Massively collaborative crowdsourced research on COVID19 and the chemical senses: insights and outcomes. Elisabeth M. Weir et al. Food Quality and Preference, November 26 2021, 104483. https://doi.org/10.1016/j.foodqual.2021.104483
Highlights
• The Global Consortium of Chemosensory Research (GCCR) was founded in Spring 2020.
• Results from GCCR research on COVID19 and the chemical senses are summarized.
• Some open science approaches used by the GCCR are highlighted.
Abstract: In March 2020, the Global Consortium of Chemosensory Research (GCCR) was founded by chemosensory researchers to address then emerging reports of unusual smell and taste dysfunction arising from the SARS-CoV-2 pandemic. Over the next year, the GCCR used a highly collaborative model, along with contemporary Open Science practices, to produce multiple high impact publications on chemosensation and COVID19. This invited manuscript describes the founding of the GCCR, the tools and approaches it used, and a summary of findings to date. These findings are contextualized within a summary of some of the broader insights about chemosensation (smell, taste, and chemesthesis) and COVID19 gained over the last 18 months, including potential mechanisms of loss. Also, it includes a detailed discussion of some current Open Science approaches and practices used by the GCCR to increase transparency, rigor, and reproducibility.
Keywords: GCCRCOVID19anosmiatasteopen sciencesmell
3. Summary of some initial findings from the GCCR
While all the GCCR papers to date have resulted in key insights on COVID-19 and its effects on all three chemosesenses, the first GCCR paper (Parma et al., 2020) provided especially vital information. This analysis asked a simple question: “did self-assessment of taste, smell, and chemesthesis during illness differ from retrospective assessment of chemosensory function prior to illness?” This was determined by asking participants to rate their smell ability on a 0-100 visual analog scale (VAS) during their illness, as well as retrospectively rating their smell ability prior to their illness on the same scale – the difference between these two self-reported ratings was taken as a measure of lost function. Parallel questions were asked for taste and oral chemesthesis. Using data from ∼4000 participants collected in 10 languages, Parma et al. found that a majority of COVID19 positive participants reported a significant drop in ability to smell, with a mean drop of almost 80 points relative to their ability to smell prior to their illness. This suggested anosmia was a characteristic symptom of COVID19 (Parma et al., 2020), confirming the anecdotal reports noted above. Taste and oral chemesthesis were also significantly blunted in individuals with COVID19, with mean decreases of ∼70 and ∼37 points, respectively (Parma et al., 2020). Notably, impairment of chemesthesis was typically accompanied by either taste and/or smell loss, although in other participants, taste and smell loss appeared when normal chemesthesis was preserved. Because the survey was based on self-report of function, rather than assessment with controlled stimuli, reports of taste loss may not actually reflect taste loss (which is very rare with other viral illnesses), but may instead reflect a common taste/smell confusion among the general public (and some clinicians; see (Boltong, Keast, & Aranda, 2011)). However, the survey also asked about alteration of specific taste qualities (i.e., salty, sweet, sour, bitter, and umami/savory) with check-all-that-apply (CATA) question, and significant deficits were also observed. Changes in specific taste qualities were consistently reported by ∼37 to 45% of participants for sweet, sour, salty and bitter; reports were slightly lower for umami/savory (∼25%) but this may reflect lack of familiarity with the concept of umami and/or translation difficulty, rather than any specific robustness of this quality against loss. As noted by Green (Green, 2020), a high endorsement of altered function was reported for salty (∼45%), a quality that is not commonly attributed to odors (in contrast to sweetness or sourness). Collectively, this suggested that reported taste loss is not merely an artifact arising from a taste/smell confusion, an observation that was subsequently confirmed by multiple studies that assess taste function using various stimuli, rather than relying on self-report. Although the quality of these studies is highly variable on balance, they support the view that some of taste loss reported by those with COVID19 reflects true disruption of the taste system.
Aligning with self-reports, over a dozen studies using various stimuli (e.g., strips, homemade solutions) indicate taste loss occurs with COVID19. Although the measures of taste in these studies were often ad hoc (given the urgency of the pandemic), it is still evident that true taste loss (and not merely a semantic taste-flavor confusion) is a characteristic feature of COVID-19. That is, validated methods, like those developed as part of the NIH Toolbox (Coldwell et al., 2013) or used for the National Health and Nutrition Examination Survey (NHANES) Chemosensory Exam (Rawal, Hoffman, Honda, Huedo-Medina, & Duffy, 2015), have not been employed to date. This is wholly understandable given the needs of clinicians and researchers to collect data as quickly as possible during early days of the pandemic, often when few resources were available. Moving forward, a prime goal of chemosensory scientists should be to apply rigorous test methods to study COVID-19 taste loss, especially in those with post-acute sequelae of COVID-19 (PASC). Still, the primary message here is that the preponderance of evidence, even with imperfect measures, indicates taste loss is a verifiable symptom of COVID-19.
COVID19 associated anosmia also appears to be distinct from the transient anosmia experienced with respiratory illnesses such as the common cold. With the common cold, smell loss typically co-occurs with a lack of airflow, as a consequence of nasal congestion – that is, volatile odor active molecules cannot reach receptors on the olfactory epithelium at the top of the nasal cavity, as mucus and/or swelling block the nasal passages. In the first GCCR survey, participants were also asked to report apparent congestion, via a rating of perceived nasal blockage, where a higher score indicated more blockage. The mean pre-COVID19 nasal blockage was almost zero, whereas the mean reported blockage during COVID19 was ∼22 points higher, suggesting some participants experienced some blockage. However, comparison of the distribution of nasal obstruction ratings before and after COVID19 indicated that many participants had little to no blockage during COVID19 (Parma et al., 2020), and more critically, principal components analysis (PCA) indicated ratings of blockage were independent to ratings of smell and taste loss. This implies acute change in olfactory function for COVID19 positive individuals was not attributable to simple conductive losses. Although mechanisms for changes in taste, smell, and chemesthesis remain to be elucidated, these early data suggested disruption of chemosensory function was a hallmark indicator of COVID19 infection.
After reporting that smell and taste loss were cardinal symptoms of COVID19, the GCCR analyzed a second data tranche from the same survey that included ∼15,000 responses in 23 languages collected between 7 April 2020 and 2 July 2020. Besides fully replicating findings from the first study among non-overlapping participants (n=4,825), the new report compared responses in COVID19 positive and negative individuals, all with recent respiratory symptoms (Gerkin et al., 2021). The primary goal was to try to predict COVID19 diagnosis from all reported symptoms, regardless of whether they were chemosensory in nature.
Using both categorical and binary responses, chemosensory symptoms were found to be more strongly associated with COVID-19 diagnosis than fever, cough, or any other non-chemosensory symptoms, including difficulty breathing. In fact, self-reported smell ability during illness was the single most predictive factor for COVID-19, followed by self-reported taste ability during illness. These results highlight the importance of chemosensory changes during COVID19 infection. As anticipated, smell, taste, and chemesthesis were reported to be greatly reduced for COVID19 positive individuals compared to COVID19 negative individuals. Also, this analysis confirmed that nasal obstruction was not predictive of COVID19. Overall, non-chemosensory symptoms were less specific symptoms than smell loss, allowing those with COVID19 to be distinguished from other respiratory illnesses. In summary, smell loss was quantified as the single best predictor of COVID19, and this specificity was used to create a novel tool, the ODoR-19 (Olfactory Determination Rating Scale in COVID-19) scale, for use in rapid screening (Gerkin et al., 2021).