Validation of Exceptional Longevity

Species of Evidence of Exceptional Longevity

by B. Jeune and J.W. Vaupel


[ References ]

Most reported instances of exceptional longevity are incorrect. This was the rule everywhere until the end of the 19th century and is still the case in the vast majority of countries. Previous monographs in this series have documented the proliferation of centenarians since 1950 - in Japan and the countries of Western Europe, the only countries with time series of valid data at advanced ages (Kannisto 1994; Jeune and Vaupel 1995; Kannisto 1996; Thatcher, Kannisto, and Vaupel 1998). Although today age reports in the Nordic countries are exceptionally accurate, Lundström (1995) shows for Sweden and Skytthe and Jeune (1995 and in this volume) for Denmark that substantial age exaggeration occured up through the late 19th century. Until then in Sweden and Denmark - and still today in most countries - age statistics were based on self-reported ages in censuses or unverified ages reported on death certificates. It is only when reliable birth registrations are available for a century or more and when reports of ages above 100 are systematically checked against these data that the quality of national statistics on exceptional longevity improves.

The cult of centenarians in the 17th and 18th centuries

The history of longevity is a history of myths. Jeune (1995) describes the widespread fascination in the 17th and 18th centuries with longlivers, whose extreme ages were uncritically repeated in numerous catalogues. In this volume Laslett elaborates on this fascination, which he calls the cult of centenarians. Leading scholars of the time, such as Bacon, Locke, Harvey, Haller, and Temple, fought prejudice, demanded empirical evidence, and questioned unverified claims. They were, however, astonishingly gullible about exceptional longevity. As Laslett emphasizes, there is no indication that any of these eminent sceptics ever doubted self-reports of ages well above 100.

Laslett opens his chapter with Locke's narrative about Alice George. Locke did not question her reported age of 108 years even though he was told that she did not marry until she was 30 years old and that she subsequently had 15 children. He did not even interview her eldest son, who lived "next door to her" and who was 77 years old. Alice George may have had her first child when she was about 20 years old, if she really had 15 children. She may have been a nonagenarian, an exceptional event at that time, if the man next door really was her son and actually was 77 years old.

Laslett seeks to explain the emergence and spread of this cult, which left its mark on the spirit of the time. Laslett's intriguing and sharply articulated hypotheses about the roots and burgeoning of the cult deserve attention and should lead to further studies. Laslett is not certain about the extent of the cult and he emphasizes that "we must not exaggerate its effects or its substantiveness ... Its origins are rather conjectural and it is not clear how widespread it was at any one time, how many versions of it may have existed, who accepted any of them, and who, like Gregory King, seem to have been unaffected".

Petersen and Jeune's chapter on Bolle Luxdorph's private library of literature on long-lived people gives an interesting glimpse into the "cult of centenarians" at that time. Luxdorph was a high-ranking Danish public servant who was acquainted with the naïve studies by Bacon and Haller of longevity and who devoted great effort to collecting portraits of the very old. As Kjærgaard (1995) stressed, he impressed on the bishops whom he had asked to report on centenarians that he wanted documentation in the form of birth and death certificates. In a case where he had the possibility to check reported information he was able to prove that an alleged centenarian had only reached the age of 93. He may have been the first to successfully carry out a validation of the age of a reported centenarian.

Laslett rightly highlights William Thoms (1803-1885) "as the protagonist, in the proper sense of that word, of the story of the emergence of the validation of centenarian claims". Thoms in his book on "Human Longevity" (1873) acknowledges Dilke and Lewis as British predecessors. As shown by Poulain et al. in this volume, the Belgian, Quetelet, may also be regarded as a predecessor, and Desjardins in this volume reports that the Canadians Taché and Tanguy validated the age of alleged centenarians in Canada at the time of Thoms but independently of his work. Nonetheless Thoms must be recognized as the first who systematically dealt with the problems of age-validation. Also, he was the first who drew up unambiguous criteria for age-validation, criteria that are still useful and reasonable, as shown in several chapters in this volume. At the end of this introductory chapter we quote Thoms' criteria, which he called "species of evidence".

Historic-demographic studies of centenarians

In the second monograph in this series, edited by Jeune and Vaupel (1995), several authors examine the hypothesis that the emergence of centenarians is a modern phenomenon, with few if any genuine long-livers reaching age 100 before about 1800. This hypothesis is examined further in this volume.

Desjardins shows how meticulous age-validation conducted by French-Canadian demographers has refuted almost all of the few centenarians whom Taché and Tanguy did not reject. Of special interest is Pierre Joubert, who until recently appeared in the Guinness Book of Records as a 113 year old man who died in 1814. In reality he died in 1766, 67 years old, whereas his namesake - his son - died in 1814. It is striking that not a single of the 178 reported centenarians among the 210,000 registered deaths among French-Canadians before 1800 could be verified, and that only one case (Jean Feuville) could not be refuted. The refuted centenarians were found to have died at an average age of 88 years and 14 died in their 70s.

Perhaps the most interesting finding of the French-Canadian study is that among 20,000 persons who were born in Canada before 1700 there were 105 who appear to have died as nonagenarians, 12 of these reaching between 95 and 99 years. It must be emphasized, though, that the age-validation of these nonagenarians was not at the same meticulous level as for the alleged centenarians. The oldest of the nonagenarians, Elisabeth Chavigny, who died a month after her 99th birthday, was however validated with a thorough family reconstitution which seems to rule out name-saking. She is thus the oldest person before 1800 to have been age-validated so thoroughly. Evidence is weak for the early centenarians in Great Britain reported by Hynes (1995) and in China reported by Zhao (1995). There is stronger evidence for another centenarian before 1800, the Norwegian Eilif Philipsen (Kjærgaard 1995), but a thorough family reconstitution is not available.

If it was possible for one person among the 20,000 born in French Canada in the 17th century to reach the age of 99, it may seem likely that it was possible for someone somewhere in the world to attain his or her 100th birthday. Elisabeth Chavigny may, however, have been an extraordinary outlier, comparable to Jeanne Calment, whose verified lifespan of more than 122 years is almost five years greater than the next-longest verified lifespan. It would therefore be interesting if Desjardins in the same thorough fashion could validate the 11 other 95-97 year olds.

Early historic materials (cf. Buffon 1749) and more recent demographic data from various well-defined populations in the 17th and 18th centuries (cf. Bourdelais 1993, Wrigley et al. 1997) may throw more light on the question of whether anyone reached age 100 before 1800. It can, however, already be concluded that at that time it was possible to become a nonagenarian, a conclusion that Thatcher (1999) recently has given statistical support.

Skytthe and Jeune (1995) show that the number of reported Danish centenarians fell sharply after 1840, i.e. from the time when the central statistical bureau started to systematically check census and death-certificate reports. It might be presumed that reported centenarians are valid after about 1840. But Skytthe and Jeune in this volume demonstrate that age-exaggeration persisted until the end of the 19th century. They were able to validate the true lifespan of less than half of the centenarians reported in the period 1840-1899. Of these validated cases, only about half genuinely attained the age of 100. The quality of the age reports improved from 1840 to 1899 but there were cases of age exaggeration even at the very end of the 19th century. Niels Thorsen provides a good example of how age-exaggeration took place successively at every census. This successive age-exaggeration, in the end summing up to 20 years, is comparable with that documented by Mazess and Forman (1979) for reported centenarians in Vilcamba in Ecuador.

The extensive Belgian material presented by Poulain et al. in this volume confirms the findings from Denmark. There is evidence of widespread age-exaggeration in the period 1586-1799. After 1800 there are clear signs of considerable age-exaggeration up until about 1890. This is clearly illustrated by an examination of the 1846 census. Quetelet started a validation of the 31 reported centenarians and could confirm only 21. The further validation by Poulain et al. of these 21 shows that the validation of Quetelet was not sufficiently rigorous, as they could confirm only 6 persons, whereas they could refute 6 and 9 remain doubtful. The Belgian material is highly interesting as sufficient data on each of the centenarians have been collected to make comprehensive family reconstitutions possible. This makes it possible to analyse the occurrence of families with long-living siblings.

In sum, several chapters of this volume provide further evidence that genuine centenarians before 1800 were nonexistent or at least extremely rare. The data materials mentioned above and other demographic data make it possible to further explore the prevalence of nonagenarians before 1800 and the emergence of centenarians in the last century. It will also be interesting to identify the time of the appearance of the first supercentenarian (i.e., a person with a lifespan of more than 110 years) as well as the first 105 year olds.

Studies in the maximal duration of life

Sufficiently thorough attempts at verifying the maximal duration of life in different periods of time have not been made. Nonetheless, various gerontologists continue to make the undocumented claim that the maximal duration of life is 110, 115 or 120 years, and that this maximal duration has not changed over the course of human history. Unreliable reports of extreme ages in various countries and at various times, including the present, continue to be cited despite their implausibility and the lack of firm evidence.

This monograph carries on the laborious work of validating the ages of the few persons who may have been those who have lived longest. We do not know when the age of 110 years was reached for the first time, but statistical considerations indicate that the event may not have taken place until this century, and perhaps not until after 1950 as suggested by Jeune (1995) on the basis of the work of Vincent (1951).

The Irish noblewoman, Kathrine Plunket, described by Thatcher in this volume may have been the first supercentenarian. She allegedly died at the age of almost 112 years in 1932. Thatcher has been able to procure much information that makes the claim very plausible. The family reconstitution presented is not as compelling as the family reconstitutions presented in this volume for supercentenarians who could be investigated while they were still alive. Kathrine Plunket is, however, the best bid so far for an age-validated supercentenarian from before 1950. Two earlier alleged supercentenarians, the French-Canadian Pierre Joubert and the Dane, Bartholomaeus Albrecht, are refuted in this volume by Desjardins and by Skytthe and Jeune.

While the research in this monograph was in progress, three exceptional long-livers died. These three were the long-livers whom the contributors to this volume scrutinized more closely than any alleged long-liver has ever been examined before. The three are Jeanne Calment, who died on August 4, 1997 at the age of 122 years and 5 months; Marie Louise Meilleur, who died on April 16, 1998 at the age of 117 years and 7 months; and Christian Mortensen, who died April 25, 1998 at the age of 115 years and 8 months.

The most thorough age-validation ever to have been undertaken was done for Jeanne Calment by Robine and Allard (1995) with the assistance of Caroline Boyer. Their family reconstitution is so extensive that it will be very difficult to replicate it on any other longliver at any other place. It was possible only because Jeanne Calment always lived in her native town, Arles, and because her family has always lived in or in the neighbourhood of Arles. This made it possible to find ancestors in five generations before her, and a great number of descendants of these ancestors (Robine and Allard 1998). Jeanne Calment is an extraordinary outlier because she lived more than 122 years.

The perhaps second longest living person, Marie Louise Meilleur, was almost five years younger than Jeanne Calment when she died. As mentioned by Desjardins in this volume, quite a number of persons were nominated by the press to be Jeanne Calment's successor after her death. For most of these persons a validation has not even been attempted. Although Marie Louise Meilleur has not yet been as thoroughly validated as Jeanne Calment and Christian Mortensen, she is the best known candidate for an age-validated 117-year-old person. However, it may be possible that Sarah Knauss from Allentown in Pennsylvania, who was born September 24, 1880, i.e. one month later than Marie Louise Meilleur (August 29, 1880), is the second longest living person, as she was still well, living in a nursing home at the age of 118 when some of the authors in this monograph (Perls, Robine, Wilmoth, and Jeune) visited her November 22, 1998. The few documents that we saw at the visit seem to support her reported age. Her age will now be thoroughly validated by Perls and his colleagues.

The longest living person in Europe after Jeanne Calment is Charlotte Hughes, who died in 1993 at the age of 115 years and 229 days, and currently the oldest reported person in Japan is Aso Takii, who died in August 1998 at the age of 114 years. It may be possible to find a few additional women from the most developed countries who have lived beyond age 115. Furthermore, the quality of age-reporting in China, as documented by Zeng et al. in this volume, suggests that it may be possible to verify some Chinese supercentenarians.

Considerably fewer men than women reach advanced ages. Among Danish centenarians three out of four are women, and among those who are 105 years or older nine out of ten are women. It is no wonder, then, that there are very few credible reports of men over age 110 in countries with valid data. When an American born in Denmark, Christian Mortensen, was mentioned in a Danish newspaper on his 112th birthday, the authors of the chapter in this volume on Mortensen read this account with considerable scepticism, not least because age-exaggeration is common in reports from the United States. A Danish genealogist, Poul Ørberg, showed an interest in Mortensen when he attained the age of 110 and had found the record of his birth in the local parish register. This, however, might have been just another instance of namesaking. So the authors of the chapter on Mortensen started a thorough age-validation that included double-checking of information from interviews with register information (cf. Wilmoth et al. 1996). In this volume this research is supplemented with additional information. The validation of Mortensen's age has been meticulous, although not up to the standards of the validation of Jeanne Calment. It was much more difficult to validate Mortensen because he was an emigrant and because he lived several places in the Unites States.

Age validation in current studies of centenarians

In Denmark and Belgium contemporary centenarians can be readily identified and their ages verified because these countries have electronic systems of civil registration built with valid documentation. In some other European countries, such as France and Italy, identification and verification of centenarians are more problematic. In these countries centenarians must be identified through municipal electoral registers. Such registers are not always kept up-to-date and are not always cleansed of everyone who has died.

However, most European countries, including France and Italy, have a long tradition of registration of births, marriages, and deaths in parish and municipal registers. It is therefore possible to carry out thorough age validation studies - as illustrated by the research on Jeanne Calment. To do so, however, the person to be validated must have been born and have lived most of his or her life in the same municipality. Alternatively, reports of deaths have to be communicated to the municipality of birth, as has been the practice in France since 1945. In the chapter of Cournil et al. it is shown how difficult it may be to estimate the correct lifespan of persons who were born at the end of the century in Jeanne Calment's native town, Arles, because of unrecorded deaths on a considerable proportion, especially of those who died before 1945.

In countries with no official birth registration until this century, age validation is even more difficult. As shown in this monograph, this is the case in China and the United States. Wang et al. in their chapter on age validation of centenarians in China demonstrate that considerable overstatement of extreme ages is found among some ethnic minorities. Age reports for Han Chinese up to the age of about 105 appear generally to be valid, on the basis of some of the demographic indices described by Kannisto later in this monograph and on the basis of a survey carried out among approximately 200 centenarians from three different regions in China. Given the lack of birth registration for the oldest Chinese and the loss of family genealogies during the Cultural Revolution, Wang et al. had to develop alternative methods of age validation.

In their chapter based on the New England Centenarian Study, Perls et al. show how it was possible to make reasonable estimates on the prevalence of centenarians in a local area of eight suburbs around Boston. Centenarians were identified from municipal electoral registers and other sources. It turned out that most of these reported centenarians were either dead or died before the date chosen for the estimation of prevalence. A few others had moved out of the area and some could not be located. Interviews were then conducted with the remaining centenarians or members of their families to procure copies of birth certificates and other corroborative pieces of evidence indicating the person's age, such as military certificates, old passports, school report cards, family bibles, and baptismal or other church documents. In many cases it was clear that the reported centenarian was actually younger. For only 46 persons out of an initial list of 289 could it be concluded that the person was living in the eight-town area and was probably a centenarian. These 46 centenarians indicate a prevalence of about 1 centenarian per 10,000 people, a prevalence a little higher than that found in Western European countries with reliable data

In a study of centenarian siblings in a family from Ohio, Alpert et al. show that it is possible in the United States to collect very detailed information about long-lived families, provided that nearly all members of the family have lived in the same area. This study is interesting in that it is the first published example of a complete age validation of the occurrence of centenarians in a family. Similar research is now being conducted in Belgium and Denmark. This body of research was discussed at a workshop, in which most of the authors of this monograph participated, on "Genes, genealogy and longevity" at Université Catholique in Louvain-la-Neuve on October 22-24, 1998. Long-lived families will be the focus of a future monograph in this series.

The main theme and conclusion of the chapters in the present monograph is that information on centenarians has to be treated with great care. Age misreports have been so common that it is judicious to be highly sceptical about any report of extreme age. In most countries and at all times up until the 20th century, most reports of centenarians have been false and demographic data on numbers of centenarians and on survival at the highest ages are incorrect. In the important final chapter of this monograph Kannisto summarizes and applies various indices that can be used to assess the validity of age-specific demographic data. His considered judgment is that very few countries have reliable data on centenarians before 1950. Furthermore, Kannisto has serious doubts about the quality of current data on centenarians in such major, highly-developed countries as the United States, Canada, and Australia. For the countries and time-periods Kannisto thinks usable data are available, age-specific death counts and population counts are available as part of the Kannisto-Thatcher database at Odense University Medical School.

Age validation criteria

As noted above Thoms was the first to propose criteria for age validation. He called these criteria "species of evidence." To honor the pioneer work of Thoms in this field - and because this pioneering work is still useful and informative - we conclude this introductory chapter with a lengthy quote from Thoms (1873):

[ Home | Contents | Return to previous page ]

Max-Planck-Gesellschaft 2003