Smell loss predicts Parkinson's

The research record

This is the project's file for this landmark, as written and checked. Sources are listed at the end; anything not yet confirmed is marked "Not yet verified".

Who

M. M. Ponsen, H. W. Berendse and colleagues (Ann Neurol 2004; VU University Medical Center, Amsterdam). G. W. Ross, H. Petrovitch, R. D. Abbott and colleagues, Honolulu-Asia Aging Study (Ann Neurol 2008).

What happened

• Ponsen 2004. 361 symptom-free first-degree relatives of people with Parkinson's were tested for smell. Forty with poor smell and 38 with normal smell were scanned for dopamine transporters (SPECT). After 2 years, 10% of the poor-smell relatives, all with strongly reduced scan binding at baseline, had developed Parkinson's; none of the others had. A 2010 follow-up reported 5 of 40 (12.5%) by 5 years.
• Ross 2008. 2,267 men aged 71 to 95, free of Parkinson's and dementia, took an odour identification test in 1991–96. In up to 8 years, 35 developed Parkinson's. In the first 4 years, incidence per 10,000 person-years fell from 54.5 in the lowest smell quartile to 26.6, 8.2 and 8.4 in the others. The adjusted odds ratio for the lowest quartile was 5.2 (95% confidence interval 1.5 to 25.6). The link was not seen beyond 4 years.

Why it moved us toward a cure

Smell centres are among the earliest sites of pathology in the Braak model (landmark 6); these studies gave that idea a test in living people. Smell loss became one of the markers in the MDS prodromal criteria (landmark 222), which pick people for prevention trials.

Outcome

Advanced as a marker, but a weak one on its own. In Ponsen's cohort most poor-smell relatives did not get Parkinson's (5 of 40 by 5 years), and Ross found the signal faded after 4 years. Ponsen's authors said the 2-step approach (smell test, then dopamine-transporter scan) needs confirmation in larger groups.

Sources

• Ponsen et al., Ann Neurol 2004, DOI 10.1002/ana.20160, PMID 15293269
• Ponsen et al., J Neurol Neurosurg Psychiatry 2010 (5-year follow-up), DOI 10.1136/jnnp.2009.183715
• Ross et al., Ann Neurol 2008, DOI 10.1002/ana.21291, PMID 18067173
• Not yet verified: the full texts (abstracts only, via Europe PMC); how much smell loss weighs in the 2019 prodromal criteria update beyond what landmark 222 says; other early smell studies.