Some authors have hypothesised that life expectancy at birth could reach 100 years in developed countries in the next decades. Using recent data for French women, it is shown that life expectancy changes from 1993 to 2016 are mainly linked to mortality rates of oldest women. In recent years, it happened that life expectancy of French women oscillated, because of for instance influenza epidemics killing mainly oldest frail people. It is hypothesised that in coming years, life expectancy of French women (and one day of men?) will only very slightly increase and will show oscillations because of increased mortality the years of severe influenza, heatwave and other events threatening the life of frail oldest people. This fate could also be that of the other developed countries in the future, which would mean that life expectancy has begun to plateau.

Life expectancy at birth (e0) has strongly increased for two centuries. It was for instance ca 33 and 37 years for French men and women, respectively, in 1806 [1], and it is ca 80 and 85 years today. There has been a debate on the limits of life expectancy since Vaupel [2] hypothesised that “half of the babies born in France this year (1996) will reach age 95 and that most of the girls will reach age 100.” At that time, life expectancy was increasing each year in developed countries, except in some cases such as the former Soviet Union after 1965 [3] where it stagnated and even decreased. Since then, it happened that e0 could stagnate and even decrease in some countries, such as the USA (e.g., [4, 5]) or the UK [6]. For these two countries, it appears that decreases are mainly due to less wealthy people, particularly young and middle-aged ones in the USA because of what has been called the opioid crisis [7], but elderly people have also been hit because of the austerity policy in the UK [8].

In France, e0 has decreased in recent years, in 2015 for men and in 2003, 2012, and 2015 for women (arrows in Fig. 1), but it does not seem that France (still?) suffers from the British and US issues. Rather, it has been argued that heatwave [9] and influenza [10] could be the culprits, and that elderly people, who are much frailer than younger ones, have suffered and died from them. It would mean that, now, e0 changes in France are mainly explained by mortality rate changes of elderly people. It was not the case in the past century when e0 quickly and strongly increased because of vaccination, sanitation, medical care, social progress, and so on, which all strongly decreased mortality rates of, particularly, babies, young, and middle-aged people, at a time when elderly people were less numerous than today. Therefore, it could be that, today, the mortality rate changes of younger people are so small that they have only a minor effect on e0, while those of the numerous elderly people have a strong impact. If it were the case, and except if mortality rates of elderly people could strongly decrease in the future, the obvious outcome is that e0 should not increase in coming years and could oscillate because of “bad” and “good” years. In other words, e0 could have already plateaued in France, in contrast to the Vaupel’s [2] hypothesis (see also [11, 12]). How to test this hypothesis?

Fig. 1.

Life expectancy at birth (e0) of French men and women, from 1993 to 2016. The arrows indicate the years when e0 decreased. The 2017 and 2018 data (including Mayotte), displayed for information, are provisional [17] and not involved in the computations. This is why they are not connected to the year 2016 on the graph.

Fig. 1.

Life expectancy at birth (e0) of French men and women, from 1993 to 2016. The arrows indicate the years when e0 decreased. The 2017 and 2018 data (including Mayotte), displayed for information, are provisional [17] and not involved in the computations. This is why they are not connected to the year 2016 on the graph.

Close modal

Mortality rates and e0 of French men and women from 1993 to 2016 were extracted from a database (metropolitan France) of the French National Institute for Statistics and Economic Studies (INSEE, https://www.insee.fr/fr/statistiques/fichier/1911933/irsocsd2013_t67_f.xls). 1993 was the first year used in the computations because it was the first year with a very low infant mortality (5‰) and 2016 is the last year for which data are available. Between 1993 and 2016, the infant mortality was similarly low in metropolitan France (ca in the range 3–5‰) and thus was not expected to strongly make e0 to vary. For each gender and age group, we computed the yearly changes of mortality rates and of e0. For instance, a negative value of the mortality rate change of 15- to 19-year-old men from 1993 to 1994 means that mortality rate has decreased in 1994 for this age group. For this age group, the 1993 and 1994 values were, respectively, 0.72 and 0.69‰, and thus the mortality rate change between 1993 and 1994 is 0.69–0.72 = –0.03‰. Similarly, a positive value of the e0 change from 1993 to 1994 means that e0 has increased in 1994. The 1993 and 1994 values were, respectively, 73.3 and 73.7 years, and thus the e0 change between 1993 and 1994 is 73.7–73.3 = 0.4 years. To sum up, the mortality rates of 15- to 19-year-old men has decreased in 1994 when compared to 1993, and e0 has increased. Yearly changes of mortality rates were then correlated with yearly e0 changes, for each gender and age group (younger than 1 year of age, 1–4, 5–9, 10–14, 15–19, 20–24, 25–29, 30–34, 35–39, 40–44, 45–49, 50–54, 55–59, 60–64, 65–69, 70–79, 80–89, 90–110 years of age; the INSEE table pools nonagenarians and centenarians, as the number of centenarians is rather low: in 2018, there were ca 1 million nonagenarians and less than 25,000 centenarians). This allows to know whether mortality rate changes of each age group were linked to e0 changes. The table reports the results of correlations for men and women between yearly mortality rate changes and yearly e0 changes, and Figures 2 and 3 show the regression graphs.

Fig. 2.

Regression of yearly e0 changes (years) according to yearly mortality rate changes (‰) for men older than 60 years of age. On each graph, the number near a point indicates the number of ties for this point. The equation is displayed at the top of the graph, and the r2 are shown in Table 1.

Fig. 2.

Regression of yearly e0 changes (years) according to yearly mortality rate changes (‰) for men older than 60 years of age. On each graph, the number near a point indicates the number of ties for this point. The equation is displayed at the top of the graph, and the r2 are shown in Table 1.

Close modal
Fig. 3.

Regression of yearly e0 changes (years) according to yearly mortality rate changes (‰) for women older than 60 years of age. On each graph, the number near a square indicates the number of ties for this point. The equation is displayed at the top of the graph and the r2 are shown in Table 1. The correlation is not significant in the 60–64 age group (see Table 1).

Fig. 3.

Regression of yearly e0 changes (years) according to yearly mortality rate changes (‰) for women older than 60 years of age. On each graph, the number near a square indicates the number of ties for this point. The equation is displayed at the top of the graph and the r2 are shown in Table 1. The correlation is not significant in the 60–64 age group (see Table 1).

Close modal

The mortality rate and e0 changes were significantly correlated in men older than 40 years, but also at 5–9 years of age (Table 1): when mortality rates decreased, e0 increased. Even if a Bonferroni correction made that some results were no longer significant, it remains that many age groups explained a part of the variance of yearly e0 changes. Figure 2 reports the regression graphs for men older than 60 years of age. The graphs show that, for most of the age groups and years, mortality rates decreased and e0 increased: most of the points in Figure 2 are located in the upper left quadrant. In other words, from 1993 to 2016, e0 of men increased and the favourable mortality rate changes of many age groups explained this result. Even when mortality rates of oldest men increased, e0 did not decrease, except in one case (in 2015, see Fig. 1).

Table 1.

Correlation between yearly mortality rate changes at various ages (in years) and yearly e0 changes: results for men and women from 1993 to 2016 (n = 23 years)

Correlation between yearly mortality rate changes at various ages (in years) and yearly e0 changes: results for men and women from 1993 to 2016 (n = 23 years)
Correlation between yearly mortality rate changes at various ages (in years) and yearly e0 changes: results for men and women from 1993 to 2016 (n = 23 years)

Thus, because e0 increases during the whole period are linked with decreased mortality rates at various ages, the increased mortality rates of elderly people have only a minor effect on e0 changes. The e0 of men does not appear to plateau because mortality rates at various ages, and not only at old age, dictate e0 changes, and mortality rates at these various ages are still decreasing. One can thus make the hypothesis that e0 of men will still increase in the coming years.

Mortality rate and e0 changes were significantly correlated only in women older than 65 years (Table 1). Figure 3 shows that, in contrast to men, many mortality rate increases were linked either to no e0 change or to a decrease, particularly in the oldest age group. In other words, from 1993 to 2016, e0 could either stagnate, increase, or decrease, and the mortality rate changes of elderly women mainly explained this result. Thus, while “bad” years did not decrease men’s e0, they could decrease e0 of women. Particularly, as e0 changes are linked only with mortality rate changes of elderly women, a “bad” year for 90–110 year-old women (Fig. 3) is often linked with an e0 decrease or a stagnation.

To sum up, one can thus make the hypothesis that e0 of women has maybe begun to stagnate and will now oscillate, depending on “bad” and “good” years. Even if a slight increase could be still observed, the main feature of e0 in coming years could be its oscillations rather than its increase, because these oscillations could be of higher importance (e.g., 15,000 extra-deaths of oldest women because of a flu) than the slight increase. A continuous increase in e0 would be only possible if it were possible to protect these very old and frail women against influenza, heatwave, and other threats, and to discover new means to decrease their mortality rates. This hypothesis appears to be fragile, as it would mean that very old women would no longer be frail.

It has been hypothesised [2, 11, 12] that median lifespan could still strongly increase in developed countries, up to 104 years for both sexes of the 2007 French birth cohort and 107 years for Japan (Table 1 in Christensen et al. [12]). One could argue that it is too early to falsify this hypothesis, but recent trends in France show that maybe it is possible right now.

Recent e0 increases in France were mainly due to elderly and not younger people. For instance, 0- to 64-year-old women added 0.27 years to the 1.92-year e0 increase between 1998 and 2010, while 65- to 110-year-old ones added 1.65 years (see Table 7 in Prioux and Barbieri [13]). However, as there are more and more women reaching very old ages, there is a small chance to strongly increase e0 because these women are close to the maximal observed lifespan. Indeed, while “saving” a baby can increase his/her lifespan by decades, “saving” a 90-year-old woman can only add a very few years to her life, in the best of the cases. In addition, because these oldest women are very frail, they are at risk to succumb to threats that are not very dangerous for younger people, such as heatwave and influenza [14]. In such conditions, e0 cannot strongly increase and can oscillate because of “bad” and “good” years. It seems that this is observed in French women, but not yet in men.

Thus, one can hypothesise that the e0 of French women is now plateauing and oscillating. Because mortality rates of younger women are now very low and no longer linked to e0 changes contrarily to those of elderly women, one can add that these oscillations are very probable in coming years. One can further hypothesise that these oscillations will not be striking in men while their e0 is still increasing, because the still decreasing mortality rates in men of various ages could buffer the oscillations of mortality rates of oldest men. Obviously, if it were possible to protect very old women against deadly threats, their mortality rates could still decrease and e0 increase without such oscillations, even if e0 does not reach 100 years and more, as hypothesised by Vaupel [2]. Indeed, stagnation of e0 is not a new phenomenon and, for instance, e0 stagnated in Sweden during the 1970s [15]. It could thus be argued that we face a new stagnation and that, one day, e0 will resume its increase. However, the main difference with previous episodes of stagnation is that e0 has already strongly increased allowing very frail women to live longer: could e0 still increase up to, say, 95 years, and thus could very frail women survive for many extra-years without succumbing to episodes of heatwave, influenza, and so on? This is perhaps a too strong hypothesis.

The coming years will tell us whether plateauing of e0 and oscillations will be observed in France, and in other developed countries. If they were observed, it would be the end of the multi-secular trend of the e0 increase, and the question would be to know whether this is linked to a biological limit [16] or to other causes.

Thanks are due to Jean-Marie Robine, INSERM, Montpellier, France, and to referees for their helpful comments on previous versions of the article.

The author has no conflict of interest to declare.

1.
Vallin
J
,
Meslé
F
.
Tables de mortalité françaises pour les XIXe et XXe siècles et projections pour le XXIe siècle.
Éditions de l’Institut national d’études démographiques, Paris, France,
2001
.
2.
Vaupel
JW
. The average French baby may live 95 or 100 years. In:
Robine
JM
,
Vaupel
JW
,
Jeune
B
,
Allard
M
, editors
.
Longevity: to the limits and beyond
.
Berlin
:
Springer
;
1997
. pp.
11
27
.
3.
Blum
A
,
Monnier
A
.
Recent Mortality Trends in the U.S.S.R.: new Evidence
.
Pop Stud
.
1989
;
43
(
2
):
211
41
.
4.
Case
A
,
Deaton
A
.
Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century
.
Proc Natl Acad Sci USA
.
2015
Dec
;
112
(
49
):
15078
83
.
[PubMed]
0027-8424
5.
Woolf
SH
,
Chapman
DA
,
Buchanich
JM
,
Bobby
KJ
,
Zimmerman
EB
,
Blackburn
SM
.
Changes in midlife death rates across racial and ethnic groups in the United States: systematic analysis of vital statistics
.
BMJ
.
2018
Aug
;
362
:
k3096
.
[PubMed]
0959-8138
6.
Hiam
L
,
Harrison
D
,
McKee
M
,
Dorling
D
.
Why is life expectancy in England and Wales ‘stalling’?
J Epidemiol Community Health
.
2018
May
;
72
(
5
):
404
8
.
[PubMed]
0143-005X
7.
Masters
RK
,
Tilstra
AM
,
Simon
DH
.
Explaining recent mortality trends among younger and middle-aged White Americans
.
Int J Epidemiol
.
2018
Feb
;
47
(
1
):
81
8
.
[PubMed]
0300-5771
8.
Watkins
J
,
Wulaningsih
W
,
Da Zhou
C
,
Marshall
DC
,
Sylianteng
GD
,
Dela Rosa
PG
, et al
Effects of health and social care spending constraints on mortality in England: a time trend analysis
.
BMJ Open
.
2017
Nov
;
7
(
11
):
e017722
.
[PubMed]
2044-6055
9.
Jougla
E
,
Hemon
D
. Consequences of the 2003 summer heat wave on mortality levels in France. In:
Carey
JR
,
Robine
JM
,
Michel Christen
Y
, editors
.
Longevity and frailty. Berlin
.
Berlin
:
Springer
;
2005
. pp.
125
38
.
10.
Campèse
C
,
Bonmarin
I
,
Savitch
Y
,
Fouillet
A
,
Sommen
C
,
Lévy-Bruhl
D
, et al
Surveillance de la grippe en France, saison 2016-2017
.
Bull Epidemiol Hebdo.
2017
;
22
:
466
75
.
11.
Oeppen
J
,
Vaupel
JW
.
Demography. Broken limits to life expectancy
.
Science
.
2002
May
;
296
(
5570
):
1029
31
.
[PubMed]
0036-8075
12.
Christensen
K
,
Doblhammer
G
,
Rau
R
,
Vaupel
JW
.
Ageing populations: the challenges ahead
.
Lancet
.
2009
Oct
;
374
(
9696
):
1196
208
.
[PubMed]
0140-6736
13.
Prioux
F
,
Barbieri
M
: L’évolution démographique récente en France : une mortalité relativement faible aux grands âges. Population-F
2012
67:597-656.
14.
Robine
JM
.
When climate change encounters the revolution in adult longevity
.
Aging Clin Exp Res
.
2017
Dec
;
29
(
6
):
1073
4
.
[PubMed]
1594-0667
15.
Gavrilov
LA
,
Gavrilova
NS
,
Nosov
VN
.
Human life span stopped increasing: why?
Gerontology
.
1983
;
29
(
3
):
176
80
.
[PubMed]
0304-324X
16.
Marck
A
,
Antero
J
,
Berthelot
G
,
Saulière
G
,
Jancovici
JM
,
Masson-Delmotte
V
, et al
Are we reaching the limits of Homo sapiens?
Front Physiol
.
2017
Oct
;
8
:
812
.
[PubMed]
1664-042X
17.
Papon
S
,
Beaumel
C.
Bilan démographique 2018. La fécondité baisse depuis quatre ans. Insee Première n°
1730
.
18.
Holm
S
.
A simple sequentially rejective multiple test procedure
.
Scand J Stat
.
1979
;
6
:
65
70
.0303-6898
Copyright / Drug Dosage / Disclaimer
Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.