Following their work, Lupton and her collaborators reached the conclusion that each extra year of work can delay the onset of Alzheimer’s by 0.13 of a year, that is to say, around a month and a half. However, it seemed to Catherine Grotz and her co-authors of the article published in Plos One that, while interesting in itself, the article of Lutpon and colleague was open to criticism from a methodological point of view. The first point to make is that Lupton’s sample (which included 382 individuals) selected from a database containing 938 individuals did not include any women. Not only did the exclusion of women reduce the size of the sample but it also impeded generalization of the results.

A second limitation to be found in Lupton’s paper is the fact that it is difficult to determine the parameter upon which the appearance of Alzheimer’s is based: the age when the first symptoms manifest themselves (a subjective, self-reported parameter) or a more objective and more reliable parameter: the age at diagnosis.

Two major biases

On an even more fundamental level, the study directed by the researcher from King's College London is affected by two biases. The first concerns the selection of the sample. In fact, only individuals diagnosed with Alzheimer’s disease are taken into consideration to the exclusion of the normal aged population.

In addition, in order to be sure that it is indeed retirement that has an impact on the time of onset of the disease and not the opposite – some people might have stopped working due to previously detected cognitive problems (reverse causality) – Lupton and her collaborators excluded from their sample all individuals that had been diagnosed with Alzheimer’s before their retirement. This is not sufficient to take account of the reverse causality, all the more so because it has already been shown that the prodromal phase of the disease can begin up to 10 years before the first diagnosis of dementia is made (3).

“Given such a selection strategy, the British researchers created an overestimation of the effect of retirement on Alzheimer’s disease”, comments Catherine Grotz. “In fact, if we apply the two conditions they defined (inclusion of Alzheimer’s patients who were diagnosed after beginning their retirement), a worker who left the working environment at 60 years of age, for example, can only have developed the disease after the age of 60, while another who continued his job until the age of 65 or 70 can only develop the disease after this age. In other words, those who leave work at a later age are necessarily those who are most likely to be diagnosed with Alzheimer’s at an older age”.

Lupton revisited

In the light of these factors, Catherine Grotz and her collaborators set themselves the objective of studying the association between retirement age and age at onset of AD (i.e., onset of first symptoms and diagnosis of AD). They did so by taking into consideration the two biases which could explain the results presented by Lupton in favor of the theory of cognitive reserve.

(2) Lupton MK, Stahl D, Archer N, Foy C, Poppe M, et al. (2010) Education, occupation and retirement age effects on the age of onset of Alzheimer's disease. Int J Geriatr Psychiatry 25: 30-36.
(3) Amieva H, Le Goff M, Millet X, Orgogozo JM, Peres K, et al. (2008) Prodromic Alzheimer's disease: Successive emergence of the clinical symptoms. Ann Neurol 64: 492-498.