Incidence of Parkinson's Disease
Clinical bottom line
Most developed countries with northern European age structures tended to have incidence rates between 12 and 20 cases per 100,000 per year.
Reference
D Twelves et al. Systematic review of incidence studies in Parkinson's Disease. Movement Disorders 2003 18: 19-31.
Systematic review
The review used MEDLINE and EMBASE, with searching to the end of 2001. Articles for inclusion had to be full publications of original studies providing an incidence rate for a whole population for idiopathic Parkinson's disease, or for Parkinsonian symptoms including Parkinson's disease.
Results
Twenty-five incidence studies were found, looking at populations throughout the world, and with populations as low as 8,000 to just under four million, though the number of incident cases was seven and about 400 (Figure 1).
Figure 1: Incidence of Parkinson's disease in individual studies
The determination of the diagnosis was mixed, with some patients seen by a specialist, varying from 30% to 100% of cases. Definition of incidence was also mixed, some studies using date of diagnosis, others date of onset, some both, and some did not give a definition. Also mixed in the studies were the diagnostic criteria used for defining Parkinson's disease.
In most studies the peak incidence was between 70 and 79 years of age, though mean age of symptom onset was 60-65 years in eight studies and over 65 years in five studies. There tended to be somewhat more men than women diagnosed (rate ratio ranged from 0.9 to 2.0).
The crude incidence per 100,000 per years ranged from 2 to 26 cases. Lower crude incidence rates came from populations in China, Libya, and Sardinia. Most developed countries with northern European age structures tended to have incidence rates between 12 and 20 cases per 100,000 per year.
Comment
Here we have an interesting study that not only gives some insight into the incidence of Parkinson's disease, but that also give an insight into the difficulty of such studies, and the need for strict criteria to ensure quality.
With Parkinson's disease, which develops predominantly in older people (Figure 2), the age structure of a population would be expected to be critical to any result. A young population would have low incidence, while an older population would have a higher incidence.
Figure 2: Parkinson's disease incidence and age
The authors of the study give a list of criteria for improving the quality and consistency of incidence studies:
- Base population neither too large nor too small. Thus for an incidence of 17 per 100,000 per year they recommend one million person years of surveillance, generating 60-80 cases.
- Studies should be prospective to maximise case ascertainment and accuracy.
- Multiple sources be used to identify cases (records, secondary and primary care sources etc, nursing homes).
- Prospective cases should be seen by an expert.
- Incident cases should be defined by specific symptoms and diagnostic criteria.
- Clear and consistent inclusion criteria applied to population. Broad screening terms should be used to avoid missing cases.
- Studies should have a period of follow up, perhaps to determine response to therapy.
- Incidence rates should be reported by standard age strata (deciles, for instance), with confidence intervals. Crude information about the study population would be useful so that calculations are transparent.