COVID-19 and older people in Mexico: the perils of premature data analysis

Apr 4, 2021 | All posts, Opinions and contributions, Relevant news and stories

By Peter Lloyd-Sherlock

4 April 2021.

Several Global Platform blogs and related publications draw attention to the limited quality and availability of age-disaggregated data on COVID-19 mortality in low and middle-income countries (LMICs) [link]. Sometimes researchers have to do the best they can with questionable data, perhaps adding a caveat or two as an insurance policy. But sometimes the data are so flawed that researchers should disengage.

Mexico is a case in point. At the end of March the Ministry of Health “revised” its estimate of all-age COVID-19 mortality from 201,832 to 294,287 [link]. Yet, unlike most other LMICs, Mexico had appeared to be an exceptional example of good practice in publishing detailed data on COVID-19. For most the pandemic, an official government website [link] appeared to offer up-to-date information on many issues, including age–disaggregated data on mortality, cases and hospitalisation. Interesting, this user-friendly site does not appear to have caught up with the official revision, claiming all age mortality to be only 204,011 as of 3 April 2021.

The same official website shows that people aged 60 or more accounted for 63 per cent of reported COVID-19 deaths. This is a lower share than reported for other Latin American countries, such as Brazil (77 per cent) or Argentina (84 per cent). The apparently lower concentration of deaths at older ages has promoted some analysis and discussion [link]. One of the more plausible explanations is that very high rates of diabetes and obesity among people at younger ages increased their case fatality rates [link]. There may be some validity in this explanation and it may indeed be the case that older people really do account for a lower share of COVID-19 deaths in Mexico than in other countries. But, given the lack of reliable data, we really can’t say: the revised estimates do not include an age-breakdown.

My guess is that older people have accounted for much more than 63 per cent of Mexico’s COVID-19 deaths to date. This is based on two things: (i) comparisons with similar countries and (ii) separate studies that reveal COVID-19 deaths among older people are less likely to be correctly diagnosed than they are for people at younger ages, and therefore go uncounted [1].

The consequences of this policy failure are serious for Mexicans of all ages. Older people, in particular, have been misled into under-estimating the risk of the pandemic. There is also a lesson for researchers. For months, Mexican health workers have been questioning the accuracy of official statistics [link], but their concerns were largely ignored. Mexico’s COVID-19 data website may looks professional and is easy to use, but the data it presents are highly misleading. Best left alone.


1. Miki, J., Rampatige, R., Richards, N., Adair, T., Cortez-Escalante, J., & Vargas-Herrera, J. (2018). Saving lives through certifying deaths: Assessing the impact of two interventions to improve cause of death data in Perú. BMC Public Health, 18(1), 1329. doi:10.1186/s12889-018-6264-1