The scent of death

PEOPLE whose hearts are failing, or who have had cancer or lung disease, have good reason to be concerned about their future. People who have merely lost their sense of smell might not be so worried. Actually, though, their prospects are worse. You are more likely to die within five years if you cannot recognise common smells than if you have ever been diagnosed with one of those more obviously deadly illnesses. That, at least, is the conclusion of a sobering study just published in PLOS ONE, by Martha McClintock and Jayant Pinto of the University of Chicago.

Dr McClintock and Dr Pinto were prompted to conduct their investigation because they knew olfactory problems can forewarn of neurodegenerative diseases such as Alzheimer’s and Parkinson’s. They are also associated with abnormally shortened telomeres (the caps on the ends of chromosomes), and that shortening is, in turn, implicated in the process of ageing. Moreover, a good sense of smell helps keep people healthy by detecting pathogens and toxins in the air, stimulating appetite, and aiding memory, emotions and intimacy. The researchers therefore had good reason to wonder if measuring smell loss might predict mortality.

As part of the National Social Life, Health and Ageing Project, a survey organised by the University of Chicago that measures the health and well-being of older Americans, 3,005 participants aged between 57 and 85 completed a three-minute smell test devised by Drs McClintock and Pinto. The survey’s researchers prepared special felt-tipped pens scented with five common odours—fish, leather, orange, peppermint and rose—and presented them one by one to volunteers. After each presentation, the volunteer was shown pictures and names of four possible answers, and was asked to select the correct one. Getting one answer wrong was considered okay, or “normosmic”, but two or three errors labelled a person as “hyposmic”, or smell-deficient, and four or five counted them as “anosmic”, or unable to smell.

Five years later, 430 of the respondents were dead. The anosmic were particularly hard hit: 39% of them had perished, compared with 19% of the hyposmic and just 10% of the normosmic. After adjusting for age (older people are more likely to die), sex (women tend to live longer than men), socioeconomic status (richer, better-educated people enjoy greater longevity) and race (some ethnicities have longer lifespans than others), Dr McClintock and Dr Pinto found that the correlation still held true. Nutrition could not explain it. Nor could smoking or alcohol use. Even when they controlled for specific diseases, like heart failure, cancer, strokes and diabetes, the researchers found the relationship between not being able to smell and being more likely to die in the next half-decade remained robust. Only severe liver damage was a stronger predictor of death.

No one is suggesting that not being able to smell led directly to any of the deaths. Rather, the researchers think that smell may be the “canary in the coal mine of human health”. They note that olfaction relies on a turnover of stem cells (from which other sorts of cell develop) to maintain its functioning. Not being able to smell, they speculate, may signal a more general inability to regenerate and renew.

One weakness of their work, they concede, is that they do not know the causes of the deaths they recorded—as that was not asked in the original survey. Another is that the study was but a snapshot. It cannot show whether those at risk have always had poor smell, or have developed it recently. Nevertheless, Dr McClintock and Dr Pinto do seem to have found the makings of a simple test that could sniff out danger ahead, and thus possibly avert it.