For the study Devanand and colleagues studied 150 patients with minimal to mild cognitive impairment. They compared them to 63 healthy elderly people and ran tests on them every 6 months.
The inability to identify 10 specific odors clearly predicted who would go on to develop Alzheimer's, they told a meeting of the American College of Neuropsychopharmacology.
These smells included strawberry, smoke, soap, menthol, clove, pineapple, natural gas, lilac, lemon and leather.
'Narrowing the list of odors can potentially expedite screening and help with early diagnosis,' Devanand said in a statement.
This makes sense, he added, because examination of the brains of Alzheimer's patients shows that the nerve pathways involved in smell are affected at a very early stage.
Several groups have tried to link the sense of smell with Alzheimer's and at least one company markets a scratch and sniff test for the disease. But Devanand said it is important to identify the specific odors that may be involved.
I'd be amazed if this were a specific test (If you can't smell them --> Alzheimer's). I suspect it's more likely to be a "sensitive" test (Can smell? --> Not Alzeheimer's -- maybe another dementing disorder?). Really way too early to tell.
Not that that's going to prevent a lot of sniffing by every 40 yo with middle-aged memories.