, 4D:5D for the right (R) and left (L) hand (D length [mm]/D length [mm]) in addition to the ratios’ directional asymmetries (right ratio–left ratio: Δ2D:3D, Δ2D:4D, Δ2D:5D, Δ3D:4D, Δ3D:5D, Δ4D:5D) and unsigned asymmetries (FAs) (|(right–left)|). We also calculated two composite asymmetries by summing (i) all six (|(right–left)|) asymmetries and (ii) and asymmetries for the “independent” ratios of 2D:4D and 3D:5D. We refer to the latter as the “Clinical Composite Asymmetry” in the Results section. All measurements (in patients and controls) were made twice by AKT using the GNU Image Manipulation Program (GIMP) version 2.10.20. For a subset of measurements, a sliding calliper was used directly on the image of the fingers on the computer screen (by JTM). Measurements were performed on the palmar side of the hand using anthropometric points lying on the digit axis: pseudophalangion—the most proximal point in the finger metacarpophalangeal crease, dactylion—the most distal point on the fingertip32. There was high repeatability of digit ratios within and between observers. The final ratios were calculated as a mean of two ratios obtained from the GIMP program. These ratios were used in the further analysis of the data.

Statistical analysis

Analysis was conducted on the differences in the digit ratios and their directional (right–left) and unsigned [|(right–left)|] asymmetries between patients hospitalized due to Covid-19 and controls. The normality of distribution of the tested variables was examined (using Shapiro–Wilk test) and the homogeneity of variances was checked (using the Bartlett test). With both assumptions met we applied univariate t-tests for differences between means in addition to two-way analysis of variance (ANOVA). If any of these assumptions were not met then non-parametric tests were used. Logistic regression was used to evaluate the relationship between the asymmetry index being the sum of the unsigned asymmetries of the ratios of the largest effect sizes estimated with omega-squared for ANOVA (“Clinical Composite Asymmetry”) and the risk of hospitalization due to Covid-19. Finally, logistic regression model included the following variables: the sum of asymmetries of 2D:4D and 3D:5D (dependent variable) and the group (patient vs. control) and sex (independent variables). Effect size for inter-group differences was evaluated with Cohen’s d for t-tests and omega-squared (ω2) for ANOVA. The interpretation of descriptors of magnitude for the former were small 0.20, medium 0.50 and large 0.80 and for ω2 > 0.01 —weak, > 0.06—medium, > 0.14—strong effect. The probability of p < 0.05 was accepted as a level of significance.


Characteristics of Covid-19 patients

Among 54 patients there were 28 men (mean age 54.7 ± 14.7 years) and 26 women (mean age 59.3 ± 18.2 years). The group of patients did not differ in age and frequency of males and females from the controls (F = 1.085; p = 0.299). Specific characteristics (i.e. BMI, comorbidity, smoking status) and Covid-19 symptoms and severity are shown in Table 1.Table 1 Characteristics of patients hospitalized because of Covid-19.Full size table

Reliability of measurements

First we checked intra-observer reliability for all twelve ratios (ratio 1 versus ratio 2) for observer AKT. The coefficient of reliability for raw measurements (R) ranged from 96.07% (for 3D:4D L) to 99.66% (for 2D:5D R). Intra-class correlation coefficients were also very high Table 2). Repeatability of signed asymmetries can be low because they contain the measurement error of four digits. However, for the signed asymmetries (R–L) and the unsigned asymmetries (|R–L|) the R ranged from 99.86% (for 2D:3D |R–L| to 99.97% (for 2D:4D R–L) also with high ICC’s (Table 2). Further analysis included mean values of ratio 1 and 2. Then, inter-observer reliability was checked (observer AKT versus observer JTM), for two ratios: 2D:4D R and 2D:4D L and their signed and unsigned asymmetries. Due to the high reliability between observers (2D:4D R: TEM = 0.0089, R = 99.66%, ICC = 98.09%; 2D:4D L: TEM = 0.0118, R = 98.91%, ICC = 97.93%; R–L: TEM = 0.0076, R = 98.66%, |R–L|: TEM = 0.0076, R = 96.13%, ICC = 96.43%) final analysis included data from AKT.Table 2 Technical error measurement (TEM) and the coefficient of reliability for raw measurements (R) for ratios and for R–L and |R–L| of six ratios for observer 1.Full size table

Digit ratios: patients vs. controls

There were no relationships between age and digit ratios in any of the twelve tests (values of r varied from − 0.14 for right 2D:3D to 0.1 for right 4D:5D, all p > 0.05).

Patient and control means and SD’s for six ratios and 12 effects (right and left ratios) are given in Table 3. Values of p and Cohen’s d are included from t-tests. There were five significant effects ranging from small to medium in magnitude. Four of these showed higher values in the patients compared to the controls, i.e. 3D:5D right d = 0.55, left d = 0.37; 4D:5D right d = 0.64, left d = 0.58. One effect showed mean patient < control (right 2D:5D d = 0.38). We note that all five significant effects were present in ratios that included 5D. Correction for multiple tests is inappropriate across Table 3 as the variables are not independent, i.e. the length of each digit is present in three ratios. We considered the effect of sex on these patient/control differences by performing two-factor ANOVA’s (independent variables: group [patients, controls], sex [males, females] with dependent variable digit ratio). All five remained significant (see effect sizes [ω2]), There were no effects of sex and no significant interactions (Table 4).Table 3 Patient and control means and SD’s for six digit ratios (2D:3D; 2D:4D; 2D:5D;3D:4D; 3D:5D; 4D:5D) and their signed and unsigned asymmetries.Full size tableTable 4 Differences in digit ratios and their asymmetries between patients and controls—(ANOVA) controlled for sex.Full size table

Digit ratio asymmetries: patients vs. controls

Two associations between age and asymmetry were significant (|R–L| 2D:4D, R = 0.17, p = 0.03 and |R–L| 4D:5D, R = 0.24, p = 0.03). However, there were no relationships between age and asymmetries in ten of the twelve tests (values of R varied from − 0.13 for R–L 2D:3D to 0.16 for |R–L| 2D:3D, all p > 0.05).

There were no significant differences in directional asymmetries (R–L) between patients and controls (Tables 3 and 4). There is some evidence in the literature that directional asymmetry of 2D:4D shows sex differences (males < females). Therefore we checked for directional asymmetry (deviations from a mean of zero) in (R–L) in patients and controls for all six ratios split by sex. For male patients (n = 23) one-sample t-tests with mean set at zero showed there were no significant deviations from zero in any ratio (means varied from 0.002 for 3D:4D to 0.049 for 2D:5D, all p > 0.05). For female patients (n = 26), for five ratios means varied from − 0.014 for 2D:4D to 0.031 for 3D:5D, all p > 0.05. For female 4D:5D there was directional asymmetry with mean of 0.034, t = 2.20, p = 0.04. With regard to controls (males n = 53, females n = 47) there was a similar pattern with evidence of directional asymmetry in 4D:5D (males: mean = 0.018, t = 2.44, p = 0.02 and females: mean = 0.016, t = 2.31, p = 0.03). For the remaining ratios means varied from − 0.006 to 0.010, all p > 0.05. Therefore, there was no evidence of significant directional asymmetry in male and female mean (R–L) ratios with the exception of 4D:5D which showed some evidence of higher ratios in the right hand compared to left hand. This suggests that the ratios we consider here (with the exception of 4D:5D) have a mean that does not significantly deviate from zero, i.e. they have the properties of ideal fluctuating asymmetry.

With regard to unsigned asymmetries (|R–L|), the distributions are “half-normal”. It may be that t-tests of means for patients versus controls are robust enough to give meaningful p values. However, in order to consider such differences in a conservative manner we applied Mann–Whitney U tests. There were four significant effects (2D:4D, d = 0.74; 2D:5D, d = 0.66; 3D5D, d = 0.79; 4D:5D, d = 0.47) and all showed patients > controls. We note that three effects are for variables that include digit 5D. Summing the unsigned asymmetries across all six ratios we found this composite measure of asymmetry was higher in the patients compared to controls (d = 0.8). We then focused on |R–L| in the two “independent” ratios with the highest effect size (i.e. 2D:4D and 3D:5D) and found they were not correlated (r = − 0.047). A composite of these two variables, a “Clinical Composite Asymmetry showed the highest effect size of all with patients > controls, d = 1.04 (Fig. 1, Table 3).

figure 1
Figure 1

We further considered the effect of sex on these patient/control differences by performing two-factor ANOVA’s (independent variables: group [patients, controls], sex [males, females] with dependent variable digit ratio). There were high effect sizes (ω2) for |Δ2D:4D| = 0.115; |Δ2D:5D| = 0.105; |Δ3D:5D| = 0.155; |Δ4D:5D| = 0.055. The effect size for the “Clinical Composite Asymmetry” of 2D:4D and 3D:5D was 0.231. Logistic regression indicated that the “Clinical Composite Asymmetry”, regardless of sex, correlates with the risk of hospitalization due to Covid-19. The area under an ROC curve (AUC) is 0.787, which shows that this classifier is better than a random classifier (AUC = 0.5) with the cut-off point of 0.087. A “Clinical Composite Asymmetry” that is higher than 0.087 discriminates hospitalized patients (sensitivity—71% and specificity 75%) (Fig. 2). The risk of hospitalization in case of the index > 0.087 is 3.5 times higher than in those with lower “Clinical Composite Asymmetry” (OR 3.667).

figure 2
Figure 2

“Clinical Composite Asymmetry” did not correlate with Covid-19 severity (R = − 0.075, p = 0.61) or with length of hospitalization (R = 0.137, p = 0.35).


This study focused on associations between digit ratios and severity of COVID-19, as evidenced by hospitalization of patients. Our results indicate that digit ratios, and their asymmetries may be regarded as simple clinical markers of the possible risk of hospitalization due to Covid-19. Additionally, the study aimed to examine the role of prenatal sex steroids and that of postnatal developmental instability on the course of Covid-19. We have found evidence for digit ratio and digit ratio asymmetry differences between hospitalized patients with COVID-19 and controls. For digit ratios the magnitude of the effect sizes was small to medium (d = 0.3–0.64) and involved all ratios that included 5D, i.e. 2D:5D, 3D:5D and 4D:5D (patients > controls). There were no significant differences between patients and controls for directional (right-left) asymmetry. The largest effect sizes (medium to large) were found for measures of developmental instability, i.e. differences in unsigned asymmetries between patients and controls (patients > controls). These included 2D:4D (d = 0.74), and all ratios that involved 5D, i.e. 2D:5D (d = 0.66), 3D:5D (d = 0.79) and 4D:5D (d = 0.47). There are likely to be inter-correlations between these asymmetry effect sizes, for example 2D is present in two of them, as is 4D and 5D is present in three. The two largest effect sizes were found in ratios that may be independent of each other in the sense that they do not share digits (i.e. 2D:4D and 3D:5D). Summing the unsigned asymmetries of 2D:4D and 3D:5D gave a composite asymmetry with a large effect size (patient > control) of d = 1.04. Removing the effect of sex in a two-factor ANOVA had little effect on the magnitude of the effect size which remained large (ω2 = 0.231). We suggest that the unsigned composite asymmetry of 2D:4D and 3D:5D may have utility in identifying individuals that are of high risk for hospitalization resulting from COVID-19. Therefore, we have referred to it as a „Clinical Composite Asymmetry”. The utility of Clinical Composite Asymmetry as a classifier was characterized by AUC = 0.787 (good classifier). In addition, the optimum cut off point ≤ 0.087 was determined, for which sensitivity and specificity were 71% and 75% respectively with OR over 3.5. Regression analysis showed that the index > 0.087 may be a prognostic factor for hospital care for patients with Covid-19. However, to verify the prognostic value of the suggested index further studies based on larger populations in different ethnic groups are needed.

Much of the work concerning effects of prenatal sex steroids on digit ratio has concentrated on 2D:4D. However, effect sizes for 2D:4D are likely to be linked to other ratios that share 2D or 4D (i.e. 2D:3D; 2D:5D; 3D:4D; 4D:5D). The 3D:5D ratio has also been described as sexually dimorphic (males < females) and may show effects that are independent of 2D:4D24,25,26. Importantly, 3D:5D is not stable during development across age ranges from 2 to 18 years. Rather it shows a reduction with age which suggests that it may be influenced by postnatal production of androgens24. Comparisons between digit ratios of hospitalized patients versus controls gave small to medium effect sizes for 2D:5D, 3D:5D and 4D:5D. In so far as these digit ratios are influenced by sex steroids, this may be evidence for a link between severity of COVID-19 and prenatal (2D:4D) and postnatal (3D:5D) testosterone and oestrogen. Studies in humans and with an animal model (Golden Hamsters) have reported that SARS-CoV2 upregulates the enzyme CYP19A1 (oestrogen synthetase) leading to a profound reduction in testosterone and an increase in oestrogen in the lungs and other organs. Dysregulated sex hormones and interferon gamma (IFN-γ) levels are associated with critical illness in Covid-19 patients. In this regard, both male and female Covid-19 patients, present elevated oestradiol levels which positively correlates with IFN-γ levels (for humans33, for an animal model34). Manning and Fink9 reported that national values of male 2D:4D are positively related to national COVID-19 CFR’s. This led them to suggest that nations with high COVID-19 mortality have male populations that have experienced low prenatal testosterone relative to oestrogen.

However, it is more likely that the differences between patients and controls have arisen as the result of elevated levels of developmental instability in the former compared to the latter. Manning24 has considered the stability of all six digit ratios during growth between the ages of 2 years and 18 years. Right 2D:4D was stable but left 2D:4D was not. All ratios that included 5D showed growth-linked instability for both the right and left hands. We suggest that ratios that include 5D are „hotspots” for developmental instability that may be triggered by stressors that include rapid growth22,24. Recently a syndemic approach, which includes biological and social interactions for prognosis, treatment, and health policy, has been proposed. Interaction between infection with SARS-CoV-2 and an array of non-communicable diseases strongly associated with poverty, including obesity, hypertension, diabetes, cardiovascular and chronic respiratory diseases, and cancer is now considered. Moreover, syndemics are characterised by biological and social interactions between conditions and states, which increase one’s susceptibility to poor health outcomes35. In this respect, considering morphological signs of exposure to prenatal sex steroids and developmental instabilities (interaction between rapid early growth and stressors such as poor maternal and childhood nutrition) in patients with severe or fatal course of COVID-19 may give insight into the syndemic nature of Covid-19.

In contrast to right and left digit ratios, differences in the magnitude of digit ratio asymmetries between the right and left hand gave medium to large effect sizes. This was not apparent in directional asymmetries (R–L), perhaps because they comprise subtle deviations from perfect symmetry. Such asymmetries have been described as weakly sexually dimorphic for right-left 2D:4D (or Dr-l: with male Dr-l < female Dr-l36). Removing the signs from directional asymmetry (|R–L|) gave us variables that showed medium to high effect sizes in comparisons between patients and controls. Digit ratio (|R–L|) is a measure which is equivalent to asymmetry differences in digit length22. However, in this case we are dealing with R–L differences in morphological patterns involving two digits rather than differences between single digits of the right and left hands. It is not known whether |R–L| is sexually dimorphic across the six digit ratios. There is evidence that the phenotype of Dr-l is influenced by variation in the gene for the enzyme CYP19A1. Thus, Zhanbing et al. (2019) have reported a CYP19A1 single-nucleotide polymorphism (rs4775936) is related to variation in Dr-l in a Chinese sample36. CYP19A1 is important in the conversion of testosterone to oestrogen and SNP rs4775936 has been linked to the incidence of breast cancer. It may not be coincidental that up-regulation of CYP19A1 occurs in the lungs and other organs of COVID-19 patients leading to dysregulation of sex hormones (acute reduction in testosterone and an increase in oestrogen) and a marked increase in interferon gamma (IFN-γ) levels. Both are associated with critical illness in Covid-19 patients33,34. Further work is indicated to investigate patterns of age in digit ratio asymmetry (|R–L|). However, if it takes the form of single trait asymmetry, such as that of digit length asymmetry, it may show high levels in children which reduce with increasing age22. Associated with these age changes in digit asymmetry we find age dependent instability of all digit ratios that include 5D24. Such a pattern would suggest unsigned asymmetries of digit ratios involving 5D are sensitive correlates of developmental instability which may be negatively associated with immune system function. Such an interpretation is consistent with the view that increased asymmetry is the result of a combination of deleterious genetic and environmental factors and is defined as small, random deviations from perfect bilateral symmetry regarded as a measure of the developmental stability of the individual and phenotypic and genetic quality23.

We acknowledge our study has limitations: (i) Our sample size of 54 hospitalized patients is small. Obtaining good quality photographs of patients’ hands during hospitalization was sometimes challenging. A larger number would have been possible if the patient’s hands had been photocopied or scanned at discharge. However, with this latter methodology one risks missing severe cases of COVID-19 that are never discharged. We plan to extend our study, adding numbers of hospitalized patients and remeasuring digit lengths at discharge. With an increase in sample size we will consider relationships between clinical variables and digit ratios and their asymmetries. It is to be noted that within Table 4 we control for sex and report effect sizes for patients versus controls. With regard to the latter differences the effect sizes are medium to strong but none of the former are significant. This may be because sex differences in digit ratios have small to medium effect sizes. If we are correct in this we would expect that larger samples to show significant effects for sex in addition to differences between patients and controls. (ii) Additionally a further confounder may be the unknown infection status of the controls. They were recruited during the same time frame as patients hospitalized due to Covid-19 among other out-patient patients age-matched and based on their negative history of any symptoms of Covid-19 infection. Such appointment of controls may have resulted in a heterogeneity of this group such that they may have included non-infected individuals as well as non-symptomatic but infected or past-infected individuals. However, this does not invalidate the idea of this study which was focused on “markers” of symptomatology related to hospitalization. It would be beneficial to perform such analysis comparing symptomatic versus asymptomatic (or mildly symptomatic treated out-patients) but infected patients. However, this was not possible during the first waves of Covid-19 as there was no nation-wide testing in Poland (in general only symptomatic individuals were tested). In this regard our results may be biased by behavioral factors—i.e. the way participants prevented infection. (iii). We were not able to make comparisons between individuals who had been vaccinated and those who have not. This was because our data were obtained during the first wave of Covid-19, so none of the participants (patients and controls) had been vaccinated. In this regard, it would be of great interest to compare the efficacy of the vaccine in individuals with high and low values of the „Clinical Composite Asymmetry”. The prediction would be that vaccine efficacy would be low when „Clinical Composite Asymmetry” is high and high when „Clinical Composite Asymmetry” is low.

In conclusion, we have found differences in digit variables between patients hospitalized for COVID-19 and controls. Overall, our findings point to high levels of developmental instability in the former compared to the latter. Our focus was on six digit ratios and for each we considered right and left ratios and their asymmetries (signed and unsigned). We found differences in digit ratios between patients and controls that were focused on ratios that included 5D. The effect sizes were small to medium. Unsigned asymmetries of four digit ratios, including three that involved 5D, yielded medium to large effect sizes with patients > controls. The largest of these asymmetries were for 2D:4D and 3D:5D. A „Clinical Composite Asymmetry” for these two variables gave an effect size which may have some utility in identifying individuals who have experienced high developmental instability. Thus, this „Clinical Composite Asymmetry” may enable us to identify individuals who are likely to experience severe COVID-19 and those who may not.


  1. WHO-China Joint Mission, Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19) https://www.who.int/docs/default-source/coronaviruse/who-china-joint-mission-on-covid-19-final-report.pdf (2020).
  2. Mattiuzzi, C. & Lippi, G. Which lessons shall we learn from the 2019 novel coronavirus outbreak?. Ann. Transl. Med. 8, 48 (2020).CAS Article Google Scholar 
  3. Tahtasakal, C. A. et al. Could we predict the prognosis of the COVID-19 disease?. J. Med. Virol. 93, 2420–2430 (2020).Article Google Scholar 
  4. van Halem, K. et al. Risk factors for mortality in hospitalized patients with COVID-19 at the start of the pandemic in Belgium: A retrospective cohort study. BMC Infect. Dis. 20, 897 (2020) (Erratum in: BMC Infect. Dis. 20, 956 (2020)).Article Google Scholar 
  5. Yi, P. et al. Risk factors and clinical features of deterioration in COVID-19 patients in Zhejiang, China: A single-centre, retrospective study. BMC Infect. Dis. 20, 943 (2020).CAS Article Google Scholar 
  6. Guan, W. J. et al. Clinical characteristics of coronavirus disease 2019 in China. N. Engl. J. Med. 382, 1708–1720 (2020).CAS Article Google Scholar 
  7. Wambier, C.G., A. Goren, A. Ossimetha, G., & Nau, A.A. Qureshi, Androgen-driven COVID-19 pandemic theory. Preprint at https://doi.org/10.13140/RG.2.2.21254.11848 (2020).
  8. Wambier, C. G. & Goren, A. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection is likely to be androgen mediated. J. Am. Acad. Dermatol. 83, 308–309 (2020).CAS Article Google Scholar 
  9. Manning, J. T. & Fink, B. Understanding COVID-19: Digit ratio (2D:4D) and sex differences in national case fatality rates. Early Hum. Dev. 146, 105074 (2020).CAS Article Google Scholar 
  10. Heurich, A. et al. TMPRSS2 and ADAM17 cleave ACE2 differentially and only proteolysis by TMPRSS2 augments entry driven by the severe acute respiratory syndrome coronavirus spike protein. J. Virol. 88, 1293–1307 (2014).Article Google Scholar 
  11. Hoffmann, M. et al. SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor. Cell 181, 271–280 (2020).CAS Article Google Scholar 
  12. Ferrini, R. L. & Barrett-Connor, E. Sex hormones and age: A cross-sectional study of testosterone and estradiol and their bioavailable fractions in community-dwelling men. Am. J. Epidemiol. 147, 750–754 (1998).CAS Article Google Scholar 
  13. Swift-Gallant, A., Johnson, B. A., Di Rita, V. & Breedlove, S. M. Through a glass, darkly: Human digit ratios reflect prenatal androgens, imperfectly. Horm. Behav. 120, 104686 (2020).Article Google Scholar 
  14. McCormick, C. M. & Carre, J. M. Facing off with the phalangeal phenomenon and editorial policies: A commentary on Swift-Gallant, Johnson, Di Rita and Breedlove 2020. Horm. Behav. 120, 104710 (2020).Article Google Scholar 
  15. Sahin, H. A further analysis of Manning and Fink. Early Hum. Dev. 105121 (2020) (Online ahead of print).
  16. Jones, A. L., Jaeger, B. & Schild, C. No credible evidence for 2D:4D and COVID-19 outcomes: A probabilistic perspective on digit ratio, ACE variants, and national case fatalities. Early Hum Dev. 152, 105272 (2020).Article Google Scholar 
  17. Manning, J. T., Scutt, D., Wilson, J. & Lewis-Jones, D. I. The ratio of 2nd to 4th digit length: A predictor of sperm numbers and concentrations of testosterone, luteinizing hormone and oestrogen. Hum. Reprod. 13, 3000–3004 (1998).CAS Article Google Scholar 
  18. Manning, J. T. Resolving the role of prenatal sex steroids in the development of digit Ratio. Proc. Natl. Acad. Sci. U. S. A. 108, 16143–16144 (2011).ADS CAS Article Google Scholar 
  19. Manning, J. T. & Fink, B. Digit ratio and personality and individual differences. In The SAGE Handbook of Personality and Individual Differences (eds Zeigler-Hill, V. & Shackelford, T. K.) (SAGE Publications Ltd, 2018).Google Scholar 
  20. Manning, J. T. Digit Ratio: A Pointer to Fertility, Behavior, and Health (ed. Manning, J.T) 24–41 (Rutgers University Press, 2002).
  21. Manning, J. T. The Finger Ratio (ed. Manning, J.T) 5–10 (Faber and Faber, 2008).
  22. Wilson, J. M. & Manning, J. T. Fluctuating asymmetry and age in children: Evolutionary implications for the control of developmental stability in children. J. Hum. Evol. 30, 529–537 (1996).Article Google Scholar 
  23. Thornhill, R. & Møller, A. P. Developmental stability, disease and medicine. Biol. Rev. Camb. Philos. Soc. 72, 497–548 (1997).CAS Article Google Scholar 
  24. Manning, J.T. Sex differences and age changes in digit ratios: Implications for the use of digit ratios in medicine and biology. In Handbook of Anthropometry, 841–851 (Springer, 2012).
  25. Manning, J. T., Callow, M. & Bundred, P. E. Finger and toe ratios in humans and mice: Implications for the aetiology of diseases influenced by HOX genes. Med. Hypotheses. 60, 340–343 (2003).CAS Article Google Scholar 
  26. McFadden, D. & Shubel, E. Relative lengths of fingers and toes in human males and females. Horm. Behav. 42, 492–500 (2002).Article Google Scholar 
  27. Trivers, R., Manning, J. & Jacobson, A. A longitudinal study of digit ratio (2D:4D) and other finger ratios in Jamaican children. Horm. Behav. 49, 150–156 (2006).Article Google Scholar 
  28. Ribeiro, E., Neave, N., Morais, R. N. & Manning, J. T. Direct versus indirect measurement of digit ratio (2D: 4D) a critical review of the literature and new data. Evol. Psychol. 14, 1474704916632536 (2016).Article Google Scholar 
  29. Fink, B. & Manning, J. T. Direct versus indirect measurement of digit ratio: New data from Austria and a critical consideration of clarity of report in 2D:4D studies. Early Hum. Dev. 127, 28–32 (2018).Article Google Scholar 
  30. Manning, J. T. et al. Photocopies yield lower digit ratios (2D:4D) than direct finger measurements. Arch. Sex. Behav. 34(3), 329–333 (2005).Article Google Scholar 
  31. Manning, J. T., Fink, B., Neave, N. & Caswell, N. The second to fourth digit ratio and asymmetry. Ann. Hum. Biol. 33, 480–492 (2006).Article Google Scholar 
  32. Kasielska-Trojan, A. & Antoszewski, B. Can digit ratio (2D:4D) studies be helpful in explaining the aetiology of idiopathic gynecomastia?. Early Hum. Dev. 91, 57–61 (2015).Article Google Scholar 
  33. Schroeder, M. et al. Sex hormone and metabolic dysregulations are associated with critical illness in male Covid-19 patients. Preprint at https://doi.org/10.1101/2020.05.07.20073817v2 (2020).Article Google Scholar 
  34. Stanelle-Bertram, S., et al. SARS-CoV-2 induced CYP19A1 expression in the lung correlates with increased aromatization of testosterone-to estradiol in male golden hamsters. Preprint at https://www.researchsquare.com/article/rs-107474/v1 (2020).
  35. Horton, R. Offline: COVID-19 is not a pandemic. Lancet 396, 874 (2020).CAS Article Google Scholar 
  36. Zhanbing, M. et al. Association of CYP19A1 single-nucleotide polymorphism with digit ratio (2D:4D) in a sample of men and women from Ningxia (China). Early Hum. Dev. 132, 58–65 (2019).Article Google Scholar 

Download references

Author information


  1. Plastic, Reconstructive and Aesthetic Surgery Clinic, Institute of Surgery, Medical University of Lodz, Kopcinskiego 22, 90-153, Lodz, PolandA. Kasielska-Trojan & B. Antoszewski
  2. Applied Sports, Technology, Exercise, and Medicine (A-STEM), Swansea University, Swansea, UKJ. T. Manning
  3. Department of Infectious and Liver Diseases, Medical University of Lodz, Lodz, PolandM. Jabłkowski & J. Białkowska-Warzecha
  4. Department of Gynecology and Reproductive Medicine, Karolinska University Hospital, Stockholm, SwedenA. L. Hirschberg


Conception or design of the work: J.T.M., A.K.T., B.A., A.H. Data acquisition: J.B.W., M.J. Data analysis: J.T.M., A.K.T. Interpretation of data: J.T.M., A.K.T., B.A. Drafting and revising the ms: J.T.M., A.K.T., B.A., A.H., J.B.W., M.J. All authors have approved the submitted version.

The length of your fingers may determine how sick you get from COVID-19

Authors: Chris Melore Studyfinds.org March 28, 2022

Your risk of ending up in the hospital with COVID-19 may literally be in your own hands. A new study finds finger length displays a link to a person’s sex hormone levels. What does this have to do with COVID-19? Researchers at Swansea University say a patient’s testosterone levels play a key role in how sick they get after infection.

Previous studies show that having a longer ring finger is a sign of higher testosterone levels in the womb. On the other hand, a longer index finger signals higher levels of estrogen. Typically, men have longer ring fingers and women have longer index fingers.

The new study examined this link between the sex hormones before birth and during puberty and the rate of COVID hospitalizations. Their findings reveal that people with “feminized” short little fingers in comparison to their other digits end up suffering more severe cases of COVID-19. Moreover, people who have larger size differences between the fingers on their left and right hands are at even greater risk.

The link between testosterone and coronavirus

Although most people only experience mild COVID symptoms, the elderly and men are more likely to have a severe case that requires urgent care. This has led scientists to wonder if a man’s testosterone levels play a role in disease severity.

One theory is that high testosterone levels cause COVID to worsen. However, another study links low levels in elderly men to a severe case of the virus.

To figure out which is right, the team examined the size ratios of the 2nd, 3rd, 4th, and 5th digits on the hands of over 150 people. Fifty-four of these individuals were COVID-19 patients, while the others served as a healthy control group.

Specifically, the results show bigger differences between the 2D:4D and 3D:5D ratios on each person’s hands had a connection to a more severe case of COVID-19.

“Our findings suggest that COVID-19 severity is related to low testosterone and possibly high estrogen in both men and women,” says Professor John Manning in a university release.

“’Feminized’ differences in digit ratios in hospitalized patients supports the view that individuals who have experienced low testosterone and/or high estrogen are prone to severe expression of COVID-19. This may explain why the most at-risk group is elderly males,” the researcher continues. “This is significant because if it is possible to identify more precisely who is likely to be prone severe COVID-19, this would help in targeting vaccination. Right-Left differences in digit ratios (particularly 2D:4D and 3D:5D) may help in this regard.”

Could testosterone drugs defeat the pandemic?

Currently, study authors say there are several trials examining anti-androgen (testosterone) drugs which may help treat COVID-19. At the same time, scientists are also looking at testosterone as a possible anti-viral medication against COVID.

“Our research is helping to add to understanding of Covid-19 and may bring us closer to improving the repertoire of anti-viral drugs, helping to shorten hospital stays and reduce mortality rates,” Prof. Manning adds. “The sample is small but ongoing work has increased the sample. We hope to report further results shortly.”

This isn’t the first study to link finger length to seemingly unrelated topics. A previous study connected children’s finger length to their mother’s income as well as vulnerability to childhood illnesses.

The study is published in the journal Scientific Reports.

A Review of COVID-19 Chilblains-like Lesions and Their Differential Diagnoses

Authors: Sachdeva, Muskaan BHSc; Mufti, Asfandyar MD, BMSc; Maliyar, Khalad BA; Lara-Corrales, Irene MD, MSc; Salcido, Richard MD, EdD, FAAPMR; Sibbald, Cathryn MD, FRCPC



To familiarize wound care practitioners with the differential diagnoses of chilblains-like lesions that could be associated with the complications of COVID-19.


This continuing education activity is intended for physicians, physician assistants, nurse practitioners, and nurses with an interest in skin and wound care.


After participating in this educational activity, the participant will:

1. Identify the population most often affected by COVID toes.

2. Select the assessments that help differentiate the various conditions that cause chilblains-like lesions.

3. Choose appropriate treatment options for the various conditions that cause chilblains-like lesions.

This review article focuses on the pathogenesis, clinical features, and diagnostic testing of the common pathologies that can manifest as chilblains-like lesions. These differentials include “COVID toes,” Raynaud phenomenon, acrocyanosis, critical limb ischemia, thromboangiitis obliterans, chilblains associated with lupus erythematosus, and idiopathic chilblains. The authors present a helpful mnemonic, ARCTIC, to assist clinicians in recognition and diagnosis.


The purpose of this educational activity is to familiarize the wound care practitioner with the differential diagnoses of chilblains-like lesions (CLLs) that may be associated with complications of COVID-19. Recently, physicians and other wound care practitioners have encountered CLLs as a potential cutaneous manifestation of SARS-CoV-2, which present as violaceous macules, papules, plaques, or nodules on extremities (ie, acral lesions).1 Also called perniosis, CLLs present in three forms: related to COVID-19 infection, lupus associated, and idiopathic. A preliminary systematic review of 46 studies (including case reports and series) found acrocutaneous lesions to be the most commonly reported skin manifestation of COVID-19 infection.2

Usually chilblains are a result of an inflammatory response to cold temperatures. A typical presentation consists of tender, pruritic erythematous or violaceous nodules located symmetrically on the dorsal aspect of the fingers, toes, ears, and nose. Less commonly, they can occur on the thighs and buttocks, usually with many local and systemic conditions, and these nodules have been described as CLLs (Figures 1–3).2

Figure 1: COVID TOESImages reprinted with patient consent.
Figure 2: CHILBLAINS-LIKE LESIONS ON THE TOESImage provided by Stedman’s.
Figure 3: CHILBLAINS-LIKE LESIONS ASSOCIATED WITH RAYNAUD PHENOMENONReprinted with permission from Vick E, Sharma D, Elwing J. Raynaud phenomenon in antisynthetase syndrome treated with epoprostenol. J Clin Rheumatol 2021;27(1):e10-1.

Although CLLs are common, many other conditions present with acral lesions and should be considered in differential diagnoses (Table).3–31 This article focuses on the pathogenesis, clinical features, and diagnostic testing of the common pathologies that can manifest as CLLs or be considered mimickers. These differentials include “COVID toes,” Raynaud phenomenon, acrocyanosis, critical limb ischemia (CLI), thromboangiitis obliterans, chilblains lupus erythematosus (CHLE), and idiopathic chilblains.Table – DIFFERENTIAL DIAGNOSIS OF CHILBLAINS-LIKE LESIONS

DiseaseRisk FactorsClinical FeaturesDiagnostic Factors
COVID toes3–12YouthFailure to wear warm footwearLow BMIMild or no systemic symptomsAppearance rather late during (suspected) infectionCutaneous, red-violaceous, edematous, rarely necrotic, itchy CLL on the toesPhysical examinationHistory of exposureHistopathologic, immunohistochemical, and immunofluorescence studiesSpike-like circular structures in biopsy
Raynaud disease13,14Age > 30 yFemaleCold climateFamily historyWhitening/bluing/reddening of hands and feetAntinuclear antibody testESR
Acrocyanosis15,16Cold climateOutdoor occupationLow BMIAnorexia nervosaCoolness and violaceous dusky discolorations of hands and feetUlceration and gangrene of the fingersIrregular and brittle nailsClammy extremitiesSusceptibility to cooling and painErythrocyanosis, perniosis, livedo reticularis, persistent and painless cyanosis of extremities, local hypothermia (extremities)Elastic infiltration of the integumentComplete hemogramAntinuclear factorRheumatoid factor, cardiolipin antibody, lupus antibody, metabolic screeningCapillary oximetryMRI, CT, capillaroscopy
Critical limb ischemia17–20AgeSmokingBMIFamily history of atherosclerosis or claudicationDiabetes, hypertension, hypercholesterolemiaGangreneNonhealing refractory woundsPain at restAnkle-brachial indexToe systolic pressure and transcutaneous oxygen pressure, auscultationDoppler ultrasound, CT angiography, magnetic resonance angiography, angiogram
Thromboangiitis obliterans21–26Age < 45 yMaleSmokingSkin ulcerations and gangrene of the digits, superficial thrombophlebitis, and Raynaud phenomenonClaudication, pain, numbness, and/or tingling in the extremitiesEarly phase: polymorphonuclear leukocytes, microabscesses, and multinucleated giant cells; Intermediate phase: progressive organization of the thrombus in the arteries and veins; End-stage: only organized thrombus and fibrosis are found in the blood vesselsComplete blood cell count, serum creatinine concentrations, fasting blood sugar levels, and ESRTests for antinuclear antibody, rheumatoid factor, serologic markers for CREST syndrome and sclerodermaScreening for hypercoagulabilityLiver function testsTransthoracic or transesophageal echocardiography, arteriography, CT, MRI
Chilblains associated with lupus erythematosus27–29Cold climateFamily historyPapuloerythematous lesionsPurplish and plaque-like infiltratesFissural hyperkeratosis and small ulceration appears on hands, fingers, soles of feet, may affect other areas such as trunk, ear, noseHistopathologic examination (fibrin deposition) or indirect immunofluorescence studies, cryoglobulin and cold agglutinin studiesResponse to antilupus erythematous therapy, other antibody studiesLatex agglutination testHypergammaglobulinemia
Idiopathic chilblains30,31Cold climateLow BMIPainful erythematous or purplish papules, CLL on hands and feet or less frequently, on nose, ear, thighC-reactive proteinESRHistologically presents without fibrin deposition

Abbreviations: BMI, body mass index; CLL, chilblains-like lesions; CREST, calcinosis cutis, Raynaud phenomenon, sclerodactyly, and telangiectasia; CT, computed tomography; ESR, erythrocyte sedimentation rate.


“COVID toes” are characterized by erythematous or violaceous, edematous, rarely necrotic papules or plaques most commonly on the toes (Figure 1).3,4 Accounting for 19% to 38% of all cutaneous lesions associated with COVID-19, this presentation has been reported in several European countries, as well as the Middle East and is most common in adolescents or young adults.3–7 Estimates on the prevalence of CLLs in patients with COVID-19 range from 2% to 20%.8 Typically occurring as a late manifestation of infection, the majority of patients test negative for COVID-19 polymerase chain reaction at the time of presentation.

The etiology and pathophysiology of COVID toes have not been confirmed, but some clinicians suggest they represent a sequela of mild infection with appropriate inflammatory response. A low body mass index and lack of warm indoor footwear during lockdown have also been associated with COVID toes, and these may be contributory or predisposing factors.6 The lesions last on average 12 days, but may persist for several weeks.7 Spiky circular structures have been identified in several biopsies of COVID toe lesions, suggestive of the spike glycoproteins present in COVID-19 viral particles.9–12 However, the relationship with COVID-19 has not been convincingly proven to date.9

The clinical diagnosis of COVID toes is based on physical examination and history of possible exposure to COVID-19.9 Lesions respond to topical steroids and environmental modification.5 Continued research is needed to assess COVID toes and their association with SARS-CoV-2.


Raynaud phenomenon is characterized by intermittent peripheral vasoconstriction to the fingers or toes that manifests as a triphasic segmental color change of white to blue to red (Figure 3).13 This phenomenon is attributed to sympathetic nervous system stimulation by emotional stress, anxiety, and cold temperatures.13 The prevalence of Raynaud phenomenon ranges from 3% to 20% in women and 3% to 14% in men.14

Primary Raynaud disease is not associated with systemic disease and is typically milder, with less frequent episodes, and most commonly presents between the ages of 15 and 25 years, with a significant female predominance (roughly 20:1).13 Conversely, secondary Raynaud phenomenon is associated with concurrent systemic disease, including connective tissue diseases, structural vascular anomalies, neurologic diseases, hyperviscosity syndromes, lymphoproliferative diseases, and exposures to medications and physical environmental triggers. Patients are typically older, with a less significant female predominance (4:1), and episodes are often more frequent.13

Symptoms include skin pallor or cyanosis of the fingers, cold digits, and numbness and soreness of digits that last approximately 20 minutes following stimulation.13 An episode typically begins with decreased blood flow to the acral regions with digits turning white in color, after which there is a cyanotic phase where the remaining oxygen in the blood left in the digits is consumed.13 The episode ends with a red phase where blood flow is restored.13 The mechanism includes a temporary constriction leading to cyanosis that is very similar to changes observed with chilblains.13

When diagnosing primary Raynaud disease or secondary Raynaud phenomenon, clinicians must assess for the presence of digital ulcerations and intensity of pain during vasoconstrictive episodes, while identifying any provoking or mitigating factors.13 In addition, nailfold capillary microscopy can be used to confirm the diagnosis by visualizing increased capillary size, often a lower density of capillaries, hemorrhage, and avascular areas in the fingers.13 Currently, careful patient history and screening for any possible associated disease are the most reliable methods for diagnosis.13 Obtain a collagen vascular disease screen and erythrocyte sedimentation rate.13 Selective patients may benefit from cryoglobulin testing, chest radiograph, pulmonary function tests, or ECG.

Unfortunately, there is no cure, and most treatment strategies focus on preventing and minimizing the severity of symptoms.13 Treatment may include insulated gloves, activated thermal warming devices, nitroglycerin gel or patches, calcium-channel blockers, phosphodiesterase E5 inhibitors (eg, sildenafil 50 mg twice a day, 200 mg once a day, or tadalafil 20 mg daily), or prostacyclin analog infusions.14


Acrocyanosis usually presents with violaceous or blue discoloration of the hands, feet, or face from constriction of the blood vessels (Figure 4).15 Acrocyanosis has a prevalence of 12% to 13% and is more common in children and young adults (usually in patients younger than 30 years) with a predominance among women (roughly 6–8:1).15,16 There are two types of acrocyanosis: primary (idiopathic) acrocyanosis without associations and acrocyanosis secondary to a predisposing condition or medication.15

Figure 4: ACROCYANOSIS OF THE FINGERTIPSImage provided by Stedman’s.

Current hypotheses propose that chronic vasospasm of small arteries combined with dilatation in the capillaries and postcapillary venules leads to cyanosis, sweating, and pain.15 Genetic defects affecting vascular musculature, erythrocyte flexibility, platelet adhesiveness, and other factors resulting in plasma hyperviscosity result in compromised blood flow and clinical manifestations.15,16

Common symptoms include cold and clammy extremities with cyanotic discoloration, enhanced susceptibility to cooling and pain, ulceration and gangrene of the fingers, irregular and brittle nails, erythrocyanosis, perniosis (bluish red discoloration of the skin), livedo reticularis, persistent and painless cyanosis of extremities, local hypothermia (extremities), and elastic infiltration of the integument.15 Symptoms often worsen with exposure to cold temperatures.16

Acrocyanosis is diagnosed based on the presence of cyanosis and local hypothermia of the extremities, sweatiness, and elastic infiltration.15 Capillary oximetry and capillaroscopy are often sufficient to rule out differential diagnoses.15 In other cases, complete blood count; MRI and computed tomography; and screening for antinuclear factor, rheumatoid factor, cardiolipin antibody, lupus antibody, and other metabolic antibodies help to identify predisposing or associated conditions.16


Critical limb ischemia is characterized by acute limb pain at rest, nonhealing ulcers, gangrene, and refractory wounds that can resemble CLLs17 and progresses over the span of months to years.18 Among Americans, there are 500 to 1,000 new cases of CLI per million people each year.19 The prevalence of CLI is 12% among adults, with men affected slightly more than women.19 It primarily presents in patients with atherosclerosis but can also be observed in patients with Buerger disease and some forms of arteritis.19

The pathogenesis of CLI involves arterial obstruction leading to impaired perfusion of peripheral tissues.19 The resulting poor oxygenation and nutrient delivery to tissues can delay healing of foot sores, ulcers, or wounds with tissue necrosis.17 Ulcers are often nonhealing or associated with necrosis or gangrene. Other signs and symptoms may include numbness; claudication; pain on limb elevation; toenail thickening; diminished pulses in the legs or feet; and shiny, smooth, or dry skin of the feet.17

In addition to clinical presentation, objective measures are used to augment the diagnosis of CLI, which is based on bedside vascular assessment and ancillary imaging techniques.17 In general, to fulfill a diagnosis of CLI, patients must exhibit all of the following observations: (1) ischemic rest pain or ulcer necrosis, (2) need for amputation if blood flow does not improve, and (3) objective ischemia meeting hemodynamic standards (ie, toe pressure of <50 mm Hg and a transcutaneous partial oxygen pressure <30 mm Hg).17 Bedside investigations to assess vascular flow include ankle-brachial pressure index, toe systolic pressure, and transcutaneous oxygen pressure. In addition, duplex segmental lower leg Doppler can assist in diagnosis. Doppler waveform changes from triphasic to biphasic to monophasic and then to stenotic waveforms can help identify arterial blockage sites.17 Last, CT angiography, magnetic resonance angiography, and (if indicated) angiography can also be used.17 If left untreated, CLI may lead to amputation of the affected limb.19,20


Thromboangiitis obliterans, also known as Buerger disease, is a nonatherosclerotic inflammatory disease strongly correlated with smoking tobacco. This condition is caused by segmental occlusive inflammation of arteries and veins in the extremities (Figure 5).21 The prevalence of thromboangiitis obliterans among patients with peripheral arterial disease ranges from 0.5% to 5.6% in Western Europe, 45% to 63% in India, 16% to 66% in Korea and Japan, and 80% in Israel (in Jews of Ashkenazi ancestry).21 Typically, this disease presents in young individuals with a smoking history (an average of 16 pack-years) and manifests as distal extremity ischemia, digit ulcerations, or digit gangrene, more commonly on the hands and feet.22 Patients may experience numbness and tingling, claudication, skin ulcerations, gangrene, superficial thrombophlebitis, Raynaud phenomenon, and pain in affected extremities.22 The accompanying CLLs are characterized by rubor or cyanosis, a discoloration termed “Buerger’s color,” and pain.23

Figure 5: THROMBOANGIITIS OBLITERANS, ALSO KNOWN AS BUERGER DISEASE, OCCURRING ON THE FEETReprinted with permission from D’Alessio I, Settembrini AM, Romagnoli S, Di Luca G, Domanin M, Gabrielli L. Successful fat grafting in a patient with thromboangiitis obliterans. Adv Skin Wound Care 2019;32(12):1-4.

Although the etiology of thromboangiitis obliterans is unclear, it is believed to be an immune-mediated endarteritis.23 Interestingly, only endothelial-dependent vasodilation is inhibited in this disease, suggesting a role of vascular endothelium in its etiology.23 Pathologically, both arteries and veins are affected by inflammatory thrombi.23

The initial stages of the disease are characterized by occlusive, highly cellular, inflammatory thrombi, with little inflammation in the endothelium.24 In the intermediate stages, there is a greater burden of thrombus in the affected area.24 However, unlike arteriosclerosis, the structure of the vessel wall is intact.24 In the final stages of the disease, blood vessels are characterized by organized thrombus and fibrosis.24

Diagnosis is confirmed by history, physical examination, vascular evaluation, and laboratory investigations.25 Disease onset before 50 years, strong history of smoking or tobacco use, infrapopliteal arterial occlusions, claudication of extremities, and exclusion of autoimmune diseases are supportive of thromboangiitis obliterans.24 Further confirmation with arteriography comparing affected and unaffected limbs is recommended.26 Unlike other forms of vasculitis, acute-phase reactants including erythrocyte sedimentation rate and C-reactive protein level are normal in this disease.26 Smoking cessation is essential to avoid limb or digit amputation.26


Chilblains lupus erythematosus manifests as the combination of cutaneous CLLs and clinical or laboratory features of lupus erythematosus (Figure 6).27 Specifically, CHLE is characterized by the appearance of an erythematous pruritic papule that often becomes painful and necrotic. Lesions are most common in an acral distribution on the dorsal surface of hands and soles of the feet.27 In certain cases, violaceous and plaque-like infiltrates, fissures with hyperkeratosis, and small ulcerations can be seen.27

Figure 6: CHILBLAINS LUPUS ERYTHEMATOSUS AND IDIOPATHIC CHILBLAINS IN THE HANDS AND LOWER LIMBSReprinted with permission from Horino T, Ichii O, Terada Y. Hydroxychloroquine-associated hyperpigmentation in chilblain lupus erythematosus. J Clin Rheumatol 2020;26(6):e192.

Epidemiologically, CHLE occurs in <1% of the population and typically affects people living in colder climates.28 Two forms of CHLE have been identified, sporadic (affecting middle-aged women) and familial (with onset at childhood).27 Familial CHLE is caused by a D18N missense mutation in the TREX1 gene.27 Lymphoblasts that carry the D18N are less sensitive to granzyme A-mediated cell death, suggesting the role of impaired apoptosis in CHLE pathogenesis.27 Conversely, there is no confirmed pathogenesis implicated in sporadic CHLE.27

The Mayo Clinic diagnostic criteria for CHLE include skin lesions in extremities induced by cold temperature, concurrent systemic lupus erythematosus, response to treatment for lupus, and negative results of cryoglobulin and cold agglutination studies.29 Further, hypergammaglobulinemia, rheumatoid factor, antinuclear and antiphospholipid antibodies, and a positive latex agglutination test are observed in a majority of patients with CHLE.29


Idiopathic chilblains are characterized by painful and pruritic erythematous lesions observed on the thighs, buttocks, and extremities, occurring in response to exposure to cold temperatures without associated connective tissue disease (Figure 6).30 A low body mass index is often associated with idiopathic chilblains.30

Epidemiologically, idiopathic chilblains is more common in women and typically affects people living in colder damp climates such as the UK or northern Europe.31 The exact pathogenesis is not proven; however, it is described as a vasculopathy that develops because of abnormal neurovascular responses to dermal temperature changes.30 Signs include erythematous or violaceous papules and skin edema, with accompanying symptoms of pruritus or burning pain sensations.31

Diagnosis is made with clinical history and examination and can be supported with histology. Differential diagnosis of this disease includes lupus chilblains; however, there is a histologic difference where lupus chilblains present with fibrin deposition that is absent with idiopathic chilblains.29 Other bloodwork including a C-reactive protein test and erythrocyte sedimentation rate may help rule out associated connective tissue disease.30 Treatments include minimizing cold exposure; vasodilatory agents including topical nitroglycerin, nifedipine, or prazosin; and antiplatelet agents including aspirin.30


The differential diagnoses of CLLs are extensive. COVID toes, Raynaud phenomenon or disease, acrocyanosis, CLI, thromboangiitis obliterans, CHLE, and idiopathic chilblains all have similar clinical manifestations. A suggested method to remember the causation and classification of dermatologic manifestations is with the mnemonic

ARCTIC: Acrocyanosis, Raynaud phenomenon or disease, COVID toes, Thromboangiitis obliterans, Idiopathic chilblains, and CLI/CHLE.

It is critical to classify CLLs to ensure proper management and limit conflation. With the advent of COVID-19, it is exceedingly important for clinicians to differentiate CLLs suggestive of alternative dermatologic pathologies from coronavirus infection.


  • Healthcare providers may encounter CLLs as a cutaneous manifestation of COVID-19 infection.
  • COVID toes, accounting for 19% to 38% of all cutaneous lesions associated with COVID-19 infection, can present as pruritic or painful swellings or purpura on the feet, most commonly the toes (as a late change).
  • Raynaud phenomenon is characterized by short periods of reduced blood flow to the fingers or toes, which manifest with segmental triphasic color change.
  • Acrocyanosis typically presents as violaceous or blue discoloration of the hands, feet, or face from constriction of the blood vessels.
  • Healthcare providers should be aware of the pathologies that can present as CLLs including COVID toes, Raynaud phenomenon or disease, acrocyanosis, CLI, thromboangiitis obliterans, CHLE, and idiopathic chilblains.

COVID-19 patients may develop skin rashes and discoloration, studies find

Authors: By Jacqueline Howard, CNN | Posted – Aug. 5, 2020 at 2:36 p.m.

CNN — As Covid-19 started to spread across the United States earlier this year, dermatology offices began to see suspicious signs on some patients’ skin: Red or purple toes, itchy hives, mottled bumps on fingers, a lacy red rash that spread across legs and arms.

But were those truly associated with the novel coronavirus? After all, many other factors could be at play.

“Many viral infections can trigger a skin rash, so when you catalog these case reports, you have to have other data. Was the patient on a medication a week before the rash began? Are there other possible causes?” asked Dr. Art Papier, an associate professor of dermatology at the University of Rochester Medical Center in New York.

“This is the challenge that Covid-19 brings up. With these different types of presentations and different rashes, is it hives because the patient just has hives or hives related to Covid-19?”

Case reports began to be released in medical journals. The latest, published Wednesday in the journal JAMA Dermatology, describes the experiences of four patients with severe Covid-19 who were admitted to hospitals in New York City in March and April.

The patients, ages 40 to 80, had discoloration of their skin as well as lesions called retiform purpura, according to the research report.

Biopsies were performed for each patient and they showed that the patients had a type of vasculopathy, meaning that their blood vessels were affected.

The researchers — from NewYork-Presbyterian/Weill Cornell Medical College — wrote in their report that the skin discoloration could represent partial occlusion or blockage of blood vessels, and the retiform purpura could represent full blockage.

Such rashes and discoloration of the skin can be a “clinical clue” to there being possible blood clotting in the body, the study said. Since early on in the pandemic, doctors have noticed that severe Covid-19 could cause abnormal blood clotting in patients.

The report comes with some limitations, including that the researchers were not able to confirm the precise timing of when rashes and other issues with the skin first appeared for each patient. Also, more research is needed to determine whether similar findings would emerge among a larger group of Covid-19 patients.

Yet overall, the researchers wrote in their report that physicians caring for Covid-19 patients should be aware of skin discoloration and rashes as “potential manifestations” of abnormal underlying blood clotting.

‘Many viral infections can affect the skin’

Doctors and researchers from around the world also have reported about other types of skin rashes among Covid-19 patients.

Covid-19 often triggers significant inflammation in its victims, in some cases producing the so-called cytokine storm that appears to be causing the worst damage in advanced patients.

The skin is particularly sensitive to inflammation, said board certified dermatologist Dr. Seemal Desai, a spokesperson for the American Academy of Dermatology.

“The cytokines that are cranking up the immune engine of the car is what then triggers a variety of these immune molecules to go into the skin and wreak havoc on the skin,” said Desai, a dermatologist in Plano, Texas.

In July, researchers from King’s College London in the United Kingdom called for skin rashes and “Covid fingers and toes” to be considered as a key symptom of Covid-19, even arguing that they can occur in the absence of any other symptoms.

Key coronavirus symptoms that are widely accepted include fever, cough and shortness of breath, but a range of other signs have been suggested. The loss of smell and taste, another outlier, was recently included on the list of most common symptoms by the US Centers for Disease Control and Prevention.

The Kings College researchers used data from the Covid-19 Symptom Study app, which is submitted by around 336,000 people in the UK. They found that 8.8% of people who tested positive for coronavirus reported a skin rash as a symptom, compared with 5.4% of people who tested negative.

The KLC team then set up a separate online survey, gathering information from nearly 12,000 people with skin rashes and suspected or confirmed Covid-19. The researchers found that 17% of respondents who tested positive for the coronavirus reported a rash as the first symptom of the disease. For 21% of people who reported a rash and had confirmed Covid-19, the rash was their only symptom.

The researchers reported their findings in a pre-print study posted to the online server medRXiv.org. The findings have not been published yet in a peer-reviewed journal.

“Many viral infections can affect the skin, so it’s not surprising that we are seeing these rashes in Covid-19,” Dr. Veronique Bataille, consultant dermatologist at St Thomas’ Hospital and King’s College London, who was involved in the pre-print study, said in a press release in July.

“However, it is important that people know that in some cases, a rash may be the first or only symptom of the disease,” Bataille said. “So if you notice a new rash, you should take it seriously by self-isolating and getting tested as soon as possible.”

Measles-like rashes and rashes inside the mouth

Preliminary research has suggested that skin rashes and lesions inside the mouth might be a symptom of coronavirus infection — but researchers say more study is needed.

In May, scientists around the world did a literature review and found patients were also presenting with red, itchy welts, and with a red or pinkish rash that looked a lot like measles.

“It’s a reaction that we typically call morbilliform which means measles, which presents in kind of pink spots, lots of little pink spots all over the skin,” said Papier, the dermatologist at the University of Rochester Medical Center.

Another study published in JAMA Dermatology in July, found that among 21 patients in Spain who were confirmed to have Covid-19 and skin rash, six of those patients or 29% had enanthem, or lesions or rash in the mouth.

The mean amount of time between the onset of Covid-19 symptoms and developing enanthem was about 12 days among the patients, according to researchers from the Hospital Universitario Ramon y Cajal in Madrid.

“This work describes preliminary observations and is limited by the small number of cases and the absence of a control group,” the researchers wrote, adding that their findings still suggest enanthem to be a possible Covid-19 symptom and not a reaction to medications, for instance.

“Despite the increasing reports of skin rashes in patients with COVID-19, establishing an etiological diagnosis is challenging,” the researchers wrote. “However, the presence of enanthem is a strong clue that suggests a viral etiology rather than a drug reaction.”

The-CNN-Wire™ & © 2020 Cable News Network, Inc., a Time Warner Company. All rights reserved.

COVID-19: Cutaneous manifestations and issues related to dermatologic care

Authors: Steven R Feldman, MD, PhD, Esther E Freeman, MD, PhD Literature review current through: Jul 2021. | This topic last updated: Apr 06, 2021.


Coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), raises many critical issues in dermatology and dermatologic care. Addressing these issues is necessary, yet also challenging, because there are few direct data on which to base recommendations [1].

This topic will discuss issues related to dermatologic care during the COVID-19 pandemic. Other relevant aspects of SARS-CoV-2 infection and patient management are discussed in detail separately.

●(See “COVID-19: Epidemiology, virology, and prevention”.)

●(See “COVID-19: Clinical features”.)

●(See “COVID-19: Diagnosis”.)

●(See “COVID-19: Infection control for persons with SARS-CoV-2 infection”.)

●(See “COVID-19: Outpatient evaluation and management of acute illness in adults”.)

●(See “COVID-19: Hypercoagulability”.)

●(See “COVID-19: Multisystem inflammatory syndrome in children (MIS-C) clinical features, evaluation, and diagnosis”.)

●(See “COVID-19: Management in hospitalized adults”.)

●(See “COVID-19: Questions and answers”.)

●(See “COVID-19: Cancer screening, diagnosis, post-treatment surveillance in uninfected patients during the pandemic, and issues related to COVID-19 vaccination in cancer patients”.)

●(See “COVID-19: Care of adult patients with systemic rheumatic disease”.)


Case series from around the world have identified a range of potential dermatologic manifestations of coronavirus disease 2019 (COVID-19) [2-5]. The frequency (ranging from 0.2 to 20.4 percent of cases) and timing of cutaneous manifestations of COVID-19 are difficult to ascertain [6-8]. Also unclear is the association of certain skin manifestations with the illness severity [9]. Moreover, it cannot be excluded that in some patients the observed skin findings may represent cutaneous reactions to the numerous treatments used for COVID-19 [9,10].

The American Academy of Dermatology’s COVID-19 Registry, a collaboration between the American Academy of Dermatology and the International League of Dermatologic Societies, is attempting to collate cases and better define the cutaneous manifestations of COVID-19 [11] (see ‘Registries’ below). Among 171 laboratory-confirmed COVID-19 patients with cutaneous manifestations from the registry, the most commonly reported were morbilliform rash (22 percent), pernio-like acral lesions (18 percent), urticaria (16 percent), macular erythema (13 percent), vesicular eruption (11 percent), papulosquamous eruption (9.9 percent), and retiform purpura (6.4 percent) [12]. Fever and cough were reported in approximately 60 percent of cases:

Exanthematous (morbilliform) rash – In several case series, a morbilliform rash predominantly involving the trunk has been reported as the most common cutaneous manifestation of COVID-19 [2,3,7,12-14]. The rash has been noted either at the disease onset or, more frequently, after hospital discharge or recovery [7].

Pernio (chilblain)-like acral lesions – Pernio (chilblain)-like lesions of acral surfaces (“COVID toes”) present as erythematous-violaceous or purpuric macules on fingers, elbows, toes, and the lateral aspect of the feet, with or without accompanying edema and pruritus (picture 1A-B). They have been described across the age spectrum in patients with confirmed or suspected COVID-19, in the absence of cold exposure or underlying conditions associated with pernio [2-4,12,15-22].

Resolution may occur in two to eight weeks. A prolonged course of more than 60 days has been reported in some patients with pernio (“long haulers”) [23].

The development of pernio-like lesions in COVID-19 may be associated with a relatively mild COVID-19 disease course [2,4,24]. In the American Academy of Dermatology/International League of Dermatologic Societies registry study, 55 percent of patients overall were otherwise asymptomatic. Ninety-eight percent of patients in the study were treated in the outpatient setting alone; this finding held true when restricted to laboratory-confirmed cases only, with 78 percent remaining in the outpatient setting [4].

Our understanding of the pathogenesis of these lesions is still under evolution, though it appears to be a primarily inflammatory process with histopathologic and direct immunofluorescence findings similar to those seen in idiopathic and autoimmune-related pernio [3,18,20,25-28]. (See “Pernio (chilblains)”.)

A French study demonstrated increased in vitro production of interferon-alpha from stimulated peripheral blood T cells in patients with pernio compared with patients with polymerase chain reaction (PCR)-positive, moderate to severe COVID-19 [29]. The histologic and biologic patterns of these patients with pernio were similar to a type I interferonopathy, suggesting that a robust, innate immune response may lead to rapid control of the virus in these patients and could, at least in part, explain the relatively mild disease course and low level of antibody production.

Pernio-like lesions may represent a postviral or delayed-onset process, with 80 out of 318 cases in the American Academy of Dermatology/International League of Dermatologic Societies registry developing lesions after the onset of other COVID-19 symptoms [4]. This finding is similar to data from Spain, where 42 out of 71 patients developed lesions after other symptoms [2].

There are several case reports and case series of patients with pernio-like lesions testing positive for immunoglobulin M (IgM), immunoglobulin G (IgG), or immunoglobulin A (IgA) for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and negative for PCR, possibly indicating a later stage in the disease process [4,24,30]. However, pernio-like lesions can, in some cases, appear while patients are still PCR positive for the virus, which has potential implications for infectivity and viral spread [4,22]:

•In a French, prospective study, 40 consecutive patients with chilblain-like lesions (median age 22 years, range 12 to 67) were tested for SARS-CoV-2 RNA with reverse transcription (RT)-PCR and SARS-CoV-2-specific IgA, IgM, and IgG antibodies with enzyme-linked immunosorbent assays (ELISAs) [24]. None of these patients were PCR positive at the time of dermatologic consultation; 12 (30 percent) had positive serology for antibodies, of whom seven had only IgA. Twenty-five patients (63 percent) were asymptomatic, and the remaining had only mild symptoms.

•In another French series of 311 patients (median age 26 years) with acral manifestations seen between March and May 2020, 150 (49 with symptoms suspicious for COVID-19) underwent nasopharyngeal swab RT-PCR and/or serology for SARS-CoV-2 [31]. Five of 75 patients were positive for SARS-CoV-2 serology, and 7 of 121 patients were positive for SARS-CoV-2 RT-PCR. Overall, 10 of 170 patients (7 percent) had confirmed COVID-19.

•In the American Academy of Dermatology/International League of Dermatologic Societies registry study, of 318 cases from eight countries, 14 of these cases were PCR positive.

•In an Italian study that screened 22 patients presenting with pernio-like lesions, 6 (26 percent) were PCR positive for SARS-CoV-2 [22].

Although the finding of PCR positivity is not universal and not all observed cases of pernio during the COVID-19 epidemic are necessarily related to COVID-19, it may be prudent that patients presenting with new-onset, pernio-like lesions that have no other clear cause be tested for SARS-CoV-2 PCR within seven days of the onset of pernio lesions [32-35]. For patients who have had these lesions for >4 weeks, IgM and IgG antibody testing may be appropriate, following local guidelines and depending on the quality of available tests, acknowledging that many of these patients may only make transient antibody responses or IgA responses that are not currently being tested in commercial laboratories. Work-up of other causes of pernio is discussed in greater detail separately [2,4,24]. (See “Pernio (chilblains)”.)

There are no treatment guidelines for COVID-19-associated, pernio-like lesions of the feet or hands. However, high-potency topical corticosteroids may be helpful if the lesions are causing discomfort.

Some patients have been noted to have “long COVID”/long-hauler COVID toes [23]. Additionally, some patients have been found to have recurrent pernio after initial SARS-CoV-2 infection, which may be triggered by cold [36].

For More Information: https://www.uptodate.com/contents/covid-19-cutaneous-manifestations-and-issues-related-to-dermatologic-care#!

Skin rash should be considered as a fourth key sign of COVID-19

May 22, 2021

Data from the COVID Symptom Study shows that characteristic skin rashes and ‘COVID fingers and toes’ should be considered as key diagnostic signs of the disease, and can occur in the absence of any other symptoms. 

The COVID Symptom Study, led by researchers from King’s College London and health science company ZOE, asks participants to log their health and any new potential symptoms of COVID-19 on a daily basis. After noticing that a number of participants were reporting unusual skin rashes, the researchers focused on data from around 336,000 regular UK app users. 

Researchers discovered that 8.8% of people reporting a positive coronavirus swab test had experienced a skin rash as part of their symptoms, compared with 5.4% of people with a negative test result. Similar results were seen in a further 8.2% of users with a rash who did not have a coronavirus test, but still reported classic COVID-19 symptoms, such as cough, fever or anosmia (loss of smell).

To investigate further, the team set up a separate online survey, gathering images and information from nearly 12,000 people with skin rashes and suspected or confirmed COVID-19. The team particularly sought images from people of colour, who are currently under-represented in dermatology resources. Thank you to all who submitted photographs of their rashes.

17% of respondents testing positive for coronavirus reported a rash as the first symptom of the disease. And for one in five people (21%) who reported a rash and were confirmed as being infected with coronavirus, the rash was their only symptom.

The rashes associated with COVID-19 fall into three categories: 

  • Hive-type rash (urticaria): Sudden appearance of raised bumps on the skin which come and go quite quickly over hours and are usually very itchy. It can involve any part of the body, and often starts with intense itching of the palms or soles, and can cause swelling of the lips and eyelids. These rashes can present quite early on in the infection, but can also last a long time afterwards.
  • ‘Prickly heat’ or chickenpox-type rash (erythemato-papular or erythemato-vesicular rash): Areas of small, itchy red bumps that can occur anywhere on the body, but particularly the elbows and knees as well as the back of the hands and feet. The rash can persist for days or weeks.
  • COVID fingers and toes (chilblains): Reddish and purplish bumps on the fingers or toes, which may be sore but not usually itchy. This type of rash is most specific to COVID-19, is more common in younger people with the disease, and tends to present later on.

For More Information: https://covid.joinzoe.com/us-post/skin-rash-covid