Part 2 · Course and diagnosis
Stages and symptoms
Parkinson's changes slowly over many years and at a different pace for each person. The Hoehn and Yahr scale sorts that course into five stages, driven mostly by which sides are affected and by balance. Scrub through them, then switch to the symptoms the scale does not count.
Schematic figure. Stage wording is the 1967 scale as commonly reproduced; the original paper could not be checked (unverified). The bar places one cohort's median months between stages end to end (Zhao 2010, Singapore, modified scale with stage 2.5); a sum of medians is not a median time to stage 5.
What the stage does not tell you
A scale weighted to balance
An MDS task force noted that Hoehn and Yahr is weighted heavily toward postural instability, mixes impairment with disability, is not linear and says nothing about non-motor problems. Its reliability and validity are best supported for the middle stages, 2 to 4. 127
Stages 1.5 and 2.5
The modified scale with half stages is widely used but had no clinimetric data. The task force recommended keeping the original five stages until the modified one is tested, reporting groups as medians and ranges, and "rate what you see", including other illnesses. 127
Stage explains little of quality of life
In a six-country survey, Hoehn and Yahr stage and medication explained only 17.3% of the variation in quality of life. Adding other factors raised this to 59.7%, and depression was the strongest predictor. 133 Non-motor symptoms as a whole affected quality of life more than motor symptoms did. 215
The pace differs
Older age at diagnosis, longer disease duration and worse motor scores predicted faster moves between stages; sex and ethnicity did not. 386 In DATATOP, people with onset at 40 or younger took longer to reach the same disability than those with onset at 70 or older (2.9 against 1.7 years), and fast progressors more often had slowness and gait problems at onset. 161
The long course: two cohorts followed from diagnosis
Per 100 people. Figures change at an even pace for illustration; only the end-point figures are data.
Survival: why there is no single number
A standardised mortality ratio (SMR) compares deaths with those expected in the general population of the same age and sex: 1 means no difference. Published ratios differ widely, from 0.9 to 3.8 across 88 studies, and they rise the longer people are followed. 205 The studies behind this guide gave no single citable figure for median survival from diagnosis or for years of life lost, so none is shown here.
Dots are estimates; the line is a 95% confidence interval where one was reported; the shaded band is the range across studies.
55% · 74%
had died 10 years after diagnosis in CamPaIGN and 20 years after in the Sydney study. 142,378 In post-mortem studies, mean disease duration until death ranged from 6.9 to 14.3 years, and survival fell by about 5% for each year of follow-up on average. 205
Age and dementia
Older age and dementia were the factors most often linked to higher mortality. There was no robust evidence that mortality fell after levodopa was introduced. 205 In Sydney, the death rate fell in the first 3 years of treatment and then rose relative to the general population. 142
Pneumonia
was the commonest cause of death in CamPaIGN. Death certificates named Parkinson's as a substantial contributor in only 20% of deaths, so it is under-recorded. 378 Planning ahead and palliative care are covered in Living with Parkinson's.
The 1967 cohort
Hoehn and Yahr's paper followed 672 people with "primary parkinsonism". Secondary summaries report that 61% were severely disabled or dead after 5 to 9 years, more than 80% of those followed beyond 10 years, and mortality about 2.9 times that of the general population. The original was not accessed, so these figures are unverified. 146