Young-Adult Writing Ability -- the Nun Study
In February 1996, University of Kentucky researchers made headlines with a controversial study suggesting that analysis of people's writing ability during their 20s could predict with astonishing accuracy their development of Alzheimer's disease 60 years later.
The study involved 93 nuns who joined the Sisters of Notre Dame in the 1920s. As part of their preparation, they each wrote autobiographies, which the Order archived. Sixty years later, by the 1980s, almost one-third of the nuns -- by then in their 80s -- had been diagnosed with Alzheimer's disease, an incidence similar to that found in the general population. Fourteen had died, and the researchers performed brain autopsies to look for the disease's characteristic abnormalities (neurofibrillary tangles and senile plaques). Five had Alzheimer's abnormalities.
The researchers then went back to the archived autobiographies of the deceased nuns to see if they might offer clues to later development of Alzheimer's. Without knowing who had written the essays, the researchers evaluated them for two criteria of language ability: idea density and grammatical complexity. The scientists judged nine of the autobiographies to be intellectually rich, packed with ideas, and grammatically complex. They considered the other five to be idea-poor and grammatically simple. Finally, they linked each of the nuns to her autobiography.
None of the nine whose writing was idea-dense and grammatically complex had developed Alzheimer's. But all of those whose work was idea-poor and grammatically simple had succumbed to the disease. The researchers suggested that young-adult writing samples might predict later development Alzheimer's with 90% accuracy. They wrote: "Our findings indicate that low linguistic ability early in life is a potent marker ... of Alzheimer's disease risk late in life."
But after the excitement died down, other researchers raised serious doubts about this study. The first objection concerned the evaluative criteria. Grammar can be evaluated objectively using standardized tests of sentence structure. It turned out that grammatical simplicity was related to poor cognition among surviving nuns, but not to development of Alzheimer's disease. Meanwhile, idea density is rather subjective, yet the researchers' conclusions were based largely on that criterion.
The second objection concerned the sample size -- just five nuns out of 93 with autopsy-confirmed Alzheimer's disease. Critics said that number of cases is not enough to conclude anything with confidence.
It remains possible that language ability during young adulthood may have some value in predicting and diagnosing Alzheimer's later in life. But today, the case is far from proven.