Relationship between COVID-19 cases and monthly mortality from all causes, cancer, cardiovascular diseases and diabetes in 16 countries, 2020–21 Authors: Margherita Pizzato, Claudia Santucci, Nazrul Islam, Carlo La Vecchia, Gianfranco Alicandro Published: 2025 ABSTRACT AbstractBackgroundDuring the COVID-19 pandemic, mortality from some chronic diseases increased. In this study, we evaluated monthly excess mortality from all causes, cancer, cardiovascular diseases (CVD) and diabetes during the months of 2020 and 2021, examining its relationship with COVID-19 cases.MethodsMonthly cause-specific mortality data were downloaded from public repositories of national statistics offices or directly requested from them, and population data were obtained from the United Nations archives. Excess deaths were estimated as the difference between observed and expected deaths. Monthly expected deaths for 2020 and 2021 were calculated using a quasi-Poisson regression model trained on 2010–19 data (or a shorter timespan if the full decade of data was not available). To quantify the correlation between COVID-19 cases and monthly excess mortality, we used the Spearman’s correlation coefficient (rs).ResultsThe study included 16 countries that provided monthly national data on causes of death (Argentina, Austria, Brazil, Switzerland, Chile, the Czech Republic, Germany, Georgia, Hungary, Italy, Lithuania, Latvia, Mexico, Serbia, Slovakia and the USA). A positive correlation was found between COVID-19 cases and monthly excess mortality from all causes in all countries (rs ranging from 0.61 to 0.91), from CVD in 11 countries (rs ranging from 0.45 to 0.85) and for diabetes in 13 countries (rs ranging from 0.42 to 0.79). Excess mortality above 5% was estimated from all causes in 14 countries for both 2020 and 2021, from CVD in seven countries for 2020 and in nine countries for 2021, and from diabetes in 11 countries for 2020 and in 12 countries for 2021. No excess above 5% was estimated for cancer mortality in any of the countries considered.ConclusionsExcess mortality from CVD and diabetes persisted in several countries throughout 2021. These increases coincide with COVID-19 peaks, supporting a short-term impact of the COVID-19 pandemic on mortality from these causes. We evaluated the monthly cause-specific excess mortality from cardiovascular diseases, diabetes and cancer in 16 countries during 2020 and 2021, in relation to COVID-19 peaks. We evaluated the monthly cause-specific excess mortality from cardiovascular diseases, diabetes and cancer in 16 countries during 2020 and 2021, in relation to COVID-19 peaks. Excess mortality from cardiovascular diseases and diabetes persisted through 2021 in several countries, whereas no excess from cancer was observed. Excess mortality from cardiovascular diseases and diabetes persisted through 2021 in several countries, whereas no excess from cancer was observed. Excess mortality from cardiovascular diseases and diabetes coincided with COVID-19 peaks, suggesting a short-term effect of the pandemic on these causes of death. Excess mortality from cardiovascular diseases and diabetes coincided with COVID-19 peaks, suggesting a short-term effect of the pandemic on these causes of death. Introduction Our prior analysis of cause-specific excess mortality in 30 countries using data from the World Health Organization (WHO) revealed a rise in mortality from diabetes and cardiovascular diseases (CVD) in the first pandemic year, with no increase in cancer mortality [1]. These findings were later confirmed by other research groups in 2021 in some countries [2–5]. Whereas these studies provided estimates on an annual basis, a more detailed evaluation of cause-specific excess mortality over shorter time frames can offer new insights into the timing of COVID-19 peaks and mortality from other conditions. Specifically, such analysis would allow for a deeper understanding of the pandemic's impact on cause-specific mortality by assessing both its immediate and its later effects. Immediate pandemic effects are expected to manifest during COVID-19 peaks and may include direct effects of SARS-CoV-2 infection on the cardiovascular system and glycometabolic control, as well as increased mortality among clinically vulnerable individuals who did not access health care or did not receive adequate treatment during these periods. Later effects may manifest outside COVID-19 peaks and encompass health complications among COVID-19 survivors, such as respiratory or cardiovascular sequelae, and increased mortality from other conditions due to delayed or deferred health care [6, 7]. In this study, we collected cause-specific mortality data from 16 European and American countries to evaluate the relationship between monthly excess mortality from all causes, cancer, CVD and diabetes and peaks in COVID-19 cases in 2020 and 2021. Methods Data source At the end of 2023, we searched public repositories of national statistics offices to obtain monthly cause-specific mortality data for countries with completeness of cause-of-death data above 90% [8], disaggregated by sex, age category and cause of death for the period 2010–21. When data were not publicly available, we sent requests to the national statistics offices to obtain data either free of charge or for a fee covering processing costs. We obtained data from 16 countries, i.e. Argentina, Austria, Brazil, Chile, Switzerland, the Czech Republic, Germany, Georgia, Hungary, Italy, Lithuania, Latvia, Mexico, Serbia, Slovakia and the USA. The first available calendar year varied by country: 2011 for most countries, 2010 for Lithuania, 2014 for Georgia and 2015 for Argentina, Italy, and Serbia. Data were available through 2021 for all selected countries except Chile, which had data only through 2020. Different age groups were provided (Supplementary Table S1, available at IJE online). Causes of death were grouped by the national statistical offices using the code of the 10th Revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10): C00-C99 for cancers, I00-I99 for CVD and E10-E14 for diabetes. For Slovakia, information on diabetes was not available. We evaluated mortality from CVD and diabetes, since in a previous work we documented excess mortality for these conditions in several countries in 2020 [1]. Although in our earlier study we did not find any excess in cancer mortality, we included it in the current analysis to investigate whether disruptions in screening and care during the first year of the pandemic affected cancer outcomes in 2021. We chose not to evaluate other causes of death, due to anticipated large fluctuations in monthly mortality data for less common causes, which could complicate their interpretation. Mid-year population data disaggregated by sex, age group and calendar year were obtained from the United Nations Archive [9]. COVID-19 cases registered over the months of 2020 and 2021 were obtained from Our World in Data [10]. This dataset contains actual confirmed cases based on official reports from national health agencies. Estimates of excess mortality Excess deaths for 2020 and 2021 were calculated as the difference between the observed number of deaths and the expected number of deaths. The expected number of deaths for each country and cause of death was obtained by extrapolating the coefficients of a quasi-Poisson regression model. The model was fitted using total and cause-specific mortality, along with population data from a pre-pandemic reference period, and then applied to the 2020 and 2021 data: logEdeathsi,j,k,l= α+β1sexi+fagej+ β2sexi · fagej+β3yeark+β4(sexi·yeark)+ fmonthl +logpopi,j,k where: E (deathsi,j,k,l​) is the expected number of deaths for sex i, age group j, year k and month l; E (deathsi,j,k,l​) is the expected number of deaths for sex i, age group j, year k and month l; α is the intercept; α is the intercept; β1, β2, ... are the coefficients associated with each term in the model; β1, β2, ... are the coefficients associated with each term in the model; f(agej) denotes the natural spline function ns(age, df= 3); f(agej) denotes the natural spline function ns(age, df= 3); sexi​⋅yeark​ is the interaction term between sex and year; sexi​⋅yeark​ is the interaction term between sex and year; f(monthl) denotes the natural spline function ns(month, df= 3); f(monthl) denotes the natural spline function ns(month, df= 3); log(popi,j,k) represents the natural logarithm of the population size, used as an offset term to control for differences in population size and age distribution over time. log(popi,j,k) represents the natural logarithm of the population size, used as an offset term to control for differences in population size and age distribution over time. The reference period varied across countries depending on the available data. Age was included in the model as the midpoint of age classes, computed by averaging the lower and upper boundaries. For open-ended age classes (e.g. ‘85+’), we used a weighted mean based on available population data for older age classes. We replaced the original age categorizations with the midpoint in order to avoid over-parameterization. Due to the limited number of categories in the mortality data from Germany and Hungary, age was included as a categorical variable for these countries. Empirical 95% confidence intervals (CIs) for the point estimate of excess deaths were derived via Monte Carlo simulation. We generated 1000 realizations of the regression coefficients from a multivariate normal distribution, using their point estimates and the variance-covariance matrix. For each realization, we calculated the difference between observed and expected deaths and obtained the 95% CI using the normal approximation. Relative excess mortality was calculated as (observed deaths—expected deaths)/expected deaths * 100. All statistical analyses were performed using software R version 4.3.0. No ethical committee approval was required since we used anonymized aggregated data. Correlation between COVID-19 cases and excess mortality To evaluate whether the cause-specific excess mortality estimated for each month was related to the number of COVID-19 cases registered in the same month, we computed the Spearman’s correlation coefficient (rs). This metric ranges from −1 to 1, with values close to −1 indicating a negative correlation and values close to 1 indicating a positive correlation. Values ≥0.7 were considered indicative of a strong correlation. Sensitivity analysis To account for the sensitivity of excess mortality estimates to the chosen baseline period for modelling expected deaths [11, 12], we estimated excess mortality using all possible combinations of calendar years preceding the pandemic. These results were presented graphically, showing the distribution of the obtained estimates, and as extreme quantiles (2.5th and 97.5th percentiles). To evaluate whether the estimated excess mortality was an exceptional event limited to the pandemic period, we used the same methodology to estimate the difference between observed and expected deaths in 2019. These results were then compared with those obtained for the pandemic years. To mitigate the potential effect of undercounting during the initial phase of the pandemic [13], the correlation analysis was conducted separately for 2020 and 2021. Results This study included 16 countries that provided monthly data on causes of death: 11 from Europe (Austria, the Czech Republic, Germany, Georgia, Hungary, Italy, Latvia, Lithuania, Serbia, Slovakia, Switzerland), four from Central and Latin America (Argentina, Brazil, Chile, Mexico), and the USA. Table 1 shows the estimates of excess deaths during 2020 and 2021 from all causes, cancer, CVD and diabetes, by country. In 2020, Mexico had the highest excess mortality from all causes (+45.9%), while Germany (+3.2%), Latvia (+3.8%) and Austria (+8.2%) had the lowest ones. Other countries reported excess deaths ranging from 10% to 17%. In 2021, 12 countries experienced higher excess mortality from all causes compared with 2020. Mexico again had the highest excess (+48%), followed by Serbia (+40.3%), Slovakia (+36.7%), Georgia (+35.2%) and Brazil (+30.9%). Table 1.Excess mortality and corresponding 95% confidence interval (CI) from all causes, cancer, cardiovascular diseases and diabetes in 16 countries in 2020 and 2021CountryYearAll causesCancersCardiovascular diseasesDiabetesExcess deaths (95% CI)% (95% CI)Excess deaths (95% CI)% (95% CI)Excess deaths (95% CI)% (95% CI)Excess deaths (95% CI)% (95% CI) ARG 202035 149 (28 568,41 729)10.4 (8.4,12.3)−2466 (-3185,-1746)−3.9 (-5.0,-2.7)738 (-969,2445)0.8 (-1.0,2.6)868 (636,1099)9.7 (7.1,12.2)202192 887 (84 532,101 241)27.6 (25.1,30.1)−1709 (-2617,-800)−2.7 (-4.1,-1.3)10 647 (8491,12 802)11.2 (9.0,13.5)847 (555,1138)9.5 (6.2,12.8) AUT 20206950 (6119,7780)8.2 (7.2,9.2)543 (270,815)2.6 (1.3, 3.8)183 (-208,574)0.6 (-0.6,1.8)−193 (-272,-113)−6.3 (-8.9,-3.7)20216994 (6025,7962)8.2 (7.1,9.4)280 (-35,595)1.3 (-0.2, 2.8)−696 (-1144,-247)−2.2 (-3.6,-0.8)248 (156,339)8.2 (5.1,11.2) BRA 2020174 504 (143 727,205 280)12.7 (10.4,14.9)−12 935 (-16 308,-9561)−5.3 (-6.7,-3.9)−11 972 (-18 602,-5341)−3.2 (-5.0,-1.4)8537 (7528,9545)12.7 (11.2,14.2)2021431 359 (395 105,467 612)30.9 (28.3,33.5)−12 578 (-16 595,-8560)−5.1 (-6.7,-3.4)10 821 (3081,18 560)2.9 (0.8,5.0)10 086 (8898,11 273)14.8 (13.1,16.5) CHE 20207868 (7190,8545)11.5 (10.5,12.5)−591 (-823,-358)−3.3 (-4.5,-2.0)302 (42,561)1.5 (0.2,2.8)−181 (-228,-133)−15.2 (-19.1,-11.2)20212498 (1706,3289)3.6 (2.5,4.8)−728 (-999,-456)−4.0 (-5.5,-2.5)−317 (-614,-19)−1.6 (-3.1,-0.1)−84 (-138,-29)−7.1 (-11.7,-2.5) CHL 202015 568 (14 434,16 701)14.1 (13.1,15.1)−356 (-648,-63)−1.2 (-2.2,-0.2)351 (17,684)1.2 (0.1,2.4)−280 (-381,-178)−7.1 (-9.7,-4.5) CZE 202017 597 (16 638,18 555)15.8 (14.9,16.6)455 (191,718)1.6 (0.7, 2.5)5362 (4917,5806)11.7 (10.7,12.6)232 (117,346)4.9 (2.5, 7.3)202128 575 (27 463,29 686)25.7 (24.7,26.7)−452 (-756,-147)−1.6 (-2.7,-0.5)3115 (2613,3616)7.0 (5.8,8.1)58 (-83,199)1.2 (-1.6, 4.0) DEU 202030 874 (18 022,43 725)3.2 (1.9,4.6)−947 (-2818,924)−0.4 (-1.2, 0.4)7496 (2644,12347)2.3 (0.8,3.7)1324 (857,1790)5.4 (3.5, 7.3)202163 831 (47 303,80 358)6.7 (4.9,8.4)−3846 (-6243,-1448)−1.6 (-2.6,-0.6)15 005 (8911,21 098)4.6 (2.7,6.5)1460 (864,2055)6.0 (3.5, 8.4) GEO 20204995 (4209,5780)11.0 (9.2,12.7)−312 (-517,-106)−3.7 (-6.2,-1.3)1258 (872,1643)6.0 (4.2,7.9)227 (175,278)30.3 (23.3,37.1)202115 600 (14 657,16 542)35.2 (33.1,37.3)−1230 (-1498,-961)−13.9 (-16.9,-10.8)1429 (956,1901)6.9 (4.6,9.2)412 (352,471)57.7 (49.3,66.0) HUN 202013 340 (10 576,16 103)10.5 (8.3,12.6)135 (-629,899)0.4 (-1.9, 2.8)3221 (2385,4056)5.1 (3.8,6.5)730 (640,819)24.1 (21.1,27.0)202129 841 (26 680,33 001)23.7 (21.2,26.2)−366 (-1233,501)−1.2 (-3.9, 1.6)4233 (3279,5186)6.9 (5.3,8.4)1009 (904,1113)33.3 (29.8,36.7) ITA 202094 809 (86 709,102 908)14.7 (13.4,15.9)−2928 (-4618,-1237)−1.6 (-2.6,-0.7)4690 (1478,7901)2.1 (0.7,3.6)3822 (3399,4244)17.6 (15.7,19.6)202155 071 (44 148,65 993)8.5 (6.8,10.2)−4752 (-7000,-2503)−2.7 (-3.9,-1.4)−1540 (-5793,2713)−0.7 (-2.7,1.2)3083 (2519,3646)14.3 (11.7,16.9) LTU 20204776 (4354,5197)12.3 (11.2,13.4)172 (17,326)2.1 (0.2, 4.0)1682 (1390,1973)7.9 (6.5,9.3)107 (71,142)18.4 (12.2,24.4)20219995 (9518,10 471)26.5 (25.3,27.8)−231 (-407,-54)−2.8 (-5.0,-0.7)2557 (2231,2882)12.5 (10.9,14.1)82 (36,127)13.0 (5.7,20.1) LVA 2020988 (629,1346)3.8 (2.4,5.1)43 (-88,174)0.7 (-1.5, 3.0)164 (-65,393)1.1 (-0.5,2.7)64 (23,104)10.6 (3.8,17.2)20216835 (6420,7249)26.0 (24.5,27.6)−116 (-268,36)−2.0 (-4.6, 0.6)1965 (1700,2229)13.6 (11.8,15.4)73 (23,122)11.5 (3.6,19.2) MEX 2020341 657 (329 120,354 193)45.9 (44.2,47.6)1195 (367,2022)1.2 (0.4, 2.1)64 261 (62 787,65 734)32.5 (31.8,33.3)41 698 (40 956,42 439)37.8 (37.1,38.5)2021360 257 (345 546,374 967)48.0 (46.0,49.9)−334 (-1304,636)−0.3 (-1.3, 0.7)64 697 (62 959,66 434)32.3 (31.5,33.2)27 931 (27 058,28 803)25.0 (24.3,25.8) SRB 202016 660 (14 965,18 354)16.6 (14.9,18.3)−578 (-973,-182)−2.6 (-4.4,-0.8)4013 (2946,5079)7.8 (5.7,9.9)212 (57,366)6.7 (1.8,11.5)202139 224 (37 117,41 330)40.3 (38.1,42.5)−1088 (-1587,-588)−5.0 (-7.3,-2.7)7196 (5886,8505)14.6 (11.9,17.2)237 (43,430)7.6 (1.4,13.8) SVK 20204971 (4364,5577)9.2 (8.1,10.3)−206 (-404,-7)−1.4 (-2.8, 0.0)2488 (2152,2823)10.1 (8.7,11.4)202119 734 (19 033,20 434)36.7 (35.4,38.0)−1265 (-1496,-1033)−8.8 (-10.5,-7.2)4262 (3883,4640)17.7 (16.1,19.3) USA 2020469 170 (439 275,499 064)16.1 (15.1,17.1)−2768 (-6166,630)−0.4 (-1.0, 0.1)44 247 (35 996,52 497)5.0 (4.1,5.9)13 564 (12 944,14 183)15.3 (14.6,16.0)2021484 547 (449 048,520 045)16.3 (15.1,17.5)−3939 (-7901,23)−0.6 (-1.3, 0.0)27 895 (18 103,37 686)3.1 (2.0,4.2)12 403 (11 664,13 141)13.6 (12.8,14.5)Country abbreviations (ISO 3166 - Country code): ARG, Argentina; AUT, Austria; BRA, Brazil; CHE, Switzerland; CHL, Chile; CZE, Czech Republic; DEU, Germany; GEO, Georgia; HUN, Hungary; ITA, Italy; LTU, Lithuania; LVA, Latvia; MEX, Mexico; SRB, Serbia; SVK, Slovakia; USA, United States of America. Excess mortality and corresponding 95% confidence interval (CI) from all causes, cancer, cardiovascular diseases and diabetes in 16 countries in 2020 and 2021 Country abbreviations (ISO 3166 - Country code): ARG, Argentina; AUT, Austria; BRA, Brazil; CHE, Switzerland; CHL, Chile; CZE, Czech Republic; DEU, Germany; GEO, Georgia; HUN, Hungary; ITA, Italy; LTU, Lithuania; LVA, Latvia; MEX, Mexico; SRB, Serbia; SVK, Slovakia; USA, United States of America. Cancer mortality was significantly lower than expected in 7 countries in 2020 and 10 countries in 2021, with most estimates being less than 5% except for Brazil (−5.3% in 2020 and −5.1% in 2021), Georgia (−13.9% in 2021) and Slovakia (−8.8% in 2021). Excess mortality above 5% from CVD was observed in 2020 in seven countries (the Czech Republic, Georgia, Hungary, Lithuania, Mexico, Serbia and Slovakia) and in 2021 in nine countries (Argentina, the Czech Republic, Georgia, Hungary, Lithuania, Latvia, Mexico, Serbia and Slovakia), with excesses up to 32.5% estimated for Mexico. For diabetes, excesses above 5% were estimated in 2020 in 11 countries (Argentina, Brazil, Germany, Georgia, Hungary, Italy, Lithuania, Latvia, Mexico, Serbia and the USA) and in 2021 in 12 countries (Argentina, Austria, Brazil, Germany, Georgia, Hungary, Italy, Lithuania, Latvia, Mexico, Serbia and the USA), with excesses in 2021 of 25% in Mexico, 33% in Hungary and up to 58% in Georgia. Figure 1 shows COVID-19 cases and estimates of excess mortality by month from January 2020 to December 2021 by country. Peaks in COVID-19 cases align with a concurrent rise in excess mortality from all causes in all countries. This pattern was partly observed also for excess mortality from CVD and diabetes. Monthly trends in COVID-19 cases (dotted line) and excess mortality (% difference from the expected number of deaths) from all causes, cancer, cardiovascular diseases and diabetes (continuous lines), by country, 2020–21. Country abbreviations (ISO 3166 - Country code): ARG, Argentina; AUT, Austria; BRA, Brazil; CHE, Switzerland; CHL, Chile; CZE, Czech Republic; DEU, Germany; GEO, Georgia; HUN, Hungary; ITA, Italy; LTU, Lithuania; LVA, Latvia; MEX, Mexico; SRB, Serbia; SVK, Slovakia; USA, United States of America. Figure 2 presents the correlation coefficients between monthly cause-specific excess mortality and reported COVID-19 cases across 16 countries. COVID-19 cases showed a positive correlation with excess mortality in all countries, except in Austria, Brazil, Chile and Italy and the USA for CVD mortality, and in Lithuania and Latvia for diabetes mortality. The correlation was strong (rs ≥ 0.7) for all-cause mortality in 14 countries, for CVD mortality in two countries and for diabetes mortality in five countries. Negative correlations were found between COVID-19 cases and estimates of excess mortality from cancer in Georgia and Slovakia. Results of the analysis conducted separately for 2020 and 2021 are reported in Supplementary Figure S1, available at IJE online. Correlation between COVID-19 cases and monthly excess mortality from all causes, cancer, cardiovascular diseases and diabetes by country, 2020–21. ALL, all causes; CAN, cancer; CVD, cardiovascular diseases; DIA, diabetes; rs, Spearman’s coefficient of correlation. *P < 0.05, **P < 0.01. Country abbreviations (ISO 3166 - Country code): ARG, Argentina; AUT, Austria; BRA, Brazil; CHE, Switzerland; CHL, Chile; CZE, Czech Republic; DEU, Germany; GEO, Georgia; HUN, Hungary; ITA, Italy; LTU, Lithuania; LVA, Latvia; MEX, Mexico; SRB, Serbia; SVK, Slovakia; USA, United States of America. Overall, estimates using different combinations of baseline periods were close to each other, with a few exceptions such as Germany for total and CVD mortality and the USA for cancer mortality (Supplementary Figure S2 and Supplementary Table S2, available at IJE online). For example in Mexico, 95% of the estimates of CVD excess deaths for 2021 fell between 63 047 and 65 475 deaths. In Germany, there was higher uncertainty, with 95% of the estimates for the same year ranging between −18 346 and 36 982 excess deaths. In the USA, 95% of the estimates of excess deaths from diabetes in 2021 were between 11 099 and 14 232 deaths. The majority of the estimates obtained for 2019 were close to 0, with some indicating lower than expected mortality. However, the excess mortality we estimated in 2020 and 2021 was not observed in 2019 (Supplementary Figure S3, available at IJE online). Discussion The novelty of our study lies in presenting new estimates of excess mortality from CVD and diabetes during 2020 and 2021, using monthly data that are not currently available in public repositories. These data allowed for a better understanding of the temporal relationship between COVID-19 waves and cause-specific excess mortality. We found persisting excess mortality from all causes in 2021 across all 16 countries, with important excess mortality from CVD in nine countries and from diabetes in 12 countries. Cancer mortality was lower than expected in 10 countries. Our study also shows that peaks in COVID-19 cases and excess mortality are closely aligned in time, indicating immediate effects of COVID-19 waves on mortality from CVD and diabetes. SARS-CoV-2 infection may be directly responsible for the excess mortality from CVD and diabetes. The virus causes cardiovascular dysfunction through multiple pathways. Endothelial damage associated with COVID-19 increases the risk of clot formation, and hypoxia forces cells into anaerobic metabolism, producing harmful metabolic by-products that cause further cellular damage [14]. SARS-CoV-2 invades cardiac cells via ACE2 receptors, leading to cardiomyocyte necrosis, releasing cytotoxic contents that weaken the heart muscle and impair its contractility [15]. The immune response to infection triggers inflammation within cardiac tissues, further compromising heart function and disrupting electrical conduction. SARS-CoV-2 infection has been associated with new-onset diabetes and severe hyperglycaemia, as well as worsening of glycaemic control in pre-existing diabetes [16]. SARS-CoV-2 replicates in pancreatic β-cells, inducing morphological and functional changes that disrupt insulin production [17, 18]. The viral entry and replication within β-cells can trigger apoptosis, diminishing the number of functional β-cells. The systemic inflammation associated with COVID-19 exacerbates this problem by increasing insulin resistance through cytokine-mediated pathways [19]. Moreover, corticosteroids have been largely used in the management of severe COVID-19 cases, potentially leading to severe hyperglycaemia in patients with undiagnosed diabetes or pre-diabetes [20]. The COVID-19 pandemic has posed important challenges for health care systems, affecting the management of chronic diseases. Hospitals and health care facilities re-allocated resources, including intensive care unit beds and medical staff, to manage COVID-19 patients, resulting in the deferral or cancellation of routine care for non-COVID conditions [21]. Many patients, fearful of virus exposure, avoided seeking medical assistance. Lockdown measures further restricted physical access to health care services, disproportionately affecting vulnerable populations reliant on regular medical support. The availability of COVID-19 vaccines had only a partial impact on excess mortality in 2021. Despite vaccine campaigns starting in early 2021, by the end of the first half of the year less than 50% of the population in most countries had been fully vaccinated [22]. In addition, the Omicron variant caused a massive increase in COVID-19 cases in December 2021, superimposed on the still prevalent Delta variant. The excess mortality from CVD and diabetes that we found in our study was previously reported in studies using annual data. In Mexico, a study based on death certificates reported excess mortality of 32.5% for ischaemic heart diseases and 36.8% for diabetes [5]. These estimates were obtained using a regression model trained on 2015–19 data to estimate the expected number of deaths in 2020–21. A comparison of death rates from CVD and diabetes in 2021 with the average figure of 2015–19 in Brazil resulted in estimates of 23% for CVD and +60% for diabetes, considerably higher than our estimates (+2.9% for CVD and +14.8% for diabetes) [3]. A study based on multiple cause-of-death data reported an +18% excess mortality in an Italian region (Veneto) in 2021, an estimate close to the one we found for the whole country (+14%) [23]. The 14% excess mortality from diabetes we estimated in 2021 in the USA is consistent with the 18% excess mortality reported in a recent study covering the period from March 2020 to March 2022 [24]. Regarding cancer mortality, other studies found no excess or even decreasing mortality in pandemic years, indicating that COVID-19 had no short-term impact on cancer mortality [2]. The decreased mortality can be partly attributed to COVID-19 acting as a competing cause of death in patients with cancer [25]. When interpreting our results, some important factors need to be considered. Estimates of excess mortality are subject to different methodological choices, including the statistical model, the baseline pre-pandemic period used to derive the model for the expected deaths, and the age categorization. Among these, the choice of baseline period emerged as a critical subjective choice in estimating excess deaths during the pandemic [26, 27]. To address this, we tried different baseline periods and found consistent excesses not only for total mortality, but also for mortality from CVD and diabetes. However, some exceptions were noted such as Germany (for all causes and CVD mortality) and the USA (for cancer mortality), where estimates showed higher uncertainty. Our model to estimate expected deaths included a linear term for the calendar year, assuming that the pre-pandemic trend would have continued unchanged in the absence of the pandemic. Cross-country comparisons, however, may be somewhat limited by the varying age categorizations in the datasets used for the analysis. Data from six countries (Austria, Brazil, Switzerland, Chile, Lithuania and the USA) were available in 5-year intervals up to 100+ years, but other countries provided less detailed data. For example in Germany and Hungary, the oldest open-ended age groups were 65+ and 60+, respectively. Although all the countries in our study followed WHO guidelines for coding causes of death, differences in diagnostic and certification practices, inadequate training of the certifying physicians and lack of understanding of the importance of accurate certification may compromise between-country comparisons. Even in high-income countries with well-established vital statistics systems, some deaths are classified as having ‘ill-defined’ causes, which include broad, uninformative diagnoses such as heart failure and senility. The frequency of these causes varies greatly between countries, with WHO reporting estimates in the countries considered in our study ranging between 3% and 65%, with the highest rates observed in Georgia and Argentina [28]. Additionally, we may have underestimated the overall impact of COVID-19 on diabetes-related mortality, as diabetes is often considered a contributing factor rather than a cause of death. This was demonstrated in a study on death certificates in the USA, which found 47.6% excess mortality when diabetes was considered as a contributing cause and only 18.4% when it was considered as an underlying cause of death [24]. In the early stages of the pandemic, the true impact of COVID-19 was not fully captured by national institutions, particularly in countries with a severe first wave and limited testing capacity. For example in Italy, about 44 000 excess deaths were estimated in March–May 2020, whereas only 33 000 COVID-19 deaths were officially recorded [29]. As the pandemic progressed, countries with robust health care systems and intensive testing protocols were better able to capture cases compared with those with limited resources [30]. This variation complicates international comparisons and may have influenced our assessment of the correlation between COVID-19 cases and excess mortality. To account for this, we provided separate estimates for 2020 and 2021, which confirmed a positive correlation between COVID-19 cases and excess mortality associated with CVD and diabetes in several countries. Finally, our analysis is limited to 2020 and 2021, as no data are available for more recent years. However, recent data indicate that excess total mortality has substantially decreased in 2023 [31]. Further studies are needed to evaluate potential long-term effects on cause-specific mortality, particularly cancer mortality. Conclusion In conclusion, our findings indicate that the excess mortality from CVD and diabetes, documented by previous studies in 2020, persisted in several countries through 2021. Additionally, the monthly analysis reveals that excess mortality from these causes coincides with COVID-19 peaks, suggesting the immediate effects of the pandemic on mortality from these conditions. Ethics approval Ethics approval was not required for this study as the data used were obtained from public repositories or provided by national statistical institutes in aggregated, non-identifiable form. Use of artificial intelligence (AI) tools Artificial Intelligence (AI) tools were not used for data collection or analysis, image or graphic production or the writing of the paper. Supplementary Material