For most of the last century, high-income countries experienced annual declines in age-specific mortality rates.¹ However, the COVID-19 pandemic created upturns in mortality rates between 2020 and 2021 in many high-income countries.^2-4 In the wake of the pandemic, some mortality rates have resumed their pre–COVID-19 pattern of annual declines. However, a portion of the mortality decline observed immediately after a pandemic can be explained by mortality displacement (also referred to as the mortality harvesting effect, harvesting, and the harvesting effect). This occurs when the pandemic causes frail or high-risk individuals, who would have died in the near future, to die earlier than they otherwise would have. As a result, the months or years following the crisis may show fewer deaths than expected because many of the most at-risk individuals had already died prematurely.⁵ This demographic phenomenon has been documented after previous health crises, including epidemics, pandemics, severe influenza seasons, and heat waves, and could therefore be plausibly observed following COVID-19 as well.^6-8

In parallel, other factors may drive postpandemic mortality decline. Rapid scale-up of vaccination campaigns reduced COVID-19–related mortality and helped prevent future waves.^9,10 On average, easing COVID-19 mobility and cross-border restrictions coincided with a rebound in trade and travel, contributing to a partial economic recovery and some improvements in employment and poverty, which are key social determinants of health, although these gains were highly unequal between and within countries.^11-17 Together, these multiple processes shape observed mortality trends after a pandemic and underscore the need to distinguish the temporary mortality displacement from longer-term recovery. Disentangling the contribution of mortality displacement from other population health drivers is crucial to accurately assess the pandemic’s true burden and to guide public health strategies during system recovery.¹⁸

There are various methods to calculate excess mortality.^19-23 Overall, these studies affirm the general approach to estimate excess deaths relative to a projected baseline of expected deaths before the pandemic. While commonly used during periods of elevated mortality, the method also applies when observed deaths fall below baseline, often termed negative excess or deficit mortality. In both cases, estimates are benchmarked against the counterfactual of uninterrupted prepandemic trends. However, a critical gap involves quantifying the extent of the mortality displacement on mortality across different age groups immediately following the COVID-19 pandemic. Prior studies focusing on excess mortality have a limited scope, concentrating on the pandemic period and not covering the postpandemic recovery phase.

Bor et al²² provided insights into postpandemic mortality declines in the US, and Riou et al²⁴ provided findings of those in Switzerland; however, a multicountry analysis of the mortality displacement across age groups, to our knowledge, is largely absent. Thus, this study systematically measured the mortality displacement using high-quality data available by country, year, age, and sex following the COVID-19 pandemic.

In this cross-sectional study, we used the Short-Term Mortality Fluctuations dataset, which compiles harmonized weekly all-cause mortality statistics from national vital registration systems of multiple countries. We extracted weekly mortality data covering the period from January 2015 through December 2024 for 34 selected countries. Most included countries were classified as high-income economies by the World Bank; Bulgaria (newly reclassified to high-income in fiscal year 2025) was included because of its reliable and harmonized mortality surveillance system.²⁵ This study represents a secondary analysis of publicly available aggregate data from the Human Mortality Database, without patient or public involvement; as such, this study did not require institutional review board approval or informed patient consent, in accordance with 45 CFR §46. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Mortality data were stratified by country, year, week, sex, and age groups (0-14 years, 15-64 years, 65-74 years, 75-84 years, and ≥85 years). The Short-Term Mortality Fluctuations dataset provided both weekly death counts and weekly death rates, with death rates calculated as the number of deaths occurring each week divided by the annual population exposure, normalized to weekly units. Data for each year typically included 52 weeks; week 53 was excluded to ensure comparability across years. Countries included in the analysis were categorized into 5 quintiles based on the severity of their cumulative excess mortality from January 2020 to December 2022. Mortality displacement was hypothesized to have been highest in countries with the highest January 2020 to December 2022 excess mortality and negligible in countries with the lowest January 2020 to December 2022 mortality. Quintiles were defined from the lowest excess mortality (quintile 1) to the highest excess mortality (quintile 5), providing a structured comparison of pandemic outcomes across countries.

We defined 2020 to 2022 (or 2020 to 2023) as the pandemic period, and 2023 to 2024 (or only 2024) as the postpandemic period; this definition was applied consistently in all analyses. To illustrate the analytic concept, a stylized mortality index for 2015 to 2025 under 2 counterfactual scenarios, 1 with mortality displacement (Figure, A) and 1 without (Figure, B), is shown. In both panels, the prepandemic trend from 2015 to 2019 is extrapolated and represents the number of deaths expected had the pandemic not occurred. A temporary spike in deaths from 2020 to 2022 is shown in the Figure. In the scenario without mortality displacement, deaths returned to the projected trend from 2023 onward; in the scenario with mortality displacement, deaths fell below the projected trend in 2023 and 2024, creating a visible gap in the prepandemic trend.

In our empirical analysis, we treated this gap as an association with mortality displacement and quantified it in 3 steps. First, for each country and year, we project the expected number of deaths in 2020 to 2024 by extending the 2015-to-2019 trend. Second, we summed the excess deaths from 2020 to 2022 (or from 2020 to 2023), defined as deaths above this expected level. Third, we summed the mortality deficit during 2023 to 2024 (or only 2024), defined as deaths falling below the expected level. The mortality displacement fraction was then interpreted as the share of the initial excess deaths in 2020 to 2022 (or 2020 to 2023) that was paid back by the mortality deficit in 2023 to 2024 or only 2024 (ie, the ratio of the postpandemic deficit to the pandemic-period excess). The methodologic details for each step are described.

The expected weekly deaths were estimated with an overdispersed Poisson generalized linear model using historical mortality data from the prepandemic period (2015-2019). For each country (c), sex (s), and age group (a), we fitted prepandemic data (week 1, 2015, to week 52, 2019) as follows in Equation 1^19,21,26:

(opens in new tab)

in which μ_c,s,a(t) is the expected number of deaths at week t; E_c,s,a(t) is the weekly person-weeks exposure used as an offset; and β_5,c,s,a(year) adjusted for temporal trend. Two harmonic pairs captured the dominant annual and semiannual seasonal components. Robust SEs were used to accommodate overdispersion. Based on each country’s parameters for age and sex, we projected death counts for week 1, 2020, to week 52, 2024. Afterward, we compared the observed weekly number of deaths in 2020 to 2024 with the expected number of weekly deaths for each country, stratified by age and sex, to estimate weekly excess mortality. Specifically, weekly excess deaths for 2020 to 2024 were calculated in Equation 2: excess_c,s,a(t) = d_c,s,a(t) − μ̂_c,s,a(t), in which μ̂_c,s,a(t) is the fitted value from Equation 1. Annual excess mortality rates per 100 000 R_c,s,a,y were obtained by Equation 3:

(opens in new tab)

Given the linear trend in expected mortality derived from 2015 to 2019, having an excess mortality rate of 0 in 2023 or 2024 would imply exactly resuming the established pre–COVID-19 mortality decline. If a country did better than achieving 0 excess mortality in 2023 or 2024, then it signals the potential presence of mortality displacement. It is also possible that a country superseded its pre–COVID-19 population health performance by improving the overall health system. However, if the country was in a high quintile for COVID-19–era excess deaths, then mortality displacement became a more plausible explanation. To isolate mortality displacement from chance, we defined mortality displacement as a sustained cumulative postpandemic mortality deficit in 2023 and 2024 or in 2024, conditional on a positive excess in 2020 to 2022 (or 2020 to 2023). Deficits were assessed against the bootstrapped 95% prediction interval of the prepandemic baseline. For each qualifying stratum, we then computed the cumulative excess mortality during the initial surge (2020-2022 or 2020-2023) and the subsequent deficit (2023-2024 or 2024). The mortality displacement fraction²⁰ was finally expressed in Equation 4:

(opens in new tab)

or Equation 5:

(opens in new tab)

To obtain 95% CIs for both R_c,s,a,y, we performed nonparametric bootstrapping with 1000 resamples. For each bootstrap replicate, we resampled the weekly mortality data with replacement within each country-sex-age stratum. For each resample, we reestimated the expected deaths by refitting Equation 1, recomputing excess deaths and the excess death rate in Equations 2 and 3, and recalculating the mortality displacement ratio in Equations 4 and 5. The 2.5 percentile of the resulting bootstrap distributions was taken as the lower 95% CI bound and the 97.5 percentile as the upper 95% CI bound.

We conducted 3 prespecified sensitivity analyses to evaluate robustness: (1) refitting the baseline model using a negative binomial specification, (2) fitting a Poisson generalized estimating equation with an autoregressive (AR[1]) working correlation structure to account for serial correlation in weekly deaths, and (3) repeating all models using broader age bands (0-64 years, 65-74 years, 75-84 years, and ≥85 years) to reduce sparsity in age-specific strata. To address the risk of an inflated type I error from multiple hypothesis testing, we used the Benjamini-Hochberg adjustment to control the false discovery rate (q ≤ 0.05). We compared the main results with the unadjusted P values to assess the outcome of the adjustment. A 2-sided P < .05 was considered statistically significant. All calculations were performed using Stata, version 18.5 (StataCorp LLC).

Our analysis included 352 182 284 deaths in 34 countries from 2015 to 2024; the median (IQR) population composition in 2020 was 50.75% (50.33%-51.44%) females, 49.25% (48.56%-49.67%) males, and 19.64% (17.74%-20.64%) aged 65 years or older. Based on Equation 1, we found that in the prepandemic period (2015-2019), 13 countries (Australia, Iceland, Israel, Luxembourg, Norway, New Zealand, Canada, Switzerland, Sweden, Belgium, Spain, England and Wales, and Austria) exhibited significant negative mortality trends (eg, Australia: rate, −0.0212 [95% CI, −0.0234 to −0.0190]; P < .001), 8 (Portugal, Slovenia, Czechia, Greece, Bulgaria, Croatia, Latvia, and Poland) showed significant positive trends (eg, Portugal: rate, 0.0096 [95% CI, 0.0033 to 0.0160]; P = .003), and the remaining 13 were fairly similar over time (eTable 3 in Supplement 1). We defined baselines as well-estimated if they were significant.

The cumulative excess death rate per 100 000 (based on mean population) for 2020 to 2022 compared with expected deaths based on 2015-to-2019 mortality is provided in Table 1 and for 2020 to 2024 in Table 2. Positive values indicate more deaths than expected; negative values indicate fewer deaths than expected. The 34 countries were grouped into 5 quintiles according to their cumulative excess death rate in 2020 to 2022 (CEDR20-22). In the lowest quintile, the CEDR20-22 values were −21 (95% CI, −35 to −6) per 100 000 for New Zealand, 3 (95% CI, −37 to 42) per 100 000 for Luxembourg, 26 (95% CI, 6 to 46) per 100 000 for Denmark, 81 (95% CI, 68 to 94) per 100 000 for Australia, 82 (95% CI, 62 to 103) per 100 000 for Norway, 108 (95% CI, 93 to 124) for per 100 000 Israel, and 113 (95% CI, 62 to 165) per 100 000 for Iceland, with all values except Luxembourg achieving statistical significance. New Zealand’s value of −21 indicates that it was able to sustain better-than-expected mortality declines even during the 2020-to-2022 period. New Zealand’s reduction persisted through 2020 to 2024 (−35), and Denmark (−21) and Luxembourg (−100) also recorded reductions. New Zealand pulled off statistically significantly negative cumulative excess mortality from 2020 to 2024 in all age groups except ages 75 to 84 years. Notably, Denmark and Luxembourg, with the lowest COVID-19–era mortality, exhibited significant cumulative negative excess mortality among individuals aged 85 years or older from 2020 to 2024. The remaining quintile 1 countries, namely Denmark, Australia, Norway, Israel, and Iceland, failed to sustain negative values and showed statistically significant excess mortality from 2020 to 2024. Conversely, CEDR20-22 values for Poland were 520 (95% CI, 457 to 584) per 100 000; Latvia, 540 (95% CI, 474 to 606) per 100 000; Slovakia, 549 (95% CI, 487 to 611) per 100 000; Croatia, 600 (95% CI, 532 to 668) per 100 000; Lithuania, 814 (95% CI, 740 to 887) per 100 000; and Bulgaria, 1070 (95% CI, 955 to 1186) per 100 000.

The cumulative excess deaths per 100 000 population from 2020 to 2024 by country and age group, relative to expected deaths based on 2015-to-2019 mortality is shown in Table 3. Notably, 22 of 34 countries recorded negative or insignificantly positive cumulative excess mortality from 2020 to 2024 for those aged 0 to 14 years. However, only 6 countries recorded reduced excess mortality for individuals aged older than 14 years, among which 4 of 6 countries were in quintile 1, and the remaining 2 countries had statistically insignificant reductions. In all countries, older age groups recorded much greater cumulative excess mortality compared with younger age groups across the entire 2020-to-2024 period.

The annual excess mortality per 100 000 population and the estimated mortality displacement by country from 2020 to 2024 are provided in Table 4. Even though in quintile 1, the highest observed mortality displacement ratios were found in Denmark at 180% and in Luxembourg at 2770%, these results were statistically insignificant and likely reflect their relatively low early excess mortality combined with substantial subsequent mortality deficits, rather than a large absolute displacement of deaths. However, 3 countries exhibited significant mortality displacement. In particular, Greece at 10% (95% CI, 4%-15%), Latvia at 21% (95% CI, 14%-28%), and Poland at 21% (95% CI, 17%-25%) showed mortality displacement ratios with statistical significance. By 2024, the US had returned to its prepandemic stable trend (3 [95% CI, −2 to 7] per 100 000). In contrast, most European countries (including Norway, France, Switzerland, the Netherlands, Belgium, Spain, the UK, Austria, Italy, and Lithuania) had not yet resumed their prepandemic trajectories, exhibiting excess mortality rates ranging from 11 (95% CI, 3 to 18) per 100 000 in France to 115 (95% CI, 94 to 135) per 100 000 in Lithuania.

The annual excess death rate and mortality displacement by country and age group are presented in eFigures 1-5 and eTable 1 in Supplement 1. Significant mortality displacement were observed in 0 of 3 countries in the 0-to-14–year age group, in 6 of 15 countries in the 15-to-64–year age group, in 5 of 15 countries in the 65-to-74–year age group, and in 4 of 8 countries in the 75-to-84–year age group, with effect magnitudes ranging from 7% to 68% in the 15-to-64–year age group, 8% to 62% in the 65-to-74–year age group, and 9% to 42% in the 75-to-84–year age group. Notably, in the 85 years or older age group, the mortality displacement was more pronounced, with 10 of 13 countries showing significant mortality displacement, ranging from 6% to 106%. Across all 25 significant age-group findings, 40% were from countries in mortality-burdened quintile 5, and 32% were from countries in mortality-burdened quintile 4, indicating that the highest burdened countries were disproportionately associated with the observed significant mortality displacement. The mortality displacement did not differ significantly between females and males in the younger age groups, as shown in eTable 2 in Supplement 1. However, in the 85 years or older age group, the difference was more pronounced, with 12 of 34 countries exhibiting insignificantly higher mortality displacement in females compared with only 3 countries where mortality displacement was higher in males.

The sensitivity analyses generally supported the robustness of our findings (eTables 4-6 in Supplement 1). For Greece, Latvia, and Poland, mortality displacement ratios and 95% CIs were very similar across the negative binomial, the AR(1) generalized estimating equation, and alternative age-band specifications, and all 3 countries remained classified as having significant mortality displacement. In the US, however, the 2024 excess mortality estimate was close to 0 and not statistically different from the baseline in the primary model (3 [95% CI, −2 to 7] per 100 000) but became a modest deficit under the AR(1) generalized estimating equation specification (−9 [95% CI, −14 to −3] per 100 000), consistent with a possible mortality displacement. Because these estimates are small in magnitude and lie near the null, we interpreted the US mortality displacement signal as sensitive to modeling assumptions about serial correlation. After applying the false discovery rate correction, the pattern of statistical significance generally remained consistent with the unadjusted P values (eTable 7 in Supplement 1). This suggests that controlling for multiple testing was not associated with the interpretation of our results.

This cross-sectional study found the magnitude and heterogeneity of the recovery and sustained decline in mortality across 34 countries following the COVID-19 pandemic, with results stratified by age, sex, and mortality quintile. The primary outcome of this research assesses how excess mortality during the pandemic was associated with subsequent mortality displacement. Statistically significant mortality displacement was seen in only 3 countries where pandemic-era excess mortality was high (Greece, Latvia, and Poland). They were predominantly observed among the oldest age group (≥85 years).

Our study also found the pace of recovery of pre–COVID-19 mortality trends. By 2024, the US had returned to its prepandemic pattern of relatively stable all-cause mortality. In contrast, as of 2024, most European countries, including Norway, France, Switzerland, the Netherlands, Belgium, Spain, the UK, Austria, Italy, and Lithuania, had still not resumed their prepandemic mortality trajectories. Among the 8 countries (Portugal, Slovenia, Czechia, Greece, Bulgaria, Croatia, Latvia, and Poland) with prepandemic positive mortality trends, 5 countries (Portugal, Slovenia, Czechia, Bulgaria, and Croatia) continued to experience significant excess mortality or sustained elevated baseline mortality. Of the 8 countries that had rising pre–COVID-19 mortality trends, only those achieving negative excess mortality would represent a return to conditions that were normal for high-income health systems. This situation prompts consideration of whether persistent outcomes from the pandemic, associated with sustained health system strain or depleted population health capital, may have continued to influence mortality patterns through 2024.

Our study also offers insight into the cumulative outcome of COVID-19 when including both the pandemic period of 2020 to 2022 and the recovery years of 2023 to 2024. Luxembourg uniquely demonstrated a statistically significantly faster-than-expected reduction in mortality, whereas New Zealand and Denmark maintained their expected rates of mortality decline. New Zealand had statistically significantly negative cumulative excess mortality from 2020 to 2024 in all age groups except ages 75 to 84 years. However, this was exceptional. For the 7 countries in the best quintile of low COVID-19–era mortality, most age groups experienced positive cumulative excess mortality when summing 2020 to 2024. Notably, Denmark and Luxembourg, with the lowest COVID-19–era mortality, exhibited significant cumulative negative excess mortality among individuals aged 85 years or older from 2020 to 2024. This is hard to explain and may be potentially due to factors such as smaller sample sizes, imprecise denominators, high mobility among older populations, and the suppression of influenza during 2020.

Most top-performing countries regarding overall excess mortality did not succeed in sustaining overall declines in mortality from 2020 to 2024. The worse the cumulative death rate was in 2020 to 2022, the worse the cumulative mortality was for 2020 to 2024. In other words, mortality displacement did not compensate for failing to protect the population during the pandemic. Although younger cohorts showed lower net mortality, associated with the 50-year trend toward better health of younger ages, the significant and persistent excess mortality among older adults, even in the presence of mortality displacement, demonstrates that targeted protection alone was insufficient. The scale of mortality displacement observed in the most affected countries may further indicate that any shielding strategies used in those countries failed to adequately protect high-risk groups.

Multiple factors contribute to the COVID-19 mortality displacement. Extensive studies have found that the pandemic disproportionately affected older adults and individuals with preexisting comorbidities, including type 2 diabetes,^27-29 high blood pressure,^30-32 and obesity.^33-35 Additionally, in countries with high COVID-19 mortality rates, individuals at elevated risk might either succumb to or recover from the virus earlier, thus reducing subsequent expected mortality rates.²⁴

Our findings have several policy implications. First, many European countries as of 2024 have yet to resume the typical pattern of mortality decline that characterized the last century of progress. The resumption of a postpandemic mortality decline in Greece, Latvia, and Poland plausibly may be driven primarily by the displaced timing of deaths of at-risk individuals, especially older adults in countries with high mortality, and may reflect a return to prepandemic trajectories rather than substantive improvements beyond the 2015 to 2019 downtrend. Pandemic management policies must prioritize protecting high-risk populations through comprehensive, population-wide measures, as targeted shielding alone was insufficient. In parallel, integrating retrospective analyses of age-specific excess mortality into routine surveillance frameworks may help detect potential mortality displacement. While real-time monitoring may be limited by data lags, periodic assessment of cumulative excess deaths across age groups, especially among older adults, may inform evaluations of recovery and health system resilience. Finally, future research should aim to disentangle temporary mortality displacement from genuine recovery, ensuring that public health strategies are based on an accurate understanding of a pandemic’s impact and ongoing vulnerabilities.

Few studies, to our knowledge, have explicitly quantified mortality displacement after COVID-19 across such a large set of countries and demographic groups, making this a unique contribution to understanding postpandemic dynamics. A key strength of this study is its comprehensive inclusion of 34 countries with high-quality data, stratified by age, sex, and mortality quintile, which enabled the detection of heterogeneous patterns that would be masked in aggregate analyses. The use of nationally reported data, subjected to rigorous quality checks, with standardized formats, longitudinal and cross-country comparability, and detailed survival data at the oldest ages further enhanced the study’s external validity and generalizability.

This study has several limitations. First, despite using harmonized and validated national data, reporting quality and completeness may vary across countries, potentially introducing bias. Second, our analysis was ecological and based on aggregated age-group and country-level data, which limits our ability to adjust for individual-level factors such as comorbidities, socioeconomic status, or vaccination status, and raises the possibility of an ecologic fallacy; that is, inferences made about individuals from group-level data may not hold at the individual level. Third, demographic dynamics during and after the pandemic, such as selective migration of older adults, shifts in the age distribution, or differential survival of frail individuals, may vary across countries and age groups and could generate patterns that resemble mortality displacement even in the absence of true mortality displacement. Fourthly, while our study quantifies the association of mortality displacement with postpandemic mortality declines and underscores the importance of considering the mortality displacement when interpreting mortality trends after 2020, it does not fully disentangle this temporary outcome from longer-term recovery associated with other factors such as vaccination, health care improvements, and socioeconomic recovery. In addition, the observational nature of the study precludes causal inference, and findings should be interpreted in light of these constraints.

The findings of this cross-sectional study suggest that 31 of 34 countries with high-quality mortality data exhibited no statistically significant evidence of mortality displacement. In contrast, for Greece, Latvia, and Poland, countries with above-average COVID-19 mortality, part of the postpandemic decline may reflect mortality displacement, particularly among the oldest age groups. However, the scale of early excess mortality far exceeded what displacement alone could make up for COVID-19. This underscores a widely seen failure of shielding strategies to adequately protect broadly defined individuals. Recognizing this distinction is essential for accurately interpreting postpandemic mortality trends and for designing policies that effectively safeguard broader populations without propagating the myth that older individuals who died in the COVID-19 era were already near death. Future research should further disentangle mortality displacement from genuine recovery to establish a clearer causal understanding of the factors driving postpandemic mortality trends, as well as unravel why so many European countries have failed to resume their pre–COVID-19 pattern of mortality decline.

Accepted for Publication: November 30, 2025.

Published: January 29, 2026. doi:10.1001/jamanetworkopen.2025.55442

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2026 Chen X et al. JAMA Network Open.

Corresponding Author: David Makram Bishai, MD, PhD, Director’s Office, G/F, School of Public Health, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Patrick Manson Building (North Wing), 7 Sassoon Rd, Hong Kong SAR, China (dbishai@hku.hk).

Author Contributions: Prof Bishai had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Bishai.

Acquisition, analysis, or interpretation of data: Chen, Ye, Cowling.

Drafting of the manuscript: Chen, Ye.

Critical review of the manuscript for important intellectual content: Chen, Cowling, Bishai.

Statistical analysis: Chen, Bishai.

Obtained funding: Bishai.

Administrative, technical, or material support: Chen, Bishai.

Supervision: Bishai.

Conflict of Interest Disclosures: Prof Cowling reported being a consultant for AstraZeneca, Fosun Pharma, GlaxoSmithKline, Haleon, Moderna, Novavax, Pfizer, Roche, and Sanofi Pasteur. No other disclosures were reported.

Funding/Support: This study is partly supported by HKU Daniel and Mayce Yu Medical Development Fund (No. 200010837).

Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 2.

Many women recognise the pattern. A routine procedure takes longer than expected. It’s more uncomfortable than promised. The doctor reassures them that this sometimes happens, or suggests anxiety or muscle tension might be playing a role. But often the explanation is simpler – and anatomical.

This mismatch between bodies and procedures isn’t related to rare conditions or specialist care. It reflects a recurring problem in everyday medicine. Many routine procedures were designed around male anatomy, and they don’t always work the same way on female bodies.

Take colonoscopy. It’s one of the most common investigations used to diagnose bowel disease and screen for cancer. Yet women are more likely than men to experience discomfort, require repositioning, or have an incomplete examination on the first attempt.

The reason lies in normal anatomy. On average, women have a longer and more mobile colon, particularly in the sigmoid segment that loops through the pelvis.

The female pelvis itself is broader and shallower, creating sharper angles as the bowel curves downward. These features make the scope more likely to bend and loop inside the bowel, slowing its progress and pulling on surrounding tissue – a major source of pain.

This isn’t abnormal anatomy. It’s normal anatomy that standard techniques don’t always take into account.

Urinary catheterisation is another routine procedure where anatomy matters. Although the urethra performs the same function in men and women, its length, course and anatomical context differ in ways that matter clinically.

In males, the urethra is long – around 18-22cm – and is usually described in three parts: the prostatic urethra, which is wide and fixed as it passes through the prostate; the membranous urethra, the narrowest segment as it crosses the pelvic floor; and the spongy (penile) urethra, which runs in a predictable course to a clearly identifiable external opening at the tip of the penis. Despite its length, the male urethra follows a stable path and ends at a prominent external landmark.

The female urethra is much shorter, usually about 3-4cm long, but lies within a more variable anatomical environment. From the bladder neck, it passes through the bladder wall and pelvic floor, before opening into the vulval vestibule at a meatus (the external opening of the urethra) closely related to the anterior vaginal wall.

Its position varies between individuals and across the life course, influenced by pelvic floor tone and hormonal status. In practice, this can make catheter insertion technically more difficult, increasing the likelihood of repeated attempts and discomfort – particularly in older women or those with atrophic tissue (thin, delicate tissue).

Lumbar puncture and spinal procedures show similar issues. Women tend to have a greater lumbar curve and different pelvic tilt, altering the angle at which a needle must pass between vertebrae. Mild spinal curvature is also more common in women. The procedure itself doesn’t change, but the geometry does, increasing the likelihood of multiple attempts and prolonged discomfort.

Even airway management, a cornerstone of anaesthesia and emergency medicine, reflects the same mismatch. Female airways are, on average, shorter and narrower. When equipment sizing and technique is based on a “standard” airway, women are more likely to experience sore throat and hoarseness afterward – effects often dismissed as minor, but rooted in anatomy rather than sensitivity.

Even something as commonplace as peripheral venous cannulation, the insertion of a small tube into a vein to deliver fluids, medications, or to take blood, reflects this mismatch. Women’s superficial veins are often smaller, less prominent and more mobile in soft tissue, making standard cannulation techniques more likely to result in repeated attempts, bruising and pain.

Design for variation, not exception

Doctors know bodies vary. In practice, many already adapt – choosing different patient positions, smaller instruments or altered techniques. But these adjustments are informal, inconsistently taught and rarely explained to patients.

Instead, difficulty is often bundled into vague categories: anxiety, tension, low pain tolerance or “one of those things”. The result is that women experience real, anatomy-driven discomfort without being told why, and may internalise it as a personal failing.

This matters. When discomfort is normalised or minimised, patients are less likely to return for screening, more likely to delay care, and more likely to mistrust reassurance that future procedures will be different.

None of this requires radical innovation. It requires naming the issue accurately. When procedures are taught and designed around a single reference body, predictable anatomical variation becomes an obstacle rather than a design feature.

Acknowledging that bodies differ – in length, curvature, mobility and spatial relationships – allows doctors to plan, explain and adapt more effectively.

Crucially, it also shifts the narrative. Instead of “this shouldn’t hurt”, the message becomes: “your anatomy means this procedure can be more challenging, and we’ll adjust it accordingly”.

SpaceX is requesting to launch up to one million satellites to create a network of orbiting data centers around Earth. 

Late on Friday, the company filed the request with the Federal Communications Commission, describing the project as a “constellation of satellites with unprecedented computing capacity to power advanced artificial intelligence (‘AI’) models and the applications that rely on them.”

The plan is shocking in its scope, dwarfing the existing Starlink constellation, which currently spans over 9,600 satellites in Earth’s orbit. 

In one 8-page document, SpaceX describes the company’s proposed “Orbital Data Center system.” “To deliver the compute capacity required for large scale AI inference and data center applications serving billions of users globally, SpaceX aims to deploy a system of up to one million satellites to operate within narrow orbital shells spanning up to 50 km each (leaving sufficient room to deconflict against other systems with comparable ambitions),” the company wrote. 

(FCC/SpaceX)

The same satellites would harness the sun’s energy, orbiting at “between 500 km and 2,000 km altitude and 30 degrees and sun-synchronous orbit inclinations,” the company adds. The orbiting data centers would also use “optical links,” or lasers, to connect with Starlink, using the existing satellite internet system to route traffic to users below. 

“Orbital data centers are the most efficient way to meet the accelerating demand for AI computing power,” the filing adds in bold, pointing to the growing energy costs of AI data centers on Earth. The company is also betting it can launch the space-based data centers at a rapid clip using SpaceX’s more powerful Starship vehicle, which is also crucial to upgrading Starlink with next-generation satellites. 

(FCC/SpaceX)

The company filed the request as SpaceX is preparing an initial public offering reportedly to help fund the orbital data center push. SpaceX CEO Elon Musk has previously indicated his Starlink technology has already created a foundation to build out the network of orbiting data centers.

However, the 1 million satellite request appears to be unprecedented and will likely face intense scrutiny from the FCC, along with potential critics and rival companies. Earlier this month, the Commission cleared a SpaceX request to operate another 7,500 satellites for the second-generation Starlink constellation, including at lower orbits. But the regulator stopped short of granting permission for the full 22,488. 

Details about SpaceX’s orbiting data centers, including their mass, were left vague. In the filing, SpaceX merely says that it “plans to design and operate different versions of satellite hardware to optimize operations across orbital shells.”

Satellite industry analyst Tim Farrar told PCMag: "This filing seems quite rushed and to be very early stage." But he said it provides a rationale for a reported merger between SpaceX and xAI, another Elon Musk company that was created to compete with OpenAI.

Farrar added: "SpaceX can't spend the $50B that the IPO is supposed to raise on its existing Starlink and Starship efforts, whereas xAI sorely needs as much money as possible to keep up with rivals."

About Our Expert

Judge rules out capital punishment, but 27-year-old still faces federal stalking charges and state-level murder charges.

A New York judge has dismissed murder and weapons charges against Luigi Mangione, meaning that prosecutors can no longer seek the death penalty in the case accusing him of killing UnitedHealthcare CEO Brian Thompson.

While district judge Margaret Garnett dismissed the charges punishable by death in her ruling on Friday, the 27-year-old still faces two counts of stalking in his federal case that could lead to a maximum punishment of life in prison, as well as state-level murder charges carrying the same penalty.

Recommended Stories

• list 1 of 2Luigi Mangione appears in court as lawyers seek to bar evidence at trial
• list 2 of 2US prosecutors to seek death penalty against CEO killing suspect Mangione

Mangione, the Ivy League-educated scion of a wealthy Maryland family, previously pleaded not guilty to federal murder, weapons and stalking charges for allegedly gunning down Thompson in December 2024.

Garnett’s decision foils the bid of US President Donald Trump’s administration to see him executed for what it called a “premeditated, cold-blooded assassination that shocked America”. The judge is an appointee of former President Joe Biden.

Thompson, 50, was killed as he walked to a midtown Manhattan hotel for UnitedHealth Group’s annual investor conference.

Surveillance video showed a masked gunman shooting him from behind, with police saying that “delay,” “deny” and “depose” were written on the ammunition, mimicking a phrase used to describe how insurers avoid paying claims.

Jury selection in the federal case is scheduled to begin on September 8.

The state trial hasn’t been scheduled yet, but Mangione had already entered separate not guilty pleas for murder, weapons and forgery charges in Manhattan state court.

This week, the Manhattan district attorney’s office sent a letter urging the judge in that case to set a July 1 trial date.

Attorney General Pam Bondi ordered Manhattan federal prosecutors last April to seek the death penalty against Mangione, following through on Trump’s pledge to pursue capital punishment.

Trump returned to office a year ago with a pledge to resume federal executions after they were halted under Biden.