In Jonathan Swift’s eighteenth-century masterpiece, Lemuel Gulliver travels to the Kingdom of Luggnagg, where every so often a child is born with a mark over its left eye indicating that it is a Struldbrug, destined to live forever. Gulliver eagerly anticipates a meeting with some of the oldest Struldbrugs, hoping to learn much from their accumulated wisdom. To his dismay, he discovers that, though they live forever, their faculties fail with advancing age and their knowledge reflects only what they had learned in youth or middle age. If we lived long enough, would we all become demented? That nontrivial question (there will be 100,000 Americans over the age of 100 by the year ZOOO!) is now addressed in two papers published in Neurology, one in February1 and one this month. 2 In both, the authors have examined the prevalence of dementia and Alzheimer’s disease in the very old. Wernicke and Reischiesl examined 78 people over age 85 (26 of them over 95) living in Berlin, and Ebly et a12 evaluated 1,834 over age 85 (104 of them over 95) living in Canada. In both papers, the authors note that, in epidemiologic~ tudies,~,~ the prevalence of dementia has consistently been found to double approximately every 5 years between the ages of 65 and eighty-five. Both papers raise the question of whether the prevalence of dementia continues to increase in the very old and whether a leveling off (as has been suggested5) would indicate a different etiology of dementia in the nonagenarians and centenarians than in younger age groups. The two studies arrive at quite opposite conclusions. Wernicke and Reischiesl found that, in their 26 Berliners over 95, the prevalence of dementia was the same as for the twenty-six 90-to 95-year olds (42.3%) and that, except for four patients with vascular dementia, all others presumably had AD. They conclude that “the group aged 95 years and older... showed a plateau near 45%, with no further increase in dementia prevalence.” They speculate that Alzheimer’s disease “may be genetically transmitted and occur in no more than approximately 50% of the population.” By contrast, Ebly et a12 found that the prevalence of dementia continued to rise, from 40.4% for the 371 Canadians 90 to 94 years old to 58.6% for the 104 people 95 and older, and the proportion of AD cases was slightly but not significantly higher over age 85. They conclude that “dementia and, in particular, AD may be processes that occur to a varying extent in all individuals, and may become universal when age is sufficiently advanced.” How can these very different conclusions be reconciled? In the so-called “dismal science” of epidemiology, things are not always what they seem. The real question here is not the percentage of people over 95 who are demented at a point in time, but rather this: what would be the true risk of developing dementia if everyone in the population lived to the maximum human life span potential of about 120 years? Would everyone eventually become demented, or would only some of the population (those with the genetic flaw) develop dementia during a limited age window, while the others (“escapees,” as with Huntington’s disease) would live on, free of this risk?