INSEE notes a link between population density and excess deaths. Between March 2 and April 19, 2020, deaths increased by 39% in the Côte-d’Or and by 16% in the Saône-et-Loire compared to the average for the last five years.
26% more deaths between early March and mid-April 2020: dense municipalities are the most affected
The health crisis linked to the spread of Covid-19 has an impact on the total number of deaths. From March 2 to April 19, 2020, 22,140 additional deaths, or 26% more from all causes, occurred compared to the average of deaths during the same period between 2015 and 2019. The excess of deaths increased significantly to from the week of March 16 then tends to decrease from the week of April 6. It affects men and women as much and is especially strong for the elderly.
Île-de-France and Grand Est are the regions most affected by this excess mortality. The increase in deaths is strongest in densely populated areas, although the population is on average younger.
Noël Gascard, Bertrand Kauffmann, Aline Labosse (INSEE)
A gradual increase in the number of deaths compared to the last five years before a decline
Since March 2020, France has experienced a health crisis linked to the spread of Covid-19. This crisis has resulted in an excess of deaths recorded by INSEE in the civil status, which covers all deaths regardless of where it occurred and whatever the cause (sources). Thus, between March 2 and April 19, 2020, France counts 22,140 additional deaths compared to the same period evaluated over the last five years (Figure 1). While in an average week of this comparison period, France has 11,950 deaths, the average is 15,100 weekly deaths in 2020. This corresponds to an average of 2,160 deaths each day, compared to 1,710 during the comparison period. The latter are therefore more numerous and progress over the weeks until April 5 before starting to decrease the following week. Thus, from March 30 to April 5, 2020, the additional mortality reached the peak of + 6,730 weekly deaths, an increase of 56%, before initiating a decline (+ 5,390 weekly deaths or + 45% of 6 as of April 12, then + 3,020 weekly deaths, ie + 25% from April 13 to 19, Figure 2a).
This additional death should not be interpreted as the number of deaths linked to Covid-19, since INSEE is not the recipient of the causes of death of the deceased. In addition, deaths from other causes may also have changed compared to the comparison period. An obvious, although limited, effect of the implementation of the containment of the population since March 17, 2020 is the reduction in the number of deaths due to road accidents. According to the provisional Road Safety report, 154 people died on the road in March 2020, a hundred people less than on average in March in the last five years. In terms of the number of deaths linked to seasonal flu, the epidemic peaks were reached between January and February in 2020 as for the previous five winters. However, the duration of epidemics and the number of attributable deaths according to Santé publique France vary from year to year. If the 2019-2020 influenza did not cause an excess number of deaths until the end of its epidemic surveillance in mid-March (like that of 2015-2016), the 2018-2019 influenza was less strong than in previous years ( 8,100 deaths in total against 13,000 and 14,400 respectively the two previous episodes). The flu in 2017-2018 had also been remarkable for its length (16 weeks of epidemic until the end of March 2018) while during that of 2014-2015, the highest number of deaths from all causes combined since the winter of 2006-2007 had been reached with 18,300 dead.
In mainland France, the Île-de-France and Grand Est regions have the highest death surpluses (+ 96% and + 59% respectively from March 2 to April 19), ahead of Bourgogne-Franche-Comté (+ 28%) , Hauts-de-France (+ 22%) and Auvergne-Rhône-Alpes (+ 21%). During the comparison period, 1,470 and 1,060 inhabitants of Île-de-France and Grand Est died on average per week, regardless of location and region of occurrence. In Île-de-France, the weekly number of deaths recorded at the place of residence jumped by 37% the week of March 16 (Figure 2b). It more than doubled from that of March 23, then tripled during that of March 30 (ie + 121% and + 213%), before decreasing to + 187% the week of April 6; between April 13 and April 19, however, it was still up 113% from an average week in the comparison period. In the Great East, the increase in mortality starts from the week of March 9 (+ 20%) and reaches + 56% the week of March 16. The number of deaths then increases less sharply than in Île-de-France. It doubled for each of the weeks of March 23 and 30 (respectively + 105% and + 111%), before also falling to + 78% the week of April 6; between April 13 and 19, the number of deaths remains up 48%. In the French overseas departments, Mayotte is the only department affected by a very large excess of deaths in 2020 (+ 54% over the whole period), but on a smaller workforce. Conversely, New Aquitaine and Guyana show little or no excess death in 2020 compared to the comparison period.
Older people are more affected
Both sexes are similarly affected by excess mortality. For men, 11,210 additional deaths are recorded compared to the comparison period. For women, this excess is another 10,930 deaths. For both men and women, the difference in mortality with the comparison period is very clear from the age of 65 (Figure 3). The excess of deaths varies from 21% to 39% depending on sex and age group. Overall, 21,900 more people died in 2020 among the 65 years or older (11,200 men and 10,700 women) compared to the comparison period. More than 84% of these additional deaths concern people aged 75 or over, and 62% only those aged 85 or over.
Men over 75 experience a higher excess mortality than that of women. It is 35% for men and 28% for women. The differences in general health status between the sexes, and in particular the higher proportion of men with risk factors (overweight for example), could be an explanation for this excess mortality.
The excess of deaths for men is identical between 65-74 years and 75-84 years. Concerning women, it is lower on the age group 75-84 years compared to the age group 65-74 years. The latter phenomenon is linked to the lower mortality recorded over two years (2017 and 2019) for women belonging to this age group, most likely linked to the seasonal flu episode.
Before age 65, the evolution of the number of deaths compared to the comparison period varies greatly by age. This must be compared with the low number of deaths before age 65 regardless of gender (around 15% of all deaths). Thus, the excess of deaths for women aged 25 to 44 corresponds to only thirty additional deaths compared to 560 deaths in this age group.
The excess of deaths recorded for men is highest in Île-de-France (+ 101%), in the Grand Est (+ 64%) and in Mayotte (+ 58%). Île-de-France is a particularly affected region: deaths there more than doubled for men from the age group of 65 years or more (8,300 deaths in 2020 compared to 3,800 deaths over the period of comparison). For women, these three regions also show the highest surpluses, with rates similar to those of men.
Six departments (Haut-Rhin, Seine-Saint-Denis, Val-d’Oise, Val-de-Marne, Hauts-de-Seine and Paris) recorded a doubling of deaths for men between the two periods. For the Haut-Rhin, the number of deaths has even tripled for men in the 75-84 age group. Concerning women, these same departments stand out, with surpluses barely lower than for men: from + 140% for Haut-Rhin to + 97% for Val-de-Marne or Paris.
An epidemic that starts in the Grand Est and spreads to the Ile-de-France
Between March 2 and April 19, deaths in the place of residence (whatever that occurred) increased by at least 50% for at least one week for the inhabitants of three out of ten departments (Figure 4). These are mainly all of the departments of Île-de-France, most of the Grand Est region (except the Ardennes and Aube), as well as the south of Hauts-de-France, eastern Burgundy-Franche-Comté, Auvergne-Rhône-Alpes and Mayotte.
The trend in excess mortality per week highlights two territories. First, a significant increase in deaths appears for residents of the Haut-Rhin during the week of March 9 (+ 75% of weekly deaths). The following week, the date on which containment started in France, mortality in this department was three times higher than during the comparison period (+ 204%). The increase in mortality is increasing in five other geographically close departments (Vosges, Doubs, Territoire de Belfort, Moselle and Bas-Rhin with deaths increasing from 52% to 75% the week of March 16), then to the rest of the Grand region Is as well as part of Burgundy-Franche-Comté.
At the end of March, mortality was still strongly excess in the Haut-Rhin (+ 270% of weekly deaths after a peak at + 283% the week of March 23). Despite a sharp deceleration in early April, the department still has deaths up 135% the week of April 6 and 68% that of April 13 compared to an average week of the comparison period. In the bordering Territory of Belfort, the situation deteriorated until the week of April 6, reaching a number of deaths up 148% before also falling to + 22% in excess of that of April 13.
The second territory with a large excess of deaths is made up of Ile-de-France and southern Hauts-de France. The excess mortality first appears for the inhabitants of Oise, in the week of March 16, with 63% more deaths, as well as in Paris and Seine-Saint-Denis (+ 51%) . The increase in mortality then spreads to all the departments of Ile-de-France as well as to Aisne, Somme and Eure. The situation worsens in Île-de-France until the end of March. In the departments of Paris and its inner suburbs as well as in the Val-d’Oise, deaths are at least three times higher than on average (above + 211%) and up to + 290% in Seine-Saint -Denis the week of March 30, department with the highest excess mortality in France in any week. The rest of the Ile-de-France departments reached an increase of 158%. The week of April 6, the excess of deaths in Île-de-France begins to decrease, but mortality remains strongly excess in Seine-Saint-Denis, in Val-d’Oise, Hauts-de-Seine and Val-de-Marne, where weekly deaths are at least three times higher than on average (+ 256%, + 211%, + 211% and + 208% respectively). The week of April 13, these four departments as well as Paris and Essonne still have an excess of deaths which ranges from + 102% to + 162% while that of Seine-et-Marne and Yvelines are respectively + 99% and + 65%.
Often different places of death and domicile in Île-de-France
People do not necessarily die in their department of residence. In 2020, for a third of the departments, more than 10% of deaths take place outside the department of domicile. For two departments, this difference exceeds 20%: Seine-Saint-Denis and Eure. In the latter, more than one death in ten takes place in a neighboring department, Paris and Seine-Maritime respectively.
In the other direction, certain departments receiving patients from other departments experience a number of deaths on their territory greater than the number of deaths of their residents. This is particularly the case in the Territoire de Belfort, where more than 60% of the deceased did not live in this department in 2020. Hospitals in the region traditionally welcome patients from the neighboring departments of Doubs and Haute-Saône. To a lesser extent, this is also the case for Paris, Hauts-de-Seine and Val-de-Marne, where more than 30% of the deaths recorded are for people domiciled outside the department. Paris has the particularity of welcoming patients from almost all departments, with a specificity for residents of Hauts-de-Seine, Val-de-Marne and Seine-Saint-Denis.
A link between population density and excess deaths
The more people there are in a small area, the greater the risk of contact. The increase in mortality between March 2 and April 19 increases with municipal density (Figure 5). It is in the densest municipalities that the increase in mortality is greatest (+ 49% against + 26% in France). Conversely, in the least dense territories, deaths in 2020 are rather stable compared to previous years.
Even excluding Île-de-France which concentrates 40% of the French population living in dense territory, the excess of deaths remains high in dense municipalities (+ 22% against + 17% on the whole territory outside Île -of France). In addition, although globally concerned by an increase in mortality, all the dense municipalities of France are not affected. Among those with more than 100,000 inhabitants, Saint-Denis and Mulhouse stand out with the highest surpluses over the period (+ 172% and + 165%), followed by Strasbourg (+ 111%) and Argenteuil (+ 109%) then Paris (+ 98%) and Montreuil (+ 90%). At the other end, Clermont-Ferrand, Brest or Caen have no excess mortality.
The excess mortality is lower in the least dense territories, although the population there is on average older, and therefore more fragile compared to Covid-19. Those aged 75 and over represent 8% of the population in dense territories, but 11% in very sparse territories. The main explanation that can be put forward is that, in sparsely populated areas, interactions between individuals are less frequent and therefore the risk of contamination is lower.
From the week of March 16, significant differences in mortality appear according to the density of the territories. The number of weekly deaths in dense municipalities is then 22% higher compared to the comparison period. These differences increase the following week and the surplus of dense municipalities increases (+ 62%) until doubling compared to the comparison period the week of March 30 (+ 106%) then decreases (+ 92%) the week of April 6 and April 13 (+ 54%). Changes per week are more moderate in municipalities with intermediate density (up to + 43% the week of March 30 before falling to + 19% the last week). They rose to a surplus of + 28% in the sparsely populated municipalities before declining to + 8%, while in the less dense territories, the increase in mortality reached its maximum during the week of March 30 (+ 14%) .
Communicated
–
