Editor's evaluation: Direct and indirect mortality impacts of the COVID-19 pandemic in the United States, March 1, 2020 to January 1, 2022

Abstract

Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Appendix 1 Data availability References Decision letter Author response Article and author information Metrics Abstract Excess mortality studies provide crucial information regarding the health burden of pandemics and other large-scale events. Here, we use time series approaches to separate the direct contribution of SARS-CoV-2 infection on mortality from the indirect consequences of the pandemic in the United States. We estimate excess deaths occurring above a seasonal baseline from March 1, 2020 to January 1, 2022, stratified by week, state, age, and underlying mortality condition (including COVID-19 and respiratory diseases; Alzheimer’s disease; cancer; cerebrovascular diseases; diabetes; heart diseases; and external causes, which include suicides, opioid overdoses, and accidents). Over the study period, we estimate an excess of 1,065,200 (95% Confidence Interval (CI) 909,800–1,218,000) all-cause deaths, of which 80% are reflected in official COVID-19 statistics. State-specific excess death estimates are highly correlated with SARS-CoV-2 serology, lending support to our approach. Mortality from 7 of the 8 studied conditions rose during the pandemic, with the exception of cancer. To separate the direct mortality consequences of SARS-CoV-2 infection from the indirect effects of the pandemic, we fit generalized additive models (GAM) to age- state- and cause-specific weekly excess mortality, using covariates representing direct (COVID-19 intensity) and indirect pandemic effects (hospital intensive care unit (ICU) occupancy and measures of interventions stringency). We find that 84% (95% CI 65–94%) of all-cause excess mortality can be statistically attributed to the direct impact of SARS-CoV-2 infection. We also estimate a large direct contribution of SARS-CoV-2 infection (≥67%) on mortality from diabetes, Alzheimer’s, heart diseases, and in all-cause mortality among individuals over 65 years. In contrast, indirect effects predominate in mortality from external causes and all-cause mortality among individuals under 44 years, with periods of stricter interventions associated with greater rises in mortality. Overall, on a national scale, the largest consequences of the COVID-19 pandemic are attributable to the direct impact of SARS-CoV-2 infections; yet, the secondary impacts dominate among younger age groups and in mortality from external causes. Further research on the drivers of indirect mortality is warranted as more detailed mortality data from this pandemic becomes available. Editor's evaluation The authors examine the impacts of the COVID-19 pandemic on excess mortality in the US up to January 2022. The authors separate direct impacts of the pandemic from indirect impacts (disruptions), finding that most excess deaths (84%) are due to direct impacts. Moreover, in individuals under 44 years of age, indirect effects predominate in mortality from external causes and all-cause mortality. The paper is well written and of interest to understant the impacts of the COVID-19 pandemic. https://doi.org/10.7554/eLife.77562.sa0 Decision letter Reviews on Sciety eLife's review process Introduction As the official death toll of the coronavirus disease 2019 (COVID-19) continues to grow, the full impacts of the pandemic on a range of conditions remain debated. By the end of January 2023, official statistics reported 1,056,000 deaths in the United States alone (Johns Hopkins University, 2022), although burden estimates that do not rely on official tallies suggest a higher death toll (Weinberger et al., 2020; Karlinsky and Kobak, 2021). A pandemic of the magnitude of COVID-19 would be expected to have secondary effects on unrelated health conditions; for instance, non-COVID-19 deaths increased in Spring of 2020 at the height of the first wave in part due to avoidance of the healthcare system (Bollmann et al., 2020; Kansagra et al., 2020; Mafham et al., 2020; Woolf et al., 2020; Woolf et al., 2021). Excess mortality approaches have been used for over a century to capture the full scope of large-scale infectious disease events, heatwaves, and earthquakes, by measuring the rise in mortality over a historical baseline (Serfling, 1963; Weinberger et al., 2020). In the early phase of the pandemic, these approaches highlighted substantial underestimation in official statistics of COVID-19 deaths due to limited viral testing (Weinberger et al., 2020; Kobak, 2021). More recent analyses have examined excess mortality patterns for specific causes of death by age and socio-demographic groups and have compared the COVID-19 death toll between countries (Banerjee et al., 2020; Islam et al., 2021; Karlinsky and Kobak, 2021; Mena et al., 2021; Rossen et al., 2021; Woolf et al., 2021; COVID-19 Excess Mortality Collaborators, 2022). Yet, separating the direct impact of SARS-CoV-2 infection on mortality from the other consequences of the pandemic remains challenging. To measure the direct and indirect effects of the COVID-19 pandemic, it is important to enumerate the mechanisms that could generate these effects and understand how they would manifest in mortality statistics. We first consider the direct effects of COVID-19 as those deaths that resulted from SARS-CoV-2 infection and its complications. When there is evidence of SARS-CoV-2 infection in the days or weeks before death, either virologically or clinically, it is likely that these deaths will receive a COVID-19 code during certification and these deaths will appear in official statistics. We would, however, expect variation between states in the death certification process for COVID-19. Additionally, a number of deaths triggered by SARS-CoV-2 infection could result from a complicated and protracted pathologic process, especially in patients with multiple underlying chronic conditions, who may lack a history of SARS-CoV-2 testing, and whose death may not be ascribed to COVID-19. For instance, a death in a diabetic patient could have been triggered by an undetected SARS-CoV-2 infection, resulting in a primary code of diabetes, with a SARS-CoV-2 code either lacking or listed as contributing condition. We would then expect a rise in diabetes mortality to coincide with a rise in COVID-19 cases. A similar phenomenon has been reported for influenza, with mortality from chronic conditions rising concomitantly with influenza-associated respiratory mortality in epidemic and pandemic seasons (Reichert et al., 2004; Quandelacy et al., 2014). In addition to the direct impacts of COVID-19, there will be positive and negative changes in mortality during the pandemic period that are not associated with SARS-CoV-2 infection and its complications. We refer to these changes as indirect impacts. Reasons for these changes include avoidance of the healthcare system for treatment of acute conditions and for management of underlying chronic conditions, stressed healthcare systems in a period of high COVID-19 incidence, mental health issues in families of patients severely affected by COVID-19, societal disruptions (Sharma et al., 2021), decreased social interaction that depresses circulation of endemic pathogens, and decreased air pollution. The indirect impacts of the pandemic on mental health, violence, and addiction remain particularly debated, with potentially large impacts on mortality (Faust et al., 2021b; Faust et al., 2021c). These indirect mortality changes may or may not coincide temporally with COVID-19 waves. In the United States, there was substantial geographic and temporal heterogeneity in the trajectory of the COVID-19 pandemic, along with differences in the strength and types of interventions implemented to mitigate COVID-19. In a large country with standardized death ascertainment like the United States, these heterogeneities provide an opportunity to separate the contributions of viral infection from other drivers of mortality. Here, we apply time series approaches to four large waves of COVID-19 from March 1, 2020 to January 1, 2022, to separate the direct consequences of SARS-CoV-2 infection on age- state- and cause-specific mortality from the indirect consequences associated with hospital strain and interventions. Our analyses indicate that the direct and indirect effects of the pandemic vary substantially by chronic condition and age group. A better understanding of these effects is particularly important for the mitigation of future large-scale pandemics. Results Overall mortality patterns We compiled weekly US mortality data by age and state, from August 1, 2014 to January 1, 2022 for eight underlying conditions (all causes, respiratory conditions, Alzheimer’s disease, cancer, cerebrovascular diseases, diabetes, heart diseases, external causes; see supplement for disease codes and Centers for Disease Control and Prevention, 2022a; Centers for Disease Control and Prevention, 2022b). External causes include suicides, accidents, homicides, and poisoning from opioids and other substances, among other conditions. Respiratory mortality includes deaths ascribed to COVID-19 (ICD code U07), influenza, pneumonia, and chronic lower respiratory diseases. This is our most specific indicator of excess deaths directly attributable to SARS-CoV-2 infection. We used the data until March 1, 2020 to calibrate seasonal regression models and project expected mortality baselines in the absence of a pandemic (See methods for details). Models were adjusted for influenza circulation. Pandemic excess mortality was the difference between observed and expected baseline mortality from March 1, 2020 to January 1, 2022 (see https://github.com/viboudc/DirectIndirectCOVID19MortalityEstimation; Viboud, 2023 for data and code). Across the United States from March 1, 2020 to January 1, 2022, there were 848,866 cumulative deaths officially attributed to COVID-19, namely, with COVID-19 as the underlying cause of death. During the same period, we estimate 757,600 (95% Confidence Intervals (CI) 725,200–788,100) excess respiratory deaths and 1,065,200 (95% CI 909,800–1,218,000) excess deaths due to all-cause (Table 1). National mortality patterns comprise four waves from March 1 to June 20, 2020 (wild-type variant); June 21 to September 19, 2020 (wild-type variant); September 20, 2020 to June 19, 2021 (wild-type and Alpha variants); June 20 to November 11, 2021 (Delta variant). A recrudescence of mortality in the last weeks of 2021 was attributable to the co-circulation of the Delta and Omicron variants. The timing and intensity of mortality varied greatly by state (Figure 1A and Appendix 1—figures 1–8). The first wave was concentrated in Northeastern states, while Southern and Western states experienced mortality increases during later waves. A sensitivity analysis based on the length of the historic data used for calibration of the model baseline is shown in Appendix 1—figure 9. All-cause and respiratory disease estimates, as well as national estimates, were particularly robust to the choice of the calibration period. Figure 1 Download asset Open asset Weekly mortality rates (per 100,000) for select US jurisdictions and validation of COVID-19 excess mortality estimates against serology. (A) Weekly all-cause mortality rate per 100,000 in the United States and top five most populated states, August 2, 2014 to January 1, 2022. Black lines show observed data. Green line shows the seasonal model baseline. The red solid line shows the seasonal variation accounting for influenza circulation. The orange shaded areas show the upper and lower 95% confidence intervals (CIs). The dotted vertical red line marks the start of the pandemic on March 1, 2020. (B) Comparison between estimated excess respiratory mortality rates and cumulative COVID-19 seroprevalence estimates from the Centers for Disease Control and Prevention (CDC) as of December 31, 2021. Each point corresponds to a state; observations are shown for 16 states which have enough resolution in respiratory mortality data. Error bars represent 95% CIs on serology and excess mortality estimates. The black line and dotted region represent a linear regression fit and the associated 95% CI for a model without intercept. Table 1 Reported COVID-19 deaths by US jurisdiction, Compared with Excess Deaths from All-Causes and Respiratory Diseases: March 1, 2020 to January 1, 2022. JurisdictionEstimated excess all-cause deaths per 100,000, (95% prediction interval)No. estimated excess all-cause deaths (95% prediction interval)No. estimated excess respiratory deaths (95% prediction interval)No. reported COVID-19 deaths*Ratio of COVID-19 deaths to all-cause excess deathsRatio of COVID-19 deaths to respiratory excess deathsUnited States318 (272–364)1,065,200 (909,800–1,218,000)757,600 (725,200–788,100)848,8860.801.12Alabama569 (406–727)26,900 (19,200–34,400)15,000 (13,000–16,800)16,4250.611.09Arizona414 (328–498)35,000 (27,700–42,100)22,900 (21,000–24,600)23,3810.671.02Arkansas450 (283–612)13,800 (8700–18,700)NA93630.68NACalifornia286 (234–336)120,500 (98,700–141,800)75,500 (69,800–80,700)81,9100.681.08Colorado284 (173–393)15,000 (9100–20,800)NA11,2800.75NAConnecticut238 (98–374)8700 (3600–13,700)NA94511.08NAFlorida342 (279–404)80,200 (65,400–94,600)54,900 (51,400–58,200)60,7040.761.11Georgia366 (287–443)39,600 (31,100–48,000)25,400 (23,100–27,600)27,7630.701.09Illinois269 (198–339)35,600 (26,200–44,800)25,300 (22,800–27,600)28,5090.801.13Indiana325 (213–436)21,600 (14,100–28,900)16,400 (14,100–18,500)19,8300.921.21Iowa195 (26–361)5900 (800–10,900)NA82951.41NAKansas241 (75–401)7000 (2200–11,600)NA73161.05NAKentucky421 (278–560)18,600 (12,300–24,800)NA13,3130.71NALouisiana451 (317–581)21,300 (15,000–27,400)NA13,9840.66NAMaryland262 (163–360)17,000 (10,600–23,400)NA12,8320.75NAMassachusetts196 (95–297)13,500 (6500–20,300)NA15,7991.17NAMichigan251 (165–336)26,800 (17,600–35,900)22,200 (19,600–24,700)27,2871.021.23Minnesota149 (42–253)8800 (2500–14,900)NA11,0151.25NAMississippi477 (305–645)14,500 (9300–19,600)NA11,0690.76NAMissouri309 (190–426)19,200 (11,800–26,400)15,400 (13,300–17,200)17,0050.891.11Nevada348 (211–479)12,000 (7300–16,500)NA91720.76NANew Jersey320 (238–401)30,300 (22,500–38,000)25,500 (23,600–27,300)27,7700.921.09New York353 (287–416)69,000 (56,300–81,500)56,700 (48,400–63,600)62,3390.901.10Ohio400 (296–502)46,600 (34,500–58,500)30,900 (27,800–33,700)35,6330.771.15Oklahoma416 (263–566)15,600 (9800–21,100)NA13,0980.84NAOregon209 (76–338)8900 (3200–14,400)NA58010.65NAPennsylvania347 (259–434)44,400 (33,100–55,500)34,500 (31,800–37,000)38,9540.881.13South Carolina437 (301–570)21,100 (14,500–27,500)NA15,2020.72NATennessee411 (292–526)27,800 (19,800–35,700)20,200 (18,000–22,200)21,5800.781.07Texas364 (311–417)104,300 (89,000–119,400)76,000 (72,200–79,500)82,3280.791.08Virginia236 (149–320)21,000 (13,300–28,500)12,800 (10,700–14,800)15,8240.751.22Washington172 (77–264)12,800 (5700–19,600)NA98680.77NAWisconsin219 (−23–43)13,100 (−1400–26,600)NA12,3620.94NA * As reported by National Center for Health Statistics. States are ordered alphabetically. No. of reported COVID –19 deaths (any death with COVID-19 as underlying cause) until December 31, 2021 as available on June 14, 2022, were obtained from the NCHS website (Centers for Disease Control and Prevention, 2022d). Next, to validate our excess mortality approach, we compared our estimates with serology (see methods for details). Excess respiratory mortality showed a significant, positive correlation with CDC seroprevalence surveys (Centers for Disease Control and Prevention, 2022g) conducted in late December 2021 in each state (Figure 1B). Seroprevalence estimates ranged between 11.1 and 47.7% across states, with a population-weighted national seroprevalence of 34.6%. New York and Alabama experienced higher than predicted excess mortality with respect to their reported serologic infection rates, while Illinois and Michigan had the reverse pattern. The nationwide infection fatality rate (IFR) was estimated at 0.67% (95% CI 0.60–0.73%) based on excess respiratory mortality and 0.89% (95% CI 0.77–1.02%) based on all-cause excess mortality (Appendix 1—figure 10). Sensitivity analyses based on the maximum reported seroprevalence at any time point of the study period indicate that New York remained an outlier, with Illinois and Texas showing the reverse pattern (Appendix 1—figure 10). Use of official COVID-19 deaths determined an IFR of 0.72% (95%CI 0.62–0.81%); interestingly, serology was more highly correlated with excess respiratory deaths than with official COVID-19 deaths (Appendix 1—figure 10). The IFR was significantly higher in individuals over 65 years, estimated at 5.5% (95% CI 4.5–6.6%) based on all-cause excess respiratory mortality. Next, we compared the mortality burden of COVID-19 and influenza. We estimated excess mortality for the severe November 2017 to March 2018 influenza A/H3N2 season and for the large wave of COVID-19 in November 2020 to March 2021 (see Appendix for details). We find that nationally, over this 5 month period, the mortality burden of COVID-19 was 5.7-fold higher than that of influenza based on all-cause excess mortality. A similar pattern was seen in all states median ratio of COVID-19 to influenza excess mortality rates across states, 5.8 (IQR, 5.0–7.8), (Appendix 1—figure 11). Direct and indirect pandemic impacts by cause of death To probe the direct and indirect mortality impacts of the pandemic, we assessed whether the trajectories of various mortality categories were synchronous with that of respiratory mortality. Synchronicity would signal a direct impact of SARS-CoV-2 infection on these mortality categories. Overall, during the March 1, 2020 to January 1, 2022 pandemic period, excess mortality increased for 6 of the 7 non-respiratory conditions studied, although the timing and intensity of the rise varied by disease (Table 2, Figure 2). Cancer was the only mortality condition that did not increase during the pandemic. Cancer deaths have remained below historic levels since March 2020, although cumulative departures from baseline were not significantly different from zero (Table 2). In contrast, mortality from chronic conditions such as Alzheimer’s, diabetes, and heart disease rose during the pandemic, with the trajectory of excess mortality matching the pattern of respiratory mortality in the 4 pandemic waves (Figure 2 for national patterns, and Appendix 1—figures 3–8 for state-specific data). Across these causes of death, the first excess mortality peaks occurred within one week of the first respiratory mortality peak on April 18, 2020, with the most pronounced synchronicity patterns observed in the first wave in New York and New Jersey. Across chronic conditions, the peak of excess mortality was highest during the winter of 2020 to 21, or the third wave of the pandemic (Appendix 1—figure 12). Table 2 Estimation of the direct impacts of COVID-19 on non-respiratory conditions. Cause of DeathNo. estimated excess deaths (95% prediction interval)% of excess deaths directly attributable to COVID-19 (95% prediction interval)*All-cause1,065,200 (909,800–121,8000)84% (65, 94)Alzheimer’s25,300 (12,600–37,600)70% (45, 89)Diabetes24,700 (15,900–33,300)70% (45, 93)Heart diseases51,300 (7,400–94,300)73% (32, 94)Cerebrovascular diseases16,600 (5,300–27,800)26% (−17, 62)External causes102,800 (81,400–123,700)−48% (−64, −23)†Cancer4,300 (−18,100–26,500)N/A‡ * Regression estimates of the direct impact of COVID-19 on cause-specific excess mortality, where weekly cause-specific excess mortality is regressed against COVID-19 intensity, strength of interventions, and ICU occupancy, using gam models. Estimates are based on comparison of predictions from the full model with counterfactual predictions where the COVID-19 term is set to zero. † COVID-19 intensity is significant but negatively associated with excess mortality from external causes, hence the estimated attributable fraction is negative. ‡ COVID-19 intensity is not retained in the cancer model. Figure 2 Download asset Open asset Weekly national mortality rates and model baselines (per 100,000) for eight causes of death. The black line shows observed data, the green line shows the seasonal model baseline, the orange shaded areas the 95% Confidence Interval (CI) on the seasonal baseline, and the red line shows model predictions with seasonal variation and influenza circulation. Excess mortality attributed to the COVID-19 pandemic is defined as the area between the black and green line from March 1, 2020 onwards. The dotted black vertical line marks the start of the pandemic on March 1, 2020. We found a significant rise in deaths from external causes during the pandemic period from March 1, 2020 to January 1, 2022, corresponding to 102,800 (95% CI 81,400–123,700) cumulative excess deaths nationally (Figure 2 and Table 2). The largest excess mortality rates from external causes were found in states that also had high baseline death rates from these conditions (Appendix 1—figure 13). However, the weekly trajectory of mortality from external causes did not align with that of respiratory mortality. We further analyzed subcategories of external causes that were available on a monthly resolution (see Figure 3 and methods for data). The largest excess death tolls observed during this period were from accidents and injuries (43,600 excess deaths (95% CI 17,200–70,000), a 12% increase over baseline), drug overdoses (25,300 deaths (95% CI 12,000–38,700), 16% increase), and assaults and homicides (8,000 deaths (95% CI 3,700–12,200), 20% increase, Table 3). Overdoses were the first to peak in May 2020, followed by accidents and assaults in July 2020. Notably, mortality from suicides remained at historic levels throughout the end of the study period. Figure 3 Download asset Open asset Monthly national deaths by subcategory of external causes of death from January 2014 to December 2021. The black line shows observed data, the green line shows the seasonal model baseline, and the orange shading represents the 95% Confidence Interval (CI) on the seasonal baseline. The dotted red vertical line marks the start of the pandemic period of excess mortality on March 1, 2020. Table 3 Excess mortality for different subcategories of external deaths during the COVID-19 pandemic period, March 2020 to December 2021. Estimates are based on a seasonal regression model fitted to monthly data (as shown in Figure 3). Underlying cause of deathNo of excess deaths (95% prediction intervals)Ratio of excess deaths to baseline deaths (95% confidence intervals)*Accidents (unintentional injuries)43,600 (17,200–70,000)0.12 (0.05–0.2) Motor vehicle accidents†9,600 (1,000–18,200)0.13 (0.01–0.24)Drug overdoses25,300 (12,000–38,700)0.16 (0.07–0.24)Assaults and homicides8,000 (3,700–12,200)0.2 (0.09–0.31)Suicides3,000 (−7,000–13,100)0.04 (−0.08–0.16) * This should be interpreted as the percent increase over baseline. For instance, mortality from accidents increased by 12% (95% CI, 5–20%) during the period March 2020 to December 2021 (p<0.05), relative to baseline pre-pandemic levels. † Motor vehicle accidents are a subcategory of accidents. We saw evidence of increased synchronicity in multiple causes of death during the pandemic, which is a signature of the direct effects of SARS-CoV-2 infection on mortality. During the period March 1, 2020 to January 1, 2022, and compared to historical patterns, all-cause mortality became more correlated with excess deaths from respiratory conditions in all 16 states with available respiratory estimates (Appendix 1—figure 14 and methods for details). States that experienced high cumulative excess respiratory deaths had concomitantly high excess mortality from all-causes (Spearman rho=0.81, 95% CI: 0.48–0.94), attesting to the large impact of COVID-19 on total mortality (Appendix 1—figure 15). Synchrony between excess deaths from underlying respiratory diseases and excess deaths from underlying chronic conditions increased during the pandemic in a subset of states (Appendix 1—figure 14), particularly for diabetes (n=8 states), Alzheimer’s (n=5), heart diseases (n=4), and cerebrovascular diseases (n=4). In contrast, excess deaths recorded as due to cancer or external causes showed either no change (in most states) or declining synchrony (in one or two states) with respiratory mortality during the pandemic. Next, to quantify the direct and indirect impacts of the pandemic on different causes of death, we used GAM to regress weekly cause-specific excess mortality on official COVID-19 deaths, the strength of non-pharmaceutical interventions, and hospital ICU occupancy (see methods for statistical approach, and Figure 4 and Appendix 1—figure 16 for results). We used official COVID-19 deaths as a proxy for the direct impact of SARS-CoV-2 infection on mortality. The variables measuring the strength of interventions (Oxford contingency index, Oxford University, 2021) and ICU occupancy (Health and Human Services, 2022) allowed for the estimation of the indirect consequences of the pandemic on mortality. We found a major direct impact of SARS-CoV-2 infection on mortality from all-cause, diabetes, heart disease, cerebrovascular diseases, and Alzheimer’s; namely, the strongest predictor of excess mortality from these causes was COVID-19 deaths. The relationship between these mortality conditions and COVID-19 deaths, while non-linear, was typically monotonically increasing. Non-pharmaceutical interventions and ICU occupancy variables were also statistically associated with excess mortality, although the form of the relationship was more complex. Non-pharmaceutical intervention variables had a curvilinear relationship with excess deaths, consistent with different mechanisms affecting different periods of the pandemic. At lower levels of interventions (measured by the Oxford contingency index between 0 and 50), representing the early stages of the lockdown in March 2020, excess mortality rose with interventions. Later in the pandemic, increased interventions were estimated to have a beneficial effect on excess mortality, driven by comparison between late 2020 when interventions were strengthened in response to increasing COVID-19 activity (Oxford index above 60), and Spring 2021 when interventions were relaxed (Oxford index between 50 and 60). The relationship with the ICU occupancy variable was more difficult to interpret, varied between causes of death, and had the lowest statistical significance of the three variables tested. Furthermore, all mortality conditions were not equally well captured by our models: the best model fit was for all-cause mortality (R2=96%) and the worst was for cerebrovascular diseases (R2=47%; Figure 4). Figure 4 Download asset Open asset Observed and predicted excess death rates by condition, United States, March 1, 2020 to January 1, 2022, using generalized additive models (GAM) with weekly COVID-19 deaths, intensive care unit (ICU) occupancy, and a proxy for the strength of interventions as covariates. Observed values are in black and predicted values are in red (mean=dark red, 95% Confidence Interval (CI) in lighter red). See also Appendix 1—figure 16 for a comparison of predicted and observed values, and Appendix 1—figures 18 and 19 for age-specific models. On a national level, the GAM approach estimated that 84% (95% CI 65–94%) of all-cause excess deaths were attributable to the direct impact of SARS-CoV-2 infection, while the proportion was 73% (95% CI, 32–94%) for heart diseases, 70% (95% CI: 45–89%) for Alzheimer’s, and 70% (95% CI: 45–93%) for diabetes (Table 2). The contribution of COVID-19 on cerebrovascular diseases was not statistically significant. Applying a similar GAM approach to excess mortality from cancer and external causes revealed that these conditions were more strongly associated with the intervention and ICU occupancy variables than with COVID-19. Stricter interventions were associated with a nearly linear increase in external cause mortality and a decline in cancer mortality. COVID-19 deaths had a negative effect on excess mortality from external causes (i.e. high COVID-19 activity coincided with fewer excess deaths from external causes, Table 2). The model for cancer had the worst fit of all conditions studied, while the model for external causes had an intermediate fit (R2=18% vs 58% respectively). State-level analyses yielded similar estimates of direct and indirect pandemic effects as in national analyses (Appendix 1—figure 17). The median proportion of all-cause excess deaths attributed to direct COVID-19 effects was 81% by the GAM approach (inter-quartile range across states, 63–90%, Appendix 1—figure 17). State-level analyses confirmed the direct impact of COVID-19 on Alzheimer’s, diabetes, and heart diseases, although the effect size was generally attenuated compared to national analyses. Consistent with national analyses, the effect of COVID-19 on mortality from cerebrovascular disease and cancer was low or non-significant, while COVID-19 had a negative effect on mortality from external causes. Pandemic age mortality patterns Next, we ran some of the same analyses on age-specific data. The total burden and direct impacts of the COVID-19 pandemic from March 1, 2020 to January 1, 2022 varied substantially by age (Figure 5, Table 4). As many prior studies have reported, all-cause excess death rates increased monotonically with age. Individuals 85 years and older, the age group with the highest death rate, accounted for 17% of excess mortality, while individuals under 25 years accoun