Artículo de revista: Acosta, Enrique (2022) “Global Estimates of Excess Deaths from COVID-19.” Nature, 613: 31-33, December (ISSN: 1476-4687) https://doi.org/10.1038/d41586-022-04138-w. These side effects include sarcopenia — an associated with younger animals (or people). changes found in aged skeletal muscle. The age-associated decline in skeletal muscle mass The result therefore suggests that strategies to changes in gene expression observed by the and function that occurs as ageing skeletal remove senescent cells might also help younger authors have also been shown to occur in the muscle becomes replaced by fat and fibrotic people to recover from muscle injury. cells of aged tissue in general6, indicating tissue, pointing to an inability of the muscle to To explore the mechanism by which senes- that age-induced changes might be driven by repair itself4. By contrast, healthy young skele- cent cells block muscle regeneration, the group senescent cells. It was not previously known tal muscle has a remarkable regenerative capac- further profiled gene expression in the cells, that senescence might be the main driver of ity, even after injury. Moiseeva et al. set out to and found a decrease in expression of genes age-related gene changes. The work therefore investigate how the presence of senescent cells related to the function of energy-producing gives fresh rationale to the strategy of seeking might underlie skeletal muscle’s diminishing organelles called mitochondria, and an treatments that selectively remove senescent ability to regenerate with age in mice. increase in inflammatory genes, among other cells to combat age-related muscle weakness. The first challenge was to selectively isolate changes. In particular, interferon-stimulated senescent cells from muscle tissue. The enzyme genes (which are associated with inflamma- David J. Glass is at Regeneron senescence-associated β-galactosidase tion) are upregulated in both ageing and senes- Pharmaceuticals, Tarrytown, New York 10591, (SA-β-gal) is highly active in senescent cells. cence. Moreover, the authors found changes in USA. The authors collected muscle tissue that had collagen production, which have previously e-mail: david.glass@regeneron.com previously been damaged, and treated it with been linked to fibrosis6. Fibrosis interferes a fluorogenic substrate that fluoresces when with regeneration by creating the equivalent 1. Peacocke, M. & Campisi, J. J. Cell. Biochem. 45, 147–155 cleaved by SA-β-gal. This enabled them to sep- of scar tissue, instead of competent normal (1991). 2. Moiseeva, V. et al. Nature 613, 169–178 (2023). arate fluorescent senescent cells from other tissue. In the case of skeletal muscle, fibrotic 3. Wiley, C. D. & Campisi, J. Nature Metab. 3, 1290–1301 cells using a well-established approach called tissue forms instead of muscle fibres, thus (2021). fluorescence-activated cell sorting. The group impinging on muscle function. 4. Roubenoff, R. & Hughes, V. A. J. Gerontol. A. Biol. Sci. Med. Sci. 55, M716–M724 (2000). observed many more senescent cells in ageing Together, Moiseeva and colleagues’ find- 5. Theret, M., Rossi, F. M. V. & Contreras, O. Front. Physiol. 12, tissue than in young tissue after injury. They ings indicate that senescent cells trigger 673404 (2021). also used their cell-labelling strategy to iden- inflammation and block regeneration through- 6. Shavlakadze, T. et al. Cell Rep. 28, 3263–3273 (2019). tify and study the positions of the cells in the out the animal’s life, and in particular seem to The author declares no competing interests. injured tissue in vivo. be responsible for many of the detrimental This article was published online on 21 December 2022. Tissue regeneration requires both stem cells and surrounding ‘niche’ cells, which can influ- Epidemiology ence the behaviour of the stem cells. Moiseeva Global estimates of excess and colleagues examined gene expression and chromatin (the DNA–protein complex in which genetic material is packaged in the nucleus) in the senescent cell populations from young and old animals. They found that senescent cells — deaths from COVID-19 even those from young animals — make up part of the niche and have inflammation-promoting Enrique Acosta characteristics, which are associated with age-related declines in health (Fig. 1). The Estimating the number of deaths attributable to COVID-19 authors showed that several different cell around the world is a complex task — as highlighted by types give rise to senescent cells, including skeletal muscle stem cells, myeloid cells and one attempt to measure global excess mortality in fibro-adipogenic cells, the last of which can 2020 and 2021. See p.130 promote inflammation, fat deposition and fibrosis in ageing animals5. The group’s discovery that senescent-cell Knowing how COVID-19 affects global mortality as the difference between all observed and numbers increased drastically after damage rates is crucial if we are to understand the fac- expected deaths in a given period — is consid- highlights a mechanism that might explain tors that govern its spread and severity, and ered the gold-standard approach for estimat- why some older people are more affected by to be able to evaluate the effectiveness of gov- ing the mortality toll of short-term events2,3. ageing processes than others — if they have ernment responses to the pandemic. In May, But it is hard to find a universally effective had an injury, more senescent cells might be a team of researchers led by the World Health way to measure excess mortality4,5, because present in their muscles. In line with this idea, Organization (WHO) and the United Nations there are substantial variations in underlying geriatric mouse muscle that had been injured Department of Economic and Social Affairs mortality trends and data availability across and so harboured senescent cells was less able published the first results from their attempt populations. to induce force than was uninjured geriatric to estimate global, COVID-19-related death Msemburi and colleagues set out to esti- tissue, even after the injury had repaired. How- rates. On page 130, Msemburi et al.1 present mate excess deaths from COVID-19 for every ever, the strength of the muscle was improved these estimates in more detail. country in the world. The authors report by giving the animals dasatinib and quercetin, Many deaths from COVID-19 went that there were between 13.2 million and drugs that can kill senescent cells. This shows undetected in official reports from 2020 and 16.6 million more deaths than expected in that removal of senescent cells can improve 2021, because of limited testing capacity and 2020 and 2021. This death toll was between muscle function. misclassification of causes of death. This lack 2.4 and 3.1 times higher than the officially Moiseeva et al. found that removal of senes- of data makes it challenging to quantify the reported number of COVID-19-related deaths. cent cells also improved muscle repair in mortality toll of short-term events, such as Four out of five excess deaths occurred in younger animals. This finding was somewhat wars and natural disasters, as well as pandem- middle-income countries (Fig. 1), with some surprising, because senescence is not usually ics. For this reason, excess mortality — defined of the worst affected in Latin America. In both Nature | Vol 613 | 5 January 2023 | 31 © 2 0 2 3 S p r i n g e r N a t u r e L i m i t e d . A l l r i g h t s r e s e r v e d . News & views overall deficit remain uncounted.) Data source Although the inferences made by Msemburi Complete monthly Partial monthly Annual Subnational No data and colleagues are not ideal, there is no obvi- ous alternative. However speculative these High estimates, most are surely closer to the Upper middle truth than are officially reported numbers of Lower middle deaths from COVID-19. To rely on confirmed deaths would imply that the pandemic spared Low low-income and lower-middle-income coun- 0 1 2 3 4 5 6 7 8 tries — vulnerable populations that have lim- Estimated excess deaths (millions) ited capacity for testing and response. This assumption is highly implausible, and even Figure 1 | Country income level and estimated excess deaths from COVID-19. Msemburi et al.1 estimated irresponsible. the number of excess deaths that occurred in every country in the world in 2020 and 2021. Here, countries are Nonetheless, the complexity of the task is grouped according to their income level (high, upper middle and so on). The authors made their estimates apparent from the fact that similar attempts to using monthly, country-level data on the total number of deaths where available. However, they had to make estimate the effect of the pandemic on global estimates of excess deaths from incomplete data for most countries. For some, data were available for only mortality have given different results. For the certain months or as annual data. More problematically, in many countries, data were available on only a same period, the University of Washington’s subnational level, or not at all. Globally, half of the estimated excess deaths came from these countries, with Institute for Health Metrics and Evaluation in this proportion varying hugely by the country’s income level. The authors’ estimates involved subtracting negative excess (where there were fewer than expected deaths) from positive excess, meaning that for some Seattle12 estimated 18.2 million excess deaths, subgroups (such as annual upper middle and lower middle), the overall number of deaths is negative. (Figure and The Economist magazine estimated 16 mil- generated using data from Supplementary Tables 5–13 of ref. 1.) lion (see go.nature.com/3uykedp). This makes the WHO estimate the most conservative of the three. Compared with these other studies, the years combined, the observed mortality in (mostly high-income countries that have approach adopted by Msemburi et al. is sim- Peru was double the expected level, and it was robust health-care systems). The researchers pler, and their estimation of uncertainties is between 41% and 51% higher than expected adjusted the patterns seen across the latter more rigorous. in Mexico, Bolivia and Ecuador. Low-income group to make estimates for the former using A sensible next step for Msemburi and col- countries had fewer deaths, mostly because country-specific proxies for socio-economic leagues — and one currently under way (see they account for only 9% of the global conditions, the intensity of the pandemic, each go.nature.com/3vhmybu) — is to include infor- population, and have younger populations, country’s susceptibility to COVID-19 and its mation on age in their data. The risk of COVID- on average, than do higher-income countries. capacity to respond to the crisis. As such, half 19 death increases with age13. Because the Compared with the published data, Msemburi of global excess deaths were estimated with- authors calculated excess mortality for entire et al. present more-detailed methods and out data on mortality, or by using data from populations, any difference between countries more-refined estimates, by adjusting the way subnational regions (Fig. 1). is affected by variations in their age composi- in which underlying mortality trends were There are other factors to consider when tions. There have been some attempts to rank forecast for several countries6. interpreting Msemburi and colleagues’ find- countries’ response to the pandemic (including However, the authors’ estimates must be ings, including ‘avoided’ and ‘displaced’ mortal- one in the current study) by estimated overall interpreted with extreme caution. This is ity. The former refers to expected deaths that death toll, but without data on age, any assess- because only 37% of countries had complete did not occur — for instance, influenza-related ments of differences in pandemic severity or data for the number of people who died from deaths that were avoided in 2020 and 2021, the effectiveness of responses will be biased. any cause in each month of 2020 and 2021 — owing to changes in people’s behaviour and Finally, the complexity of estimating the an essential figure for accurately calculating social-isolation measures9. The latter refers to effect of the pandemic on global mortality excess mortality. There were no data at all for deaths of frail or sick people that were expected underscores the urgent need to build robust, 43% of countries; 2% of countries had data for to occur during the observation period, but centralized systems that allow for real-time only some regions; 5% had only annual data; were brought forward by COVID-19, producing monitoring of global mortality. The construc- and 13% had incomplete monthly series. The a temporary surplus in mortality followed by a tion of such systems will require considerable availability of data is heavily correlated with deficit, and consequently ignored when calcu- global efforts to strengthen civil registration income. For example, only 2% of European lating cumulative excess deaths. and crucial statistics systems worldwide, espe- countries had no data for 2020, whereas for Msemburi et al. did not adjust for these cially in low- and middle-income countries. But, African countries this figure was 87%. factors, because they aimed to identify all once built, they will serve as an essential early As a result, the authors had to make some changes in mortality during the pandemic. warning for future pandemics and health crises. problematic inferences. First, for countries Nevertheless, adjusting estimates for these that lacked data in some regions, they had numbers is essential if the goal is to under- Enrique Acosta is at the Centre for to scale death counts from subnational to stand COVID-19 fatality better. There are Demographic Studies, 08193 Barcelona, national levels. This assumes a constant share already models available that can adjust for Spain, and at the Max Planck Institute for of deaths between regions before and during influenza-related avoided mortality10, and Demographic Research, Rostock, Germany. the pandemic. But the spread, timing and new approaches have been proposed for use e-mail: acosta@demogr.mpg.de severity of COVID-19 within countries was far in the context of the COVID-19 pandemic11. For from uniform7,8. displaced mortality, partial adjustments can 1. Msemburi, W. et al. Nature 613, 130–137 (2023). 2. Leon, D. A. et al. Lancet 395, e81 (2020). Second, the researchers had to infer the be made by excluding mortality deficits when 3. Helleringer, S. & Queiroz, B. L. Int. J. Epidemiol. 51, 85–87 number of expected and excess deaths in computing the cumulative number of excess (2022). countries with no mortality data (mostly deaths over the months of the observation 4. Nepomuceno, M. R., Klimkin, I., Jdanov, D. A., Alustiza-Galarza, A. & Shkolnikov, V. M. Popul. Dev. Rev. low-income countries) by extrapolating pat- period. (The adjustment is only partial because 48, 279–302 (2022). terns from those with more-complete data COVID-19 deaths that occur during periods of 5. Schöley, J. Preprint at medRxiv 32 | Nature | Vol 613 | 5 January 2023 © 2 0 2 3 S p r i n g e r N a t u r e L i m i t e d . A l l r i g h t s r e s e r v e d . https://doi.org/10.1101/2021.06.04.21258353 (2021). 11. Shkolnikov, V. M. et al. SSM Popul. Health 18, 101118 6. Nature 606, 242–244 (2022). (2022). 7. Lima, E. E. C. et al. Genus 77, 30 (2021). 8. Konstantinoudis, G. et al. Nature Commun. 13, 482 (2022). 12. Wang, H. et al. Lancet 399, 1513–1536 (2022). 13. Sasson, I. Demogr. Res. 44, 379–396 (2021). From the archive 9. Simonsen, L. & Viboud, C. eLife 10, e71974 (2021). 10. Thompson, W. W. et al. Influenza Other Respir. Viruses 3, The author declares no competing interests. How kangaroo rats avoid eating too much 37–49 (2009). This article was published online on 14 December 2022. salt, and searching for trout Materials science 50 years ago Electric switch found The harsh environment of the desert poses many problems, such as [that] of … obtaining enough food and water, to for a superconductor animals like the kangaroo rats … in the deserts of western North America. The possibility of obtaining food and water from such plants as can survive there is Kenji Yasuda confounded by the fact that one of the Ferroelectricity has been found in a superconducting adaptations shown by desert plants such as the saltbush (Atriplex) is to have a layer compound. Strong coupling between these two properties of salt-filled tissue at the surface of their enables ferroelectric control of the superconductivity, which leaves. This seems to help the salt–water could prove useful for quantum devices. See p.48 balance of the plants but would make them dangerous eating for most animals because of the high concentration of salt. When some materials are cooled, their elec- of this polarization can be switched by an In one species of kangaroo rat, Dipodomys trons flow without friction. This is a property external electric field are known as ferroelec- microps, this particular difficulty has been known as superconductivity, which makes trics. Superconductivity and ferro­electricity partially solved … G. J. Kenagy reports ... them lose their electrical resistance at ultralow are thought to be mutually exclusive, because that this species has broad, chisel-shaped temperatures. Superconductivity is particu- the electrons in a superconductor pair up to lower incisors … [that] enable it to strip larly promising for the development of quan- conduct electricity extremely well, and this off the salty outer layers of the leaves tum devices, and the ability to switch between high conductivity should cancel the electrical and obtain the inner portions … which a superconductor and a normal metal adds an polarization2. But Jindal et al. show that these are relatively salt-free. D. microps holds extra functionality to such devices —informa- properties can coexist in bilayer MoTe2 (Fig. 1). a saltbush leaf in its forefeet and draws tion storage. On page 48, Jindal et al.1 report The authors do so by combining two pre- it over its lower teeth about ten times, so that the superconductivity of ultrathin sheets vious discoveries. In 2018, it was shown that shaving off the outer layers of one side. It of the semimetallic compound molybdenum the electrical polarization of bilayer and then turns the leaf over and does the same ditelluride (MoTe2) coexists with another trilayer tungsten ditelluride (WTe2) could thing to the other side. property, called ferroelectricity, that can be be switched with an external electric field3. From Nature 5 January 1973 used to switch the material’s superconductiv- This ferro­electric metal belongs to a broad ity on and off. group called van der Waals materials, which 150 years ago In some crystals, the arrangement of atoms have layered crystal structures that can be Some months ago I inquired through separates positive and negative charges in a thinned down to atomic thickness through the columns of the Field newspaper if any way that generates a macroscopic electrical mechanical exfoliation. It is this layered struc- sportsman, fisherman, or naturalist would polarization. Materials in which the direction ture that enables the coexistence of metallic oblige me by replying to the following queries respecting the rarer Salmonidæ of Great Britain. Firstly, whether Salmo ferox (the great lake trout) had ever been taken in any lake in Wales, and, secondly, whether Mo any of the Gwiniad tribe (Coregoni) … or Te any of the Chars (Salmo salvelinus) have ever been taken in any lake which is not Polarization a glacial lake … ? The only reply with respect to the first query … establish[ed] Electron the fact that [Salmo ferox] still lingers in North Wales. To the second question I Electric field have received no reply. Is it possible that switched Electron MoTe2 I may be more fortunate among the many pair naturalists and geologists who take Nature? From Nature 2 January 1873 Figure 1 | Switching superconductivity on and off. Jindal et al.1 show that two layers of molybdenum ditelluride (MoTe2) exhibit coexisting properties: superconductivity (in which electrical resistance is zero and electrons move without friction in pairs) and ferroelectricity (in which an external electric field can switch a material’s electrical polarization). In the authors’ MoTe2 device, an external field can be changed to switch the electrical polarization, turning the superconductor into a normal metal (in which electrons encounter friction) and back again. Nature | Vol 613 | 5 January 2023 | 33 © 2 0 2 3 S p r i n g e r N a t u r e L i m i t e d . A l l r i g h t s r e s e r v e d .