Antoine Darveau-Bernier, an iREx student at Université de Montréal, submitted his thesis in the fall of 2022. Here he summarises the research projects he conducted as part of his Ph.D.
Studying the atmosphere of the hottest giant exoplanets, from Earth and from space
My PhD aimed at developing data analysis techniques to study the atmosphere of exoplanets. Afterwards, I was able to apply one of these methods to observe the second hottest exoplanet known to date, WASP-33 b. To do so, we used the SPIRou instrument located at the Canada-France-Hawaii Telescope. This planet, that has a size and mass similar to Jupiter, is located so close to its star that it takes only 29 hours to orbit around it! This proximity has the consequence of forcing WASP-33 b to always show the same side to its star, a bit like the Moon with the Earth. This “day side” is constantly irradiated by its star and can thus reach extremely high temperatures (up to 3400°C!), approaching those of the coldest stars. Our analysis has allowed us to measure for the first time the abundance of carbon monoxide present in the atmosphere of this planet, in addition to measuring the temperature present at different altitudes.
To achieve this, we had to apply a novel method to distinguish the light emitted by the atmosphere of WASP-33 b, and that is buried in the signal of its star. By decomposing this light according to its different colors (wavelengths), it is possible to identify the signature of several molecules present in the atmosphere, like carbon monoxide or water. Indeed, each molecule absorbs light at specific wavelengths, which gives a distinct signature resembling a bar code. This signature is influenced by the variations of speed due to the (very fast!) movement of the planet around its star. This is called the “Doppler effect”. The star also moves, but much more slowly; it is therefore possible to use this effect to distinguish the signal of the planet from that of the star.
I also had the chance to work on the development of an algorithm called ATOCA dedicated to the analysis of data from the NIRISS instrument, the Canadian contribution to the famous James Webb Space Telescope. ATOCA will make it possible to get the most out of the observations taken by this instrument, in order to reveal the mysteries of the atmosphere of various exoplanets.
My thesis is therefore about perfecting data analysis techniques applied to the atmosphere of exoplanets, both for observations taken on Earth (with SPIRou) and from space (with JWST).
Antoine completed his Ph.D. at Université de Montréal between 2016 and 2022, under the supervision of Prof. René Doyon and David Lafrenière. His thesis will be available soon.
Serial cross-sectional estimation of vaccine-and infection-induced SARS-CoV-2 seroprevalence in British Columbia, Canada
Background: The evolving proportion of the population considered immunologically naive versus primed for more efficient immune memory response to SARS-CoV-2 has implications for risk assessment. We sought to chronicle vaccine- and infection-induced seroprevalence across the first 7 waves of the COVID-19 pandemic in British Columbia, Canada.
Methods: During 8 cross-sectional serosurveys conducted between March 2020 and August 2022, we obtained anonymized residual sera from children and adults who attended an outpatient laboratory network in the Lower Mainland (Greater Vancouver and Fraser Valley). We used at least 3 immunoassays per serosurvey to detect SARS-CoV-2 spike and nucleocapsid antibodies. We assessed any seroprevalence (vaccineor infection-induced, or both), defined by positivity on any 2 assays, and infection-induced seroprevalence, also defined by dual-assay positivity but requiring both antinucleocapsid and antispike detection. We used estimates of infection-induced seroprevalence to explore underascertainment of infections by surveillance case reports.
Results: By January 2021, we estimated that any seroprevalence remained less than 5%, increasing with vaccine rollout to 56% by May–June 2021, 83% by September–October 2021 and 95% by March 2022. Infection-induced seroprevalence remained less than 15% through September–October 2021, increasing across Omicron waves to 42% by March 2022 and 61% by July–August 2022. By August 2022, 70%–80% of children younger than 20 years and 60%–70% of adults aged 20–59 years had been infected, but fewer than half of adults aged 60 years and older had been infected. Compared with estimates of infection-induced seroprevalence, surveillance case reports underestimated infections 12-fold between September 2021 and March 2022 and 92-fold between March 2022 and August 2022.
Interpretation: By August 2022, most children and adults younger than 60 years had evidence of both SARS-CoV-2 vaccination and infection. As previous evidence suggests that a history of both exposures may induce stronger, more durable hybrid immunity than either exposure alone, older adults — who have the lowest infection rates but highest risk of severe outcomes — continue to warrant prioritized vaccination.
The British Columbia Centre for Disease Control (BCCDC) has a long-established serosurvey protocol to monitor population susceptibility to emerging or re-emerging respiratory viruses. The approach was first deployed during the influenza A (H1N1) pandemic in 2009 to monitor changes in seroprevalence across successive pandemic waves and the mass vaccination campaign.1–7 The methodology is predicated upon serial cross-sectional convenience sampling of anonymized residual sera from children and adults of all ages in the most populated Lower Mainland region of BC.8,9
Adapting this protocol, the BCCDC launched its first SARS-CoV-2 serosurvey in March 2020, just before the World Health Organization’s declaration of the COVID-19 pandemic. 10 Baseline assessment was followed by additional serosurveys that spanned the time from mRNA vaccine availability in mid-December 2020, through 7 pandemic waves associated with multiple variants of concern to August 2022 (Figure 1).11–13 Using these serosurveys, we sought to track the evolving proportion of the population that remained immunologically naive and, thus, fully susceptible to COVID-19, versus the evolving proportion that was immunologically primed (through vaccination or infection) and, thus, poised for more efficient memory response in mitigating the risk of SARS-CoV-2. Recognizing the spectrum of illness, including asymptomatic or mild infections, and variable diagnostic access, case identification and reporting, we also used estimates of infection-induced seroprevalence to explore the potential underascertainment of infections by surveillance case reports.
50 people prohibited. Provincial state of emergency declared. †Interactions limited to households or “core bubble” (immediate family or those in same dwelling) or to a maximum of 2 other people if living alone. ‡Dine-in food services and indoor fitness activities banned, only essential travel permitted. §Gradual return to gatherings, recreational travel, in-person work, which was interrupted by the fourth wave. ¶Indoor and personal gatherings limited, 50% capacity limit at venues of > 1000 people, sports tournaments paused. Social restrictions lifted during epidemiological week 7, 2022. **Mask mandates lifted during epidemiological week 10, 2022. ††The first 2 spike-based mRNA vaccine formulations were authorized during epidemiological weeks 50 and 52, 2020, respectively, with mRNA vaccines comprising most doses (> 90%) administered in BC and Canada across the pandemic. In epidemiological week 8, 2021, a chimpanzee adenoviral-vectored (ChAdOx1) vaccine was also authorized. ‡‡Vaccines (mRNA) initially deployed to high-risk individuals, including residents and staff of long-term care and assisted-living facilities, essential visitors within those settings and health care workers. §§Community-based vaccine roll-out, prioritized by age, beginning with the oldest adults in mid-March 2021. Access to booster doses followed similar prioritization sequence, inclusive of clinically extremely vulnerable individuals of any age. ¶¶Single-dose vaccine card required for entry into social and recreational settings starting in epidemiological week 37, 2021; 2-dose cards were required beginning in epidemiological week 43, 2021. Vaccine cards were ultimately repealed in epidemiological week 14, 2022.” class=”highwire-fragment fragment-images colorbox-load” rel=”gallery-fragment-images-1567424728″ data-figure-caption=”
” data-icon-position data-hide-link-title=”0″>
Study design and setting
Eight cross-sectional serosurveys were undertaken between March 2020 and July–August 2022 in the Lower Mainland (Greater Vancouver and Fraser Valley) region of BC, where about 60% of the provincial population (of about 5 million) resides.8,9 The timeline of SARS-CoV-2 serosurveys in relation to pandemic waves, publicly funded nucleic acid amplification testing, vaccine roll-out and other mitigation measures are shown in Figure 1, with additional details provided in Appendix 1, Supplementary Material 1, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.221335/tab-related-content.11–13
We obtained anonymized residual sera from children and adults visiting LifeLabs, the only outpatient laboratory network in the Lower Mainland. Two health authorities are responsible for surveillance in the Lower Mainland, namely the Fraser Health Authority (population 1.9 million) and Vancouver Coastal Health Authority (population 1.2 million).8,9 Residents of either health authority could participate, with eligible municipalities shown in Appendix 1, Supplementary Figure 1. At each serosurvey, a convenience sample of sera was selected by age group, equally by sex, from the LifeLabs central processing centre. For the first 2 serosurveys, we sampled 100 sera per age group, but thereafter, sampling increased to 200 per age group (< 5 yr, 5–9 yr, 10–19 yr, 20–29 yr, 30–39 yr, 40–49 yr, 50–59 yr, 60–69 yr, 70–79 yr, ≥ 80 yr).14 We excluded people who were specifically seeking SARS-CoV-2 antibody testing (which was limited in BC) and residents of long-term care, assisted-living or correctional facilities because of different pre-test likelihood of positivity.
At each serosurvey, we used at least 3 commercially available chemiluminescent immunoassays that targeted spike (S1) or nucleocapsid (NP) proteins.15,16 For seroprevalence estimation, we defined seropositivity as a signal above the manufacturer’s cut-off threshold on at least 2 chemiluminescent immunoassays (i.e., dual-assay positivity). Before the availability of S1-based vaccines,13 we assumed any dual-assay seropositivity to be from infection. From January 2021, infection-induced seropositivity required that at least 1 of the 2 positive assays included anti-NP detection.
We undertook serological testing in real time, with adjustment based on evolving understanding of assay characteristics and their local availability. For the first 3 serosurveys in 2020, we screened sera with Ortho (S1 total antibody) and Abbott (NP immunoglobulin [Ig] G) assays at the BCCDC Public Health Laboratory. For specimens positive on either of these assays, we also tested with the Siemens (S1 receptor-binding domain IgG/IgM) assay. With vaccine roll-out, anti-NP detection became more important, but concerns related to waning antibody levels and reduced anti-NP sensitivity also arose, particularly for the Abbott assay.17–21 For the fourth and fifth serosurveys, we supplemented testing with the Roche (NP total antibody) assay at the Providence Health Care Special Chemistry Laboratory, as volume permitted. In the event a specimen returned discordant results on the Abbott and Roche NP assays, we accepted anti-NP positivity on either assay (in conjunction with anti-S1) as evidence of infection. For the sixth and seventh serosurveys, all sera were tested by Ortho, Siemens and Roche assays. By the eighth serosurvey, BCCDC no longer offered Ortho testing, replacing it instead with the Abbott (S1 receptor-binding domain IgG) assay.15,16,22
We assessed 2 seroprevalence categories: any seroprevalence (induced by vaccine, infection or both), defined by any dual-assay positivity, and infection-induced seroprevalence, also defined by dual-assay positivity but requiring both anti-NP and anti-S1 detection. Detection of anti-NP indicated infection-induced antibody as no vaccines used in Canada contained NP antigen. Primary seroprevalence estimates with 95% credible intervals (CrIs) were based on Bayesian analysis,23–25 standardizing for age, sex and health authority. We derived cumulative and period-specific estimates, the latter conservatively reflecting the rate of new infections between specified serosurveys under the assumption of no meaningful waning of antibody levels and no reinfections. Bayesian methods are detailed in Appendix 1, Supplementary Material 2. High assay sensitivities and specificities have been reported for each chemiluminescent immunoassay, 15,16,22,26 but typically without addressing potential variation by vaccination status, time since exposure, severity or age.27–29 Like others,30,31 we did not adjust for sensitivity or specificity in the primary analyses but explored their effects as outlined in Appendix 1, Supplementary Material 2, based on assumptions detailed in Appendix 1, Supplementary Material 3.
Surveillance underascertainment ratios
All cases of SARS-CoV-2 confirmed by nucleic acid amplification testing were laboratory-reportable to local public health authorities and to BCCDC. Provincial surveillance reporting excluded reinfections and those positive only by rapid antigen test.11 We used infection-induced seroprevalence estimates and health authority–specific population census statistics to derive the estimated number of infections in the Lower Mainland. We derived surveillance underascertainment ratios with 95% CrIs by dividing estimated infections by surveillance reports from both health authorities, including cumulative and period-specific surveillance underascertainment ratios, the latter assuming no reinfections as per surveillance reporting. Additional methodological details are provided in Appendix 1, Supplementary Material 4, including exploratory investigation that included reinfections as 10% or 25% of all infections.
Sera were provided to BCCDC under legal order of the Provincial Health Officer (B.H.), and the study was approved by the University of British Columbia Clinical Research Ethics Board (H20–00653).
Of 14 000 sera collected, 13 765 (98.3%) contributed to the study. Of 235 sera excluded owing to insufficient volume, 215 (91.5%) were collected during the earliest 2 serosurveys, mostly (n = 189, 80.4%) from children younger than 10 years (Table 1 and Appendix 1, Supplementary Table 1).
Age and sex distributions reflected the Lower Mainland source population (Table 1 and Appendix 1, Supplementary Table 2). Sera disproportionately came from the Fraser Health Authority (59%–74% by serosurvey) compared with the proportion of this health authority’s population in the Lower Mainland (61%), notably among children younger than 10 years (Appendix 1, Supplementary Table 3). The Fraser Health Authority also reported disproportionately more cases of SARS-CoV-2 (about two-thirds) of Lower Mainland SARS-CoV-2 cases (Figure 1).
Overall vaccine- and infection-induced seroprevalence remained 1% or lower through the first 3 serosurveys to September 2020, and was less than 5% by the fourth serosurvey in January 2021 (Figure 2, Table 2 and Appendix 1, Supplementary Table 4). Seroprevalence rose to 56.2% by May–June 2021 (fifth serosurvey) and was higher with increasing age, consistent with age-prioritized vaccination, except among the oldest adults (≤ 70 yr) who were the earliest vaccinated by age (Appendix 1, Supplementary Material 1). By September–October 2021 (sixth serosurvey), overall seroprevalence reached 82.7%, reflecting increased vaccination of younger adults and adolescents, as well as delivery of second doses (Appendix 1, Supplementary Material 1). By March 2022 (seventh serosurvey) and July–August 2022 (eighth serosurvey), seroprevalence reached 95% or more, reflecting both higher vaccination (including third doses) and infection rates.
<a href=”https://www.cmaj.ca/content/cmaj/194/47/E1599/F2.large.jpg?width=800&height=600&carousel=1″ title=”Seroprevalence (any and infection-induced) by age group and serosurvey (serosurvey 4 in January 2021, serosurvey 5 in May–June 2021, serosurvey 6 in September–October 2021, serosurvey 7 in March 2022, serosurvey 8 in July–August 2022). Darker bars represent the infection-induced seroprevalence, which may or may not include vaccinated individuals. Lighter plus darker bars together provide a combined estimate of “any” seroprevalence (vaccine-induced, infection-induced or both). Displayed seroprevalence estimates are based on Bayesian analysis, standardized for age, sex and health authority within the Lower Mainland, British Columbia, Canada. Analysis details are in Appendix 1, Supplementary Material 2, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.221335/tab-related-content. Full results are in Table 2 and Appendix 1, Supplementary Table 4 (any seroprevalence) and Appendix 1, Supplementary Table 5 (infection-induced seroprevalence). Note: CrI = credible interval.” class=”highwire-fragment fragment-images colorbox-load” rel=”gallery-fragment-images-1567424728″ data-figure-caption=”
” data-icon-position data-hide-link-title=”0″>
Cumulative infection-induced seroprevalence remained less than 15% overall through September–October 2021 (sixth serosurvey) (Figure 2, Table 2 and Appendix 1, Supplementary Table 5). At least one-third were newly infected between the sixth and seventh serosurveys (Figure 3 and Table 3), with cumulative infection-induced seroprevalence reaching 42.5% by March 2022. Thereafter, one-fifth were newly infected between the seventh and eighth serosurveys, with 61.1% having evidence of previous infection by the July–August 2022 serosurvey.
<a href=”https://www.cmaj.ca/content/cmaj/194/47/E1599/F3.large.jpg?width=800&height=600&carousel=1″ title=”Difference in infection-induced seroprevalence by age group between the sixth and seventh (September–October 2021 to March 2022), and the seventh and eighth (March to July–August 2022) serosurveys. Displayed seroprevalence estimates are based on Bayesian analysis — standardized for age, sex and health authority within the Lower Mainland, British Columbia, Canada — and are predicated on the assumption of no reinfections and no antibody waning. In that context, estimates represent the rate of new infections between specified serosurveys, stratified by age group. Analysis details are in Appendix 1, Supplementary Material 4, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.221335/tab-related-content. Full results are in Table 3 and Appendix 1, Supplementary Table 9. Note: CrI = credible interval.” class=”highwire-fragment fragment-images colorbox-load” rel=”gallery-fragment-images-1567424728″ data-figure-caption=”
” data-icon-position data-hide-link-title=”0″>
Infection-induced seroprevalence decreased with increasing age. In general, age groups with the highest period-specific infection rates between the sixth and seventh serosurveys had the lowest rates between the seventh and eighth serosurveys. The highest rate of new infections was between the sixth and seventh serosurveys for all age categories younger than 50 years, whereas adults aged 70 years and older had their highest rates of new infections between the seventh and eighth serosurveys. Adults aged 50–59 and 60–69 years had comparable rates of new infection during both periods.
About half (45%–55%) of children aged 0–4, 5–9 and 10–19 years were newly infected between the sixth and seventh serosurveys (Figure 3 and Table 3). Rates of new infections were slightly lower (34%–44%) but with overlapping 95% CrIs among young adults aged 20–29, 30–39 and 40–49 years. Cumulatively, more than half of children were already infected by March 2022, reaching about three-quarters (70%–76%) by August 2022; rates were comparable or slightly lower (64%–70%), with overlapping 95% CrIs, among young adults (Figure 2 and Table 2). By March 2022, less than one-quarter (14%–25%) of older adults aged 60–69, 70–79 or 80 years and older had been infected. With their highest period-specific infection rates between the seventh and eighth serosurveys, still fewer than half (38%–43%) of these older adults were infected by July–August 2022.
Estimates of any seroprevalence were comparable by health authority, but infection-induced estimates were consistently higher for the Fraser Health Authority (Appendix 1, Supplementary Tables 4 and 5). Seroprevalence estimates did not differ meaningfully when stratified by sex (Appendix 1, Supplementary Tables 4–7). Crude and Bayesian-adjusted estimates were similar (Appendix 1, Supplementary Tables 4–7), and are also shown by individual assay in Appendix 1, Supplementary Table 8.
Surveillance underascertainment ratios
Surveillance case reports underestimated infections by 12-fold between the sixth and seventh and 92-fold between the seventh and eighth serosurveys, more than in previous periods (Table 3, Figure 4 and Appendix 1, Supplementary Table 9). Surveillance underascertainment ratios were highest among children aged 10–19 years and lowest among adults aged 80 years and older, with overlapping 95% CrIs between most other pediatric and adult age groups. Cumulative surveillance underascertainment ratios by serosurvey are also shown in Appendix 1, Supplementary Table 10.
<a href=”https://www.cmaj.ca/content/cmaj/194/47/E1599/F4.large.jpg?width=800&height=600&carousel=1″ title=”Period-specific surveillance underascertainment ratios (SUARs), overall and by age group between (A) the sixth and seventh (September– October 2021 to March 2022) serosurveys, and (B) the seventh and eighth (March 2022 to July–August 2022) serosurveys, Lower Mainland, British Columbia, Canada. Precise values, including period-specific surveillance case report tallies, new infection rates and SUARs, are in Table 3 and Appendix 1, Supplementary Table 9, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.221335/tab-related-content. Note: CrI = credible interval.” class=”highwire-fragment fragment-images colorbox-load” rel=”gallery-fragment-images-1567424728″ data-figure-caption=”
” data-icon-position data-hide-link-title=”0″>
Exploratory sensitivity analysis
Adjustment for assay sensitivity and specificity had little impact on estimates of seroprevalence or surveillance underascertainment ratios (Appendix 1, Supplementary Table 11 and Table 12). Assuming reinfections constituted as much as one-quarter of all period-specific infections did not affect the order of magnitude of estimates of surveillance underascertainment ratios between the sixth and seventh (16-fold), or the seventh and eighth serosurveys (123-fold) (Appendix 1, Supplementary Table 13).
Through 8 serosurveys spanning the first 2.5 years of the COVID-19 pandemic, we chronicled evolution of pediatric and adult seroprevalence in the Lower Mainland, BC. During the first year of the pandemic, when extraordinary measures were in place to curtail transmission, virtually everyone remained immunologically naive. Thereafter, age-based vaccine roll-out dramatically changed the immunoepidemiological landscape such that, by September 2021, more than 80% of the study population had antibody evidence of immunological priming, while more than 85% remained uninfected. By August 2022, after a series of Omicron waves, overall vaccine and infection-induced seroprevalence exceeded 95%, with 60% having been infected, including at least three-quarters of children but less than half of older adults.
Multiple immunological, epidemiological and modelling studies suggest that having had both vaccination and infection exposures contributes to stronger, broader and more durable hybrid immunity than with either exposure alone, especially against severe outcomes.32–47 The extent to which such exposure history should guide recommendations regarding booster doses depends on several factors, recognizing that a large proportion may not even be aware of their previous infection status.48 Moreover, the antigenic relatedness and immunological interactions between previously infecting viruses, the original monovalent vaccines, more recently updated bivalent vaccine strains, and currently circulating or emerging variants are complex and dynamic. Overall, our age-related findings to date are consistent with children being the least vaccinated and most infected subgroup, whereas older adults are the most vaccinated and least infected. Although everyone may benefit somewhat from additional vaccine doses, the relative incremental value of boosting by age depends on individual- and population-level risk assessment, notably related to severe outcomes. Over the longer horizon, the determinants and potential impact of post-COVID-19 conditions may further add to the complexity of risk assessment.49–53 Amidst this uncertainty, however, the prioritization of older adults, who are still at greatest risk of severe outcomes, remains most consistent with immunization goals to prevent serious morbidity and preserve health care capacity as the 2022–23 respiratory virus season begins.11,13
A strength of our serosurveillance approach is our sampling all age groups and both sexes simultaneously, enabling their direct comparison and extending the information available from more restricted population subsets (e.g., prenatal sera from women of childbearing age, or blood donors who are mostly younger adults). We found the highest infection rates among children, closely followed by young adults, which may reflect their greater interconnectedness, including between siblings and parents in the household, as well as with peers in schools and the community.54–56 The lowest cumulative infection rates were among older adults, which may reflect their greater vaccination rates and social isolation. Their increased rate of new infections between March and August 2022, after relaxation of public health measures and societal reopening, may reflect their lower likelihood of having previously acquired hybrid (vaccine-plus infection-induced) immunity.
In the United States, similar age-related gradation in accumulated infection rates (highest in children and lowest in older adults) has been reported.57 Pediatric seroprevalence estimates elsewhere in Canada are limited. Among children aged 17 years and younger who attended emergency departments in the Greater Montreal Area, 50%–60% had detectable anti-NP by June 1, 2022,58 similar to what we observed in March 2022, but lower than what we observed in July–August 2022. Differences may reflect provincial variation in the implementation of public health measures such as school closures or masking requirements. 59 Among Canadian adult blood donors 17 years of age and older, 54% had serological evidence of infection by the end of July 2022; estimates were highest among younger adults aged 17–24 years (71%) and lowest among adults aged 60 years and older (38%), which is also similar to our own findings.60,61
Our serosurveillance findings showed substantial underestimation of infections by standard case-based surveillance reporting, notably during the post-Omicron period. More restricted access to nucleic acid amplification testing and abundant community access to nonreportable rapid antigen testing likely contributed to underascertainment. Although other surveillance indicators may be warranted, including those for which access to testing is more consistent (e.g., among patients admitted to hospital) or sustainable (e.g., wastewater sampling), the derivation of severe outcome risks per SARS-CoV-2 case still requires accurate case tallies. In that regard, ongoing serosurveillance and associated estimates of surveillance underascertainment ratios are needed to inform the magnitude of increase in case denominators (and commensurate fold-decrease in severe outcome risks per case) required for accurate risk assessment and the optimal targeting of interventions.
By assuming no antibody waning or reinfection, our cumulative and period-specific infection-induced seroprevalences are likely underestimates and may best be summarized as “at least” that percentage infected. Among children younger than 5 years, discrepancy between our estimates of any and infection-induced seroprevalence by August 2022 may be a measure of such underestimation, given their very recent vaccine eligibility and negligible vaccine coverage. As vaccination may reduce viral loads, underestimation of infection-induced antibody may be greater among more highly vaccinated individuals.62,63 To improve upon anti-NP detection, we used both Abbott and Roche assays beginning in January 2021 (as described in Appendix 1, Supplementary Table 1), switching to the latter (with its improved sensitivity) for the final 2 serosurveys, when waning antibody levels may have been a greater concern.17–21 Convenience sampling is inherently subject to bias, but we show good concordance in the age and sex profiles of our participants with our source population, which we further standardized in Bayesian analyses. We cannot comment on discrete ethnic or socioeconomic groups who, although not specifically excluded, were also not specifically evaluated. Residual clinical specimens are more likely to come from people with underlying comorbidities who may differ in their exposure risk and immune responses, which could contribute to an underestimation of infection-induced seroprevalence, as would our exclusion of individuals who were specifically seeking SARS-CoV-2 antibody testing. In the other direction, sera collected in the follow-up of post-COVID-19 sequelae may have contributed to some overestimation. All surveillance data, as used here in estimation of surveillance underascertainment ratios, are subject to incomplete or missing information. Given our assumption of no reinfections, the higher the actual rate of reinfection, the greater the extent to which our surveillance underascertainment ratios may be conservative underestimates; however, in exploratory analyses in which we allowed reinfections to comprise as much as 25% of all infections, period-specific estimates were of similar order of magnitude. Finally, extrapolation to other geographic areas should take into account the specific context we provide here, such as in-person school attendance, mask mandates, vaccination program adjustments and other mitigation measures that may differ elsewhere.
By August 2022, most children and adults younger than 60 years in the Lower Mainland, BC, had acquired evidence of both SARS-CoV-2 vaccination and infection, which likely provides stronger, broader and more durable hybrid immunity than either exposure alone, especially against severe outcomes. With the lowest infection rates but highest risk of severe outcomes, older adults continue to warrant prioritized vaccination.
The authors acknowledge the serological testing oversight and contribution of the British Columbia Centre for Disease Control (BCCDC) Public Health Laboratory and Providence Health Care Special Chemistry Laboratory, including Tamara Pidduck, Jesse Kustra, Laura Burns, Gahyun Cheon and Bonny So. They thank Rhonda Creswell and Iva Tong of LifeLabs for diligent supervision of serum collection; Dr. Manish Sadarangani for his co-leadership of the SARS-CoV-2 Immunity Study referenced in Supplementary Material 3; and Dr. May Ahmed, Hannah Caird and Macy Zou for their supportive surveillance analyses at the BCCDC, as well as the many other frontline, regional and provincial practitioners, including clinical and public health providers, epidemiologists, medical health offficers and others for their significant contributions to surveillance case reporting and COVID-19 control measures in British Columbia.
Competing interests: Danuta Skowronski reports grants from the Canadian Institutes of Health Research and the British Columbia Centre for Disease Control Foundation for Public Health, paid to her institution, for other SARS-CoV-2 work. Romina Reyes is chair of the BC Diagnostic Accreditation Program committee. Mel Krajden reports grants paid to his institution from Roche, Hologic and Siemens. No other competing interests were declared.
This article has been peer reviewed.
Contributors: Danuta Skowronski, Samantha Kaweski and Suzana Sabaiduc contributed to the conception and design of the work. Samantha Kaweski, Shinhye Kim, Suzana Sabaiduc, Bonnie Henry and Romina Reyes contributed to data acquisition. Samantha Kaweski, Michael Irvine, Erica Chuang, Mieke Fraser, Suzana Sabaiduc, Paul Levett, Martin Petric, Mel Krajden and Inna Sekirov contributed to data analysis. Danuta Skowronski, Samantha Kaweski, Michael Irvine and Bonnie Henry contributed to data interpretation. Danuta Skowronski drafted the manuscript. All of the authors revised it critically for important intellectual content, gave final approval of the version to be published and agreed to be accountable for all aspects of the work.
Funding: Funding was provided in part by the Public Health Agency of Canada (Grant number: 2021-HQ-000067) and the Michael Smith Foundation for Health Research (Grant number 18934). The views expressed herein do not necessarily represent the views of the Public Health Agency of Canada.
Data sharing: Data sharing will be considered upon reasonable request to the corresponding author with appropriate review and aggregation, as required to comply with the provincial legislation under which the data were assembled, and respecting privacy and confidentiality requirements.
- Accepted October 27, 2022.
This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY-NC-ND 4.0) licence, which permits use, distribution and reproduction in any medium, provided that the original publication is properly cited, the use is noncommercial (i.e., research or educational use), and no modifications or adaptations are made. See: https://creativecommons.org/licenses/by-nc-nd/4.0/
Banned in the U.S., not approved for breastfeeding — why are so many moms taking this drug?
WARNING: The story contains details about suicidal thoughts and attempts.
For Jamie Robinson, the changes were subtle at first.
She found herself playing with her hair and bumping into things. But soon, she was having six or seven panic attacks a day. Things then escalated to intrusive thoughts to take her own life and punch herself in the face. Robinson, 39, knew at that point something was terribly wrong.
“I’m looking at my own baby,” the Montreal woman recalled, “this warmth that floods me had just died completely.”
“This reoccurring thought that she had been replaced, that this was not my baby, that this was maybe even a robot baby because there was no emotional engagement from my side… And the emotions are rushing to that space. The guilt, the feeling of panic. Like, am I a bad mom? Am I losing my mind?”
When her psychologist saw the distress she was in, she zeroed in on a breastfeeding medication Robinson had recently stopped taking. The medication was domperidone, a gastrointestinal drug that can also induce lactation.
But domperidone also acts as an antipsychotic, and psychologist Karen White believes Robinson was suffering withdrawal symptoms.
“[It] kind of clicked because I’ve seen people have very extreme reactions to stopping different medications,” White recalled. “And we kind of went, ‘oh, that could be it.'”
Drug prescribed off-label
Domperidone, which blocks dopamine in the brain, is approved in Canada as an aid to speed up digestion, but it also has a side effect: lactation. Doctors and midwives routinely prescribe it off-label for this purpose. More than 120 million prescriptions for domperidone were filled in 2020, according to Health Canada.
Thousands of mothers describe it in online forums as a wonder drug that helped them produce enough milk to breastfeed their babies.
“It kind of sounded like a miracle drug,” said Emily Matreal, 29, who lives just outside Detroit and took domperidone in 2021 to help her breastfeed her son, Conner.
Health Canada told CBC that although the agency is aware the drug is routinely prescribed to stimulate lactation, it is not approved for that purpose.
CBC spoke with nine women in Canada, the U.S. and Australia who say they had debilitating psychological side effects when they tried to come off the drug. They described extreme anxiety, panic attacks, insomnia and intrusive thoughts so severe they were left unable to function or care for their children, often for months. Some were forced to stop working or move in with family. At least one attempted to take her own life. They all say no one warned them these things could happen.
Multiple experts interviewed by CBC said they believe such side-effects are rare.
“It’s very unpredictable,” said researcher Janet Currie, who wrote her doctoral thesis on postpartum domperidone prescriptions in British Columbia. She says she’s helped between 15 and 20 postpartum women with severe psychological side effects slowly taper off the drug in the last year.
“No one can tell you exactly in advance whether you’ll have these symptoms and how intense they will be.”
Domperidone is not approved as a lactation aid anywhere in the world and there are no large-scale clinical trials that shed any light on how often these side effects occur.
Canadian data does not give the reason a person was prescribed a drug. But a CBC analysis of partial data from B.C., Saskatchewan, Manitoba and publicly-insured residents of Quebec found that of the nearly two million people prescribed domperidone between 2000 and 2021, more than three-quarters were women in their childbearing years.
The only published accounts of severe psychological withdrawal symptoms are case studies, including three published last month in the Journal of Breastfeeding Medicine.
Health Canada has issued several warnings about domperidone, but for cardiac side effects, not withdrawal symptoms. In 2012, 2015, and 2022, the agency noted it can cause irregular heart rates and sudden cardiac death.
Health Canada’s warnings about domperidone, which echo those from the manufacturer in the product monograph, say it should be prescribed at doses no higher than 30 milligrams per day for the shortest possible period. The European Medicines Agency has similar guidelines.
Banned in the U.S.
In the United States, domperidone is banned, for any purpose, by the Food and Drug Administration (FDA) due to cardiac risks. But the FDA ban hasn’t stopped Americans desperate to breastfeed, like Matreal outside Detroit, from seeking it out.
When Matreal’s breast milk supply dropped off at three months postpartum, she posted in a Facebook mom’s group asking for advice. That was where she learned about domperidone — and how to get it in the U.S., through a well-known Canadian doctor.
“I thought, ‘well, that sounds safe. I will reach out to the doctor and kind of see how this goes’,” Matreal said.
That doctor is Jack Newman, a pediatrician who runs the International Breastfeeding Centre in Toronto and is one of the best-known physicians in the field. Newman’s books and online reference materials on using domperidone to stimulate lactation are widely cited as evidence the medication is safe for this purpose in breastfeeding support groups with members around the world.
In an interview with CBC, Newman emphasized that if women were well supported by the health-care system to breastfeed from the beginning, domperidone wouldn’t be needed. Lactation consultants at his clinic watch mothers nurse and recommend other techniques, such as correcting a latch or breast compression, before turning to medication, he added.
He says the risks identified by regulators are overblown.
“We’ve never had a mother have a cardiac arrest. And I’m talking about thousands of mothers that we’ve treated over the years,” Newman said.
“The dose of domperidone that Health Canada recommends — it’s not a ‘you must do this’ — is useless, it’s not going to work. And so we with experience know that three tablets three times a day, and sometimes we go higher than that, actually helps, and it helps in the majority of mothers.”
Newman starts patients at 90 milligrams per day — three times Health’s Canada’s maximum daily recommendation — and sometimes goes as high as 160 milligrams.
‘Our lives kind of started to unravel’
Matreal paid $100 and was able to get a virtual appointment with a lactation consultant at Newman’s International Breastfeeding Centre in Toronto the next day.
The consultant presented her case to Newman, who prescribed domperidone at 90 milligrams per day. The clinic sent the prescription to a pharmacy in Vancouver, which shipped the medication to Matreal’s doorstep.
When her breast milk supply didn’t increase, Matreal got in touch with the lactation consultant at the clinic, who recommended increasing the dose to 120 milligrams.
At this dose, Matreal said she started producing “a good amount of milk.” Three months later, she decided to stop taking the drug.
Matreal says she was warned by the Newman clinic to wean slowly so her milk supply was not disrupted, and that there could be some anxiety. She tapered slowly at first, but then, in her eagerness to be done with pumping and freezing milk, decided to stop altogether.
Two days after going off the drug, Matreal noticed changes: dry eyes, hot flashes and sweating.
“There was just a deep feeling of panic. I didn’t have an appetite, I couldn’t eat, I couldn’t unwind, I couldn’t sleep…. And then our lives kind of started to unravel from there.”
Matreal tried to get answers from the medical community, including from Newman. In an email dated Oct. 10, 2021, viewed by CBC, she wrote to Newman, saying she was “suffering horrible, horrible anxiety” trying to come off the drug.
In an email response the next day, also viewed by CBC, Newman suggested she either take an anti-anxiety medication her doctor had recommended, or go back on domperidone and wean “very slowly, over six months, say.”
“Your situation is very unusual, by the way, since I have not heard of anyone having symptoms like you describe after only three months of taking it,” he added.
Matreal tried going back on domperidone, she said, but her symptoms persisted. She said she found some comfort when she went back to the online mom’s forums and found dozens of other women who said they experienced the same symptoms when they stopped taking the drug.
It’s a familiar story to Dr. Kaitlyn Krutsch, an assistant professor at the InfantRisk Center at the Texas Tech University Health Sciences Center School of Medicine, and author of three recently published case studies on domperidone withdrawal.
The Center, which studies the amounts of drugs that get into breastmilk, gets about half a dozen calls a week from American women in crisis trying to come off domperidone and unable to find answers from their doctors, Krutsch said. Women are reluctant to disclose they’ve been taking a banned drug, Krutsch explained. And even when they do, she said, American doctors don’t know what it is.
Many of the women, she added, get the drug from Canada.
“It’s heartbreaking,” she said. “A lot of times we’ll hear from the same moms over and over that they’re really struggling, that they’re having a lot of trouble with depression, that they’re having trouble functioning in their daily lives, that they don’t want to get out of bed. We even hear that they are suicidal.”
Krutsch said domperidone withdrawal can look a lot like postpartum depression, which can lead some health-care providers to assume that’s what’s wrong.
But there are important differences, Krutsch noted. In the case studies she published, women would experience symptoms within days of stopping the drug. When they went back on, symptoms would wane, she explained. In addition, one of the women in her case study was an adoptive mother who used domperidone to induce lactation. She was never pregnant.
In Montreal, Robinson’s psychologist Karen White also had her doubts that what she was seeing was postpartum depression. White, who has Robinson’s permission to discuss her case with CBC, said Robinson had what she considered to be a normal amount of anxiety for a first-time parent after the birth.
But when White saw Robinson several months later, Robinson’s level of distress alarmed her.
“I’ve seen very severe postpartum depression and anxiety, and this looks similar, but occurring so long after the birth and when she had been doing very well, didn’t make sense to me.”
Robinson says a lactation consultant at the Herzl-Goldfarb Breastfeeding Clinic in Montreal suggested she try domperidone. A doctor working at the clinic prescribed it.
The Herzl-Goldfarb clinic did not respond to a request for comment.
‘It would be better for everybody’
Matreal eventually stopped getting out of bed. She and her husband sold their house and moved in with Matreal’s parents so her mother could help take care of the baby. Unable to function, Matreal says she started to feel like a burden.
She tried to take her own life shortly after Conner turned one, and again around Mother’s Day.
“I 100 per cent felt like if I wasn’t here anymore, causing all of this, it would be better for everybody,” she said.
When CBC discussed Matreal’s case with Newman, he said anyone prescribed through his clinic gets comprehensive information about how to wean off domperidone slowly and is welcome to get in touch with the lactation consultants at the clinic if they run into problems.
He said such side effects are rare in his patients, and what’s more likely is the drug was masking symptoms of postpartum depression or anxiety that were already there.
“What is common in our patients, or anybody’s patients, is postpartum depression or anxiety. These mothers have been on the domperidone for several months and then small amounts of domperidone enter into the brain and act as an antidepressant. And when they go off it, especially if they go off it quickly, [they have] symptoms of maybe what is really the postpartum depression rather than the effect of the domperidone,” he said.
He said he prescribes the drug to American women about five or six times every few weeks, and that the FDA’s reasons for banning the drug are baseless.
“The FDA has said a lot of rubbish in the years …and they’re wrong about domperidone, to say it’s a particularly dangerous drug.”
Newman added that in the U.S., domperidone “isn’t actually banned because veterinarians can use it. So, you know, a million-dollar race horse is much more important than a mother.”
Asked if he’s ever advised American women to get the drug from a veterinarian, Newman replied: “Yes, but they don’t do it.”
Newman also said the clinic’s patients are warned about psychological withdrawal effects if they don’t taper off slowly, but CBC found no evidence of such warnings in reviewing the documentation Matreal received from his clinic.
The website of Newman’s International Breastfeeding Centre notes some women can experience anxiety and depression if they stop taking the drug too quickly, but says it is unlikely domperidone is responsible and that the same thing can happen to women who abruptly stop breastfeeding.
The College of Physicians and Surgeons of Ontario says doctors who provide care in other jurisdictions must do so in compliance with the college’s drug prescription guidelines — “as well as any relevant legal or professional obligations in their patient’s jurisdiction,” the college said in a statement.
Sharing the stories
A year after she stopped taking domperidone, Matreal laughs as she lunges for a toy Conner tosses in her direction as he toddles by. The family has moved into their own apartment and life is starting to feel normal again, she said.
“When I hear music, it kind of feels good again, and I’m spending time with my son.”
Matreal said she wants to share her story because hearing from other women who had been through the same thing was what helped her when she was at her lowest.
“It was displayed as an overall pretty safe drug, but then it’s so powerful that it can flip your life upside down,” she reflected.
“I guess that’s been the biggest thing, is just wanting to get my story out there and try to help people and kind of make them more aware that it might not just be a drop in breast milk and some anxiety. It could be a lot more.”
Back in Montreal, Robinson has created a website where she posts stories of other mothers who have had traumatic withdrawals from domperidone.
She said she’s doing it so other women have the information she was never given.
“I think that if women knew what the potential risk was … I don’t think almost any mother would take this risk of not being able to care for their child. It’s a nightmare.”
If you or someone you know is struggling, here’s where to get help:
CBC obtained data on the number of domperidone prescriptions from B.C., Saskatchewan and Manitoba, broken down by age range and gender, from the Canadian Institute for Health Information. Data from B.C. and Saskatchewan was from 2006-2021. Data from Manitoba was from 2015-2021. CBC designated “childbearing years” for these provinces as 15-54 (the 15-24 through 45-54 age buckets).
Data for Quebec came from the Régie de l’assurance maladie du Québec and included prescriptions for domperidone between 2000 and 2021 for publicly insured residents, which constitutes just under half the province. CBC designated “childbearing years” for Québec as 11-50 (the 11-20 through 41-50 age brackets). In total, the data included prescriptions for 1,974,475 unique individuals. The data did not include the reason the person was prescribed domperidone.
Steady Decline of HIV Prevalence in Somalia Says WHO
Flu surges on heels of RSV, COVID-19 to overwhelm children’s hospitals in Canada
The Most Famous Online Slots for Canadian Players
‘Bumbling and stumbling’: Alberta’s UCP caucus votes for changes to sovereignty bill
Silver investment demand jumped 12% in 2019
Iran anticipates renewed protests amid social media shutdown
Search for life on Mars accelerates as new bodies of water found below planet’s surface
News24 hours ago
‘Paris moment:’ COP15 conference in Montreal seeks hard targets on biodiversity
Politics22 hours ago
How Harlem Shaped Warnock’s Faith and Politics
Health10 hours ago
Serial cross-sectional estimation of vaccine-and infection-induced SARS-CoV-2 seroprevalence in British Columbia, Canada
Economy20 hours ago
Vermont Outdoor Recreation Economy
Tech21 hours ago
ASUS confirms Radeon RX 7900 XTX/XT TUF Gaming clock speeds
Sports19 hours ago
World Cup Daily: England, France set up first World Cup knockout stage clash
Health11 hours ago
Banned in the U.S., not approved for breastfeeding — why are so many moms taking this drug?
Sports21 hours ago
Fred McGriff Baseball Hall of Fame Barry Bonds Roger Clemens