Face masks modulate non-verbal communication in selfies, but do not affect beauty ratings

doi:10.21203/rs.3.rs-1996820/v1

Download PDF

Article

Face masks modulate non-verbal communication in selfies, but do not affect beauty ratings

https://doi.org/10.21203/rs.3.rs-1996820/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Widespread use of face masks during the COVID-19 pandemic has stimulated interest in their effect on interpersonal communication. We asked whether face masks modulate posing during selfie-taking and whether they affect the attractiveness of the poser. We collected a dataset of selfie pairs (one with and one without the mask) from a population of university community members. Results replicated well-known posing biases in selfies, including a preference for displaying the left cheek and a sex-related difference in facial prominence. However, these biases were modulated by face masks, which caused a marked increase of frontal poses in comparison to three-quarter poses (for both left- and right-cheek face rotations), and a weaker but still significant increase in facial prominence of both male and female takers. In contrast, beauty ratings did not differ between mask and no-mask selfies by the same individuals. These results support models of selfies as means of non-verbal communication, whereas do not support recent reports that face masks enhance facial beauty. (word count = 170)

During the COVID-19 pandemic most countries, as well as the World Health Organization, have recommended wearing face masks as a measure to counteract the spread of the SARS 2 coronavirus. These measures continue to be relevant as the disease evolves through different variants, both in clinical contexts and for the protection of elderly individuals and of individuals with special health needs (Centers for Disease Control and Prevention, 2021). It is generally believed that wearing face masks impacts interpersonal communication in various ways (Mheidly, Fare, Zalzale & Fares, 2020), although the current literature on the issue features mixed results. We studied how face masks may impact the composition of selfies, conceived as a novel mean of non-verbal communication and self-presentation within digital media (Bruno, Uccelli, Pisu et al, 2020). Our results complement those of numerous studies published within the last two years on in-person communication while wearing masks, and clarify some contradictions in the literature.

The effect of face masks on interpersonal interactions has been examined by several research groups in the recent years. There is general agreement that masks hinder the perception of several face dimensions, such as gender, age, identity, and displayed emotion, affecting both speed and accuracy of categorization (Fioutusi, Rotschild, Pnini & Azizi, 2021), lowering overall confidence (Carbon, 2020) and producing specific confusion patterns, such as misinterpreting disgust as anger and rating faces displayng other emotions (such as happines or sadness) as displaying neutral expressions (Carbon, 2020), and generally impairing the recognition of facial expressions (Grundmann, Epstude & Scheilbe, 2021; Marini, Ansani, Paglieri, Caruana & Viola, 2021; Mijazaki, Kamatani, Suda et al, 2022). However, assessements of the effect of face masks on ratings of perceived personality traits, such as trustworthiness, likeability, or closeness have yielded mixed results. Although most studies have suggested that masks have a negligible effect on ratings of such traits (Grundmann, Epstude & Scheilbe, 2021; Oldmeadow & Koch, 2021; Twele, Thierry & Mondloch, 2022), a study employing a more sensitive trust attribution paradigm reported an enhnacement of perceived trustworthiness in untrustworthy, but not in trustworthy, faces when wearing a standard face mask in comparison to unmasked faces (Marini, Ansani, Paglieri, Caruana & Viola, 2021). Finally, reports on the effect of face masks on beauty ratings have been especially contradictory, with some studies suggesting that masks lower facial attractiveness (Mijazaki & Kawahara, 2016), and other studies that they increase it (Hies & Lewis, 2022; Patel, Mazzaferro, Sarwer & Bartlett, 2020). Given these partly mixed findings, and given the potentially important impact of the issue on social interactions, it is important to further investigate the effect of masks on interpersonal communication.

In this paper, we purport to extend assessments of the effect of masks on interpersonal interactions to non verbal communication through digital media. Specifically, we have investigated mask effects on how individuals compose selfies, individual self-portraits taken with a mobile device digital camera by amateur photographers (Bruno, Uccelli, Pisu et al, 2020). Over the last decade, there has been a continuous increase in the use of selfies in online interactions, by posting on media such as Facebook, Instagram, Snapchat, or Tinder, or by sharing on an individual basis with friends or loved ones via emails, text messages, or chats. According to a recent estimate, approximately 90 million selfies will be taken daily across all devices in 2022 (Broz, 2022). With the introduction of health care measures during the epidemic, many of these now consist of individuals wearing face masks, even leading to the introduction of the #maskie hashtag on applications such as Instagram and Twitter. Despite the popularity of the face mask selfie subgenre, however, there has been essentially no research on the impact of masks on selfie taking behaviors. Evidence bearing on this issue is interesting not only on a general level, as an extension of existing data on masks and interpersonal communication in presence, but also as a test of models that seek to understand how non-verbal communication in selfies may be modulated by different communicative intentions and other contextual factors (Bruno, Pisanski, Sorokowska & Sorokowski, 2018; Bruno, Uccelli, Pisu et al, 2020; Soranzo & Bruno, 2020).

Specifically, in the current study we purported to do four things. First, we sought to evaluate whether wearing a face mask impacts posing preferences with regard to the bias favoring three-quarter over frontal presentations, and within the former with regard to the bias in favor of three-quarter poses displaying the left over the right cheek. Both biases have been amply documented in the literature (Bruno & Bertamini, 2013; Bruno, Bertamini & Protti, 2015; Bruno, Bode & Bertamini, 2017; Lindell, 2017; Manovich, Ferrari & Bruno, 2017) and are generally interpreted as due to the right-hemispheric specialization for emotional expression (Nicholls, Clode, Wood & Wood, 2019; see also Zeng, Wang, Silveira et al, 2022). Second, we aimed at evaluating how face mask may impact on pictorial facial prominence, the ratio of the area occupied by the face to that occupied by the whole body. It is well known that facial prominence is systematically higher in male portraits and self-portraits in comparison to female, and that this sex bias generalizes to selfies (Babic, Ropert & Musil, 2018), suggesting a fundamental cultural difference in the mode of self presentation across the sexes. Third, we sought to determine whether face masks modulate choices of camera angle in the selfie composition. It has been reported that males tend to favor angles from below whereas females prefer angles from above or at eye level (Sedgewick, Flath & Elias, 2017, Bruno, Uccelli, Pisu et al, 2020). Suggesting that this sex-related difference is related to different communicative intentions, it has been documented that choice in camera angle is modulated by the type of social media (the general photo sharing application Instagram vs the dating application Tinder) chosen for posting (Soranzo & Bruno, 2020). Fourth, and final, we puported to assess the impact of wearing a face mask on the perceived attractiveness of the selfie taker, an issue that has remained controversial as stated above. To these aims, we designed a study involving paired comparisons between selfies taken by the same individual, with and without a face mask. (1022 word)

Participants

We contacted members of the university community at Venezia - Ca' Foscari using internal maillists of students and staff. To each we sent a brief message stating that we were conducting a research on selfies and requesting them to to donate two such selfies, one with and one without a facemask (Appendix A). Two-hundred and thirty-four responded. Our full dataset therefore consisted of 468 selfies (two per participant). Although we did not collect demographical information for privacy reasons, the large majority of participants were university students in the age range 19- 24 years old, a desirable feature as this age group corresponds to individuals that are most likely to be familiar with selfies as a means of self presentation in non-verbal communication.

Independent variables

We assessed posing biases and poser attractiveness ratings as a function of two main predictors: taker sex (male, female) and presence of mask (masked, unmasked). In addition, visual inspection of the submitted selfies revealed that about one-third of participants failed to comply with our instruction in full, and submitted selfies that were not taken previously in different occasions but rather clearly taken one after the other, specifically for the purpose of sending them to us. To evaluate potential differences between the two subsets of our data, we classified selfie pairs from each participant as "on request" vs "spontaneous" and included this distinction as a fifth predictor.

Dependent variables

A combination of expert classifications, quantitative measures, and panel ratings was employed to derive assessments of posing biases in relation to head rotation (frontal presentation, three-quarter presentations showing more the left cheek, three-quarter showing more the right cheek), camera angle (approximately eye-level, from above, from below), body-relative facial prominence (the ratio of the area occupied by the face to that occupied by the whole body), frame-relative facial prominence (the ratio of the area occupied by the face to that of the picture frame), and poser attractiveness. Specifically, all selfies were classified as falling into one of the three head rotation and of the three camera angle categories by one of the authors (A.C.J.) and classifications were then double-checked in a random sample of the dataset by the other author (N.B.). Assessments of prominence were instead derived from measurements of screen-relative x, y positions of the top of the head, chin, top right corner of frame, and bottom left corner of frame. These measurements were collected using a simple R-script that presented the image on screen and recorded mouse clicks by the user. Poser attractiveness, finally, was assessed by 59 independent volunteers. Each was presented a random sample of thirty selfies from our database and asked to rate the attractiveness of the portrayed individual on a five level scale, where 1 corresponded to a judgment of "normal" (level of attractiveness) and 5 to "extremely beautiful" (Appendix B). The estimated attractiveness for each individual in each selfie was then estimated as the average of (two or more) independent ratings.

Table 1 presents the absolute frequencies of frontal, left, or right cheek poses, in the no-mask and mask groups. We observe the familiar prevalence of poses favoring the left cheek over the right in both groups. However, in the no-mask group, the probability of the observed frequencies (74 vs 46 for left and right, respectively) occurring by chance is p = 0.011 (test of goodness of fit to a rectangular distribution p = 0.5, chi-squared (1) = 6.5). In the mask group, the corresponding probability is p = 0.23 (chi-squared (1) = 1.5). We also observe a relative increase of frontal poses when wearing masks (135 selfies vs 98 when not wearing a mask), and a corresponding reduction of three-quarter poses (from 134 without masks to 85 with mask). Under the null hypothesis of no statistical association of pose choice and masks presence - absence, the probability of the relevant observed frequencies is p < 0.001 (test of association, chi-squared (1) = 11.24).

	frontal	left cheek	right cheek
no mask	98	74	46
mask	135	47	36

Table 1. Frequencies of posing frontally, displaying the left, or displaying the right cheek, in standard selfies (no mask) and in selfies taken while wearing a mask.

Table 2 presents the relative frequencies of three posing categories as a function of the selfie type, taken specifically to comply with the request to participate in the study, or already stored on the participant's phone (and therefore more representative of spontaneous selfie-taking behavior). The ratio of left to right cheek poses is approximately the same (3:2 ratio) in both types. However, the relative prevalence of frontal to three-quarter poses increases markedly in the on-request group (2:1 ratio) in comparison to the spontaneous group (approximately 1:1 ratio). Under the null hypothesis of no statistical association of pose choice and selfie type, the probability of the relevant observed frequencies is p < 0.0003 (test of association, chi-squared (1) = 13.4).

	frontal	left cheek	right cheek
on request	101	31	22
spontaneous	132	90	60

Table 2. Frequencies of posing frontally, displaying the left, or displaying the right cheek, in standard selfies (no mask) and in selfies taken while wearing a mask.

Table 3 presents the relative frequencies of selfies taken with the camera above, below, or at eye level, in the no-mask and mask groups and in females vs. males. In all groups, we observe similar frequencies of eye-level and from-below camera angles, both angles being chosen much more often than from-above angles.

	eye-level	below	above
females - no mask	74	82	11
males - no mask	24	24	2
females - mask	71	90	5
males - mask	25	25	2

Table 3. Frequencies of camera angle choices in standard selfies (no mask) and in selfies taken while wearing a mask.

Figure 1 presents body-relative facial prominences in the male and female selfies, in the mask and no-mask groups. Mean facial prominence was 0.7 in males vs. 0.64 in female selfies, and a 99% confidence interval on the between-participant mean difference ranged from 0.02 to 0.1, strongly supporting the conclusion that male selfies showcase higher facial prominence than female ones. Similarly, mean facial prominences were 0.66 with mask vs 0.64 without mask, and a 99% confidence interval on the within-participant mean difference in facial prominence ranged from 0.0007 to 0.047, reflecting poorer precision in estimating the parameter but still not covering zero. Our results therefore also support the conclusion that masks tended to increase facial prominence, although the size of this effect may well be less conspicuous than the effect of sex.

Figure 2 presents frame-relative facial prominences in the male and female selfies, in the mask and no-mask groups. Mean frame-relative prominence was approximately 0.53 in all groups expect in unmasked females, where this average was approximately 0.51. Standard errors for these means were of the order of 0.02 in smaller-sized male group, and of the order of 0.01 in the larger female group. In addition, the pattern of mean differences was not mirrored by a similar pattern in the group medians, as is visibile in the boxplots in Figure 2. Taken together, these features suggest that, although the size of the difference (about 0.02) is comparable in size to the mask effect on body-relative prominence, the precision of the estimates is too low to support reliable conclusions.

Finally, we computed means of beauty ratings for masked and unmasked selfies. The mean rating for masked selfies (2.65) was slightly lower than the rating for unmasked ones (2.8). In addition, the sign of this difference was the same in the groups of selfies produced upon request (2.3 vs. 2.6 for masked and unmasked, respectively) and in spontaneous selfies (2.8 vs 2.9). Thus, our dataset provided no evidence of an beauty-enhancing effect of masks.

Our results indicate that face masks modify how individuals take selfies. A surprisingly strong effect concerns posing choices with regard to head rotation relative to the camera direction. Although we clearly replicated a bias for showing the left cheek in selfies opting for a three-quarter pose, face masks strongly reduced the prevalence of such poses in comparison to frontal poses. The tendency to more often prefer a frontal pose when wearing a face mask is consistent with the idea that three-quarter poses are related to enhanced emotional expression, presumably by avoiding excessive compositional symmetry and showcasing one's "best" side (Lindell, 2017). When wearing the mask, emotional expression is instead entirely confined to the eyes, which are best displayed in a frontal pose. Interestingly, we observed a similar increase in frontal posing in selfies taken "on demand" to comply with the experimenter's request, in comparison to "spontaneous" selfies taken in different occasions. This difference mirrors differences in the frequency of frontal selfies reported by Bruno & Bertamini (2013), who analyzed in-lab selfies, and by Bruno, Bertamini & Protti (2015), who analyzed a sample of spontaneous selfies posted on Instagram. Although not the focus point of the current research, we believe this factor should not be neglected in future studies of composition in selfies as it clearly affect posing choices. In addition, we also observed a clear effect of masks on facial prominence. Although, again, we were able to replicate the sex-related difference involving higher prominence in males in comparison to females, we also observed that in both male and female selfies prominence was enhanced when wearing a mask in comparison to unmasked selfies. Importantly, we observed these effects only in body-relative prominence and not in frame-relative prominence, suggesting that the effect is related to choices in the presentation of the individuals' male and female bodies, and not merely to the distance of the smartphone when taking the selfie. We suggest that this effect is again consistent with a preference to emphasize the eyes in the selfie composition.

Our results, however, failed to corroborate two effects that have been reported previously in the literature. The first concerns the choice of camera angle. Previous studies have reported that choices of camera angle are related to the sex of the taker, with males preferring angles from below and female preferring angles from above or at eye-level (Sedgewick, Flath & Elias, 2017, Bruno, Uccelli, Pisu et al, 2020). In addition, a related study provided some evidence that camera angle is modulated by communicative intention, such that angles from above become less frequent in female selfies posted on Tinder in comparison to Instagram (Soranzo & Bruno, 2020). These effects have been interpreted in relation to the expressiveness of views from below, which may be associated with dominance, in contrast to views from above that may be associated to submission. That in our data these findings are not replicated suggests that the above effects are not so general after all, and that interpretations in terms of sex-role stereotypes may be premature. It seems instead that camera angle may be more related to trends in selfie-taking, possibly related to geographically defined subgroups. The second effect that we have not replicated concerns beauty ratings. Our database consisted of pairs of selfies of the same subjects with and without a mask. A similar paired comparison design has never been adopted in previous studies, and provides a more powerful test of the hypothesis that masks enhance beauty ratings. Our results however suggest that the same individual is perceived as equally beautiful, whether he or she is masked or unmasked. It remains to be determined why under other conditions other studies detected an effect of masks on beauty. Although several alternatives need to be considered, the possibility that the results of these studies were simply consequences of task demands needs to be taken seriously.

To conclude, we have provided evidence that facemasks modulate how individuals take selfies, at least in the current database. This conclusion is broadly consistent with the duplex model proposed by Bruno, Uccelli, Pisu et al (2020), which seeks to understand sefies as a vehicle of non-verbal communication and to analyze relevant communicative elements as "selfie-related proxemic indices", that is, as pictorial features that can be analyzed as indicators of the taker's manipulation of pictorial space for communicative purposes. To the extent that masks remove some elements of the face from view, and bring the eyes into greater prominence, one would expect modifications in such indices and that is indeed what we have observed. It may argued that a limitation of our approach is that our sample size was not comparable to that of other studies using images sampled from social media, and that the latter are more representative of ecologically valid selfie - taking. We emphasize, however, that our method was designed to sample actual selfies, that individuals had stored on their phones for whatever purpose (most likely, because they were posted or somehow shared). In that, we suggest, the analyzed images were actually quite representative. In addition, we not that our paired-comparisons design provided a much more powerful test of the hypotheses under consideration. We therefore believe that this limitation is not a serious one, although future studies may shed further light on these issues. (1700 words)

Babic, N. C., Ropert, T., and Musil, B. (2018). Revealing faces: gender and cultural differences in facial prominence of selfies. PLoS ONE 13:e0205893. doi: 10.1371/journal.pone.0205893

Broz, M. (2022), 28 Selfie Statistics Demographics & Fun Facts, Photurotial, https://photutorial.com/selfie-statistics/.

Bruno N, Bertamini M, Tyler CW (2019) Eye centring in selfies posted on Instagram. PLoS ONE 14(7): e0218663. https://doi.org/10.1371/journal.pone.0218663

Bruno N, Bertamini M (2013) Self-Portraits: Smartphones Reveal a Side Bias in Non-Artists. PLoS ONE 8(2): e55141. https://doi.org/10.1371/journal.pone.0055141

Bruno N, Bertamini M. & Protti F (2015) Selfie and the City: A World-Wide, Large, and Ecologically Valid Database Reveals a Two-Pronged Side Bias in Naïve Self-Portraits. PLoS ONE 10(4): e0124999. doi:10.1371/journal.pone.0124999

Bruno, N., Bode, C., & Bertamini, M. (2017). Composition in portraits: Selfies and wefies reveal similar biases in untrained modern youths and ancient masters. Laterality, 22(3), 279–293. https://doi.org/10.1080/1357650X.2016.1185108

Bruno, N. Pisanski, K., Sorokowska, A. & Sorokowski, P. (2018). Editorial: Understanding Selfies. Frontiers in Psychology. 9. 10.3389/fpsyg.2018.00044

Bruno N, Uccelli S, Pisu V, Belluardo M & De Stefani E. (2020) Selfies as duplex non-verbal communication: human-media interaction, human-human interaction, case study, and research manifesto. Frontiers in Computer Science. 2:12.

Carbon C. (2020). Wearing Face Masks Strongly Confuses Counterparts in Reading Emotions. Frontiers in psychology, 11, 566886. https://doi.org/10.3389/fpsyg.2020.566886

Fitousi, D., Rotschild, N., Pnini, C., & Azizi, O. (2021). Understanding the Impact of Face Masks on the Processing of Facial Identity, Emotion, Age, and Gender. Frontiers in psychology, 12, 743793. https://doi.org/10.3389/fpsyg.2021.743793

Freud, E., Stajduhar, A., Rosenbaum, R. S., Avidan, G., & Ganel, T. (2020). The COVID-19 pandemic masks the way people perceive faces. Scientific reports, 10(1), 22344. https://doi.org/10.1038/s41598-020-78986-9

Grundmann, F., Epstude, K., & Scheibe, S. (2021). Face masks reduce emotion-recognition accuracy and perceived closeness. PloS one, 16(4), e0249792. https://doi.org/10.1371/journal.pone.0249792

Hies, O. & Lewis, M.B. (2022) Beyond the beauty of occlusion: medical masks increase facial attractiveness more than other face coverings. Cogn. Research 7, 1. https://doi.org/10.1186/s41235-021-00351-9

Lindell, A. K. (2017). Consistently showing your best side? Intra-individual consistency in #selfie pose orientation. Front. Psychol. 8:246. doi: 10.3389/fpsyg.2017.00246.

Manovich, L., Ferrari, V., & Bruno, N. (2017). Selfie-takers prefer left cheeks: Converging evidence from the (extended) selfiecity database. Frontiers in Psychology, 8, Article 1460. https://doi.org/10.3389/fpsyg.2017.01460

Marini M., Ansani, A. Paglieri, F. Caruana, F. & Viola, M. (2021). The impact of facemasks on emotion recognition, trust attribution, and re-identification. Scientific Reports. 11. 10.1038/s41598-021-84806-5

Mheidly, N., Fares, M. Y., Zalzale, H., & Fares, J. (2020). Effect of Face Masks on Interpersonal Communication During the COVID-19 Pandemic. Frontiers in public health, 8, 582191. https://doi.org/10.3389/fpubh.2020.582191

Miyazaki, Y., Kamatani, M., Suda, T., Wakasugi, K., Matsunaga, K., & Kawahara, J. I. (2022). Effects of wearing a transparent face mask on perception of facial expressions. i-Perception, 13(3), 20416695221105910. https://doi.org/10.1177/20416695221105910

Nicholls, M. E., Clode, D., Wood, S. J., and Wood, A. G. (1999). Laterality of expression in portraiture: putting your best cheek forward. Proc. R. Soc. Lond. B 266, 1517–1522. doi: 10.1098/rspb.1999.0809

Oldmeadow, J. A., & Koch, C. (2021). Effects of Face Masks on Person Perception. Perception, 50(10), 876–889. https://doi.org/10.1177/03010066211045172

Patel, V., Mazzaferro, D. M., Sarwer, D. B., & Bartlett, S. P. (2020). Beauty and the Mask. Plastic and reconstructive surgery. Global open, 8(8), e3048. https://doi.org/10.1097/GOX.0000000000003048

Sedgewick, J. R., Flath, M. E., and Elias, L. J. (2017). Presenting your best self(ie): the influence of gender on vertical orientation of selfies on tinder. Front. Psychol. 8:604. doi: 10.3389/fpsyg.2017.00604

Soranzo, A., & Bruno, N. (2020). Nonverbal communication in selfies posted on Instagram: Another look at the effect of gender on vertical camera angle. PloS one, 15(9), e0238588. https://doi.org/10.1371/journal.pone.0238588

Stajduhar, A., Ganel, T., Avidan, G., Rosenbaum, R. S., & Freud, E. (2022). Face masks disrupt holistic processing and face perception in school-age children. Cognitive research: principles and implications, 7(1), 9. https://doi.org/10.1186/s41235-022-00360-2

Twele, A., M Thierry, S., & J Mondloch, C. (2022). Face masks have a limited influence on first impressions: evidence from three experiments. Perception, 51(6), 417–434. https://doi.org/10.1177/03010066221091729

Zeng, Y., Wang, X., Silveira, S., von Trott Zu Solz, J., Simmank, F., Zaytseva, Y., Bao, Y., & Paolini, M. (2022). Symmetric in the striate but asymmetric in the extrastriate cortex when processing three-quarter faces: Neural underpinnings for aesthetic appreciations. PsyCh journal, 10.1002/pchj.539. Advance online publication. https://doi.org/10.1002/pchj.539

Appendix A

Below is the English translation of the mail message that was send by the first author (A.C.J.) to internal student and staff maillists at Università di Venezia - Cà Foscari:

"We are asking you to collaborate on a research on selfies we are currently conducting. To collaborate, just send me two selfies, one without a facemask and one wearing a COVID facemask, taken in different occasions (see examples attached). Please send the selfies privately to me, using whatsapp, mail, or messenger. Add a statement authorizing us to use them for research purposes. They will be used only for that purpose, stored anonymously, and will not be divulged in any way. If you have a chance, please help. We are trying hard to collect enough images to perform meaningful statistical analyses. Many thanks in advance to all of you!!!!"

Appendix B

Below are the English translations of the instructions and the rating scale provided to the raters that collaborated to the study by reporting on the perceived attractiveness of the individuals in the selfies.

"You will be presented a small set of selfies. Some of these selfies will be of individuals wearing a COVID facemask, whereas others will be standard selfies of unmasked people. Your job is to provide us your assessment of how beautiful you consider the individual in the selfie, using a five-point rating scale as specified below:

1 - normal

2 - not-too -bad looking

3 - beautyful

4 - very beautyful

5 - extremely beautyful

For each displayed image, please respond by pressing the corresponding number key and then entering a carriage return to see the next image. Please remember that you have to rate the attractiveness of individual, not the overall quality of the photograph. Thank you for your help."

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Face masks modulate non-verbal communication in selfies, but do not affect beauty ratings

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

DISCUSSION AND CONCLUSION

References

Appendix

Additional Declarations

Status:

Version 1