Skip navigation

Coffee and Parkinson's Disease


How big is your coffee mug? How much do you drink every day? There may be all sorts of reasons for worrying about excess coffee of caffeine intake, but for those, like Bandolier , addicted to double espressos, there is some good news linking higher intake of coffee with a lower incidence of Parkinson's disease [1].


This was part of the Honolulu Heart Programme, a prospective study of 8,000 men of Japanese ancestry aged 45 to 68 years in Oahu in 1965. The survey has continuous surveillance of hospital and death records, with examinations at various times up to 1995.

At enrolment a detailed nutrient record was performed, verified by a full week dietary record in a subset of men. The record included details about coffee and tea intake (type and quantity) and about other caffeine containing drinks or sources. Other information, like smoking history, was also recorded.

Incident cases of Parkinson's disease were obtained through a full review of all hospital records, ongoing reviews of death certificates, and a cross check with local neurologists. Between 1991 and 1993 all subjects were examined and questioned about any history of Parkinson's disease, including signs and symptoms, and any medicines likely to be taken with someone with Parkinson's disease. A further survey was conducted between 1994 and 1996. All diagnoses were confirmed by a consensus of two neurologists using standard criteria.


Two men had Parkinson's disease at the start of the study, so there was information on 8004 men whose intake was determined in 1965 and who were followed up over 30 years. Their median age at enrolment was 53 years (range 45 to 68 years), and the median duration of follow up was 27 years (0.8 to 30 years). The median age of development of Parkinson's disease in 102 men was 74 years (range 54 to 89 years).

The Table shows the numbers of men in each category of coffee intake ( Bandolier has used the number of 8 fluid ounce mugs, each equivalent to two small cups, rather than fluid ounces or millilitres), the crude 30 year incidence, the age adjusted incidence per 10,000 person years, and the adjusted hazard ratio. The same information is available for quintiles of total daily caffeine intake.

Table: Coffee and caffeine consumption and Parkinson's disease

Daily coffee consumption (mugs) Parkinson disease / Total Percent (95% CI) Age-adjusted incidence per 10,000 person-years Adjusted relative hazard ratio (95% CI)
0 32/1286 2.5 (1.6 to 3.3) 10.4 5.1 (1.8 to 14)
1 33/2576 1.3 (0.8 to 1.7) 5.3 2.7 (1.0 to 7.8)
2 24/2149 1.1 (0.7 to 1.6) 4.7 2.5 (0.9 to 7.3)
3 9/1034 0.9 (0.3 to 1.4) 3.7 2.0 (0.6 to 6.4)
≥4 4/959 0.4 (0.0 to 0.8) 1.9 reference
Daily caffeine consumption (mg) Parkinson disease / Total Percent (95% CI) Age-adjusted incidence per 10,000 person-years Adjusted relative hazard ratio (95% CI)
0-123 35/1522 3.3 (1.5 to 3.1) 17.3 3.0 (1.1 to 8.4)
124-208 17/1396 1.2 (0.6 to 1.8) 5.4 1.1 (0.4 to 2.9)
209-287 26/1607 1.6 (1.0 to 2.2) 5.9 1.1 (0.4 to 3.0)
288-420 12/1485 0.8 (0.4 to 1.3) 4.3 0.8 (0.3 to 2.6)
421-2716 6/1481 0.4 (0.1 to 0.7) 5.0 reference
One mug of coffee assumes a volume of 8 fluid ounce or 200 mL; it is equivalent to two standard cups

Coffee drinkers had a significantly lower incidence of Parkinson's disease than men who did not drink coffee (Table, Figure). The same overall result was apparent for total caffeine intake as for coffee intake (Table). Coffee intake at enrolment was significantly related to Parkinson's disease that occurred during the first and second 15 years of follow up.

Figure: Incidence of Parkinson's disease and coffee drinking


This is the first time, apparently, that this inverse relationship between coffee intake and Parkinson's disease has been demonstrated conclusively. Is there a reason for this? Obviously there must be a reason, and there are many candidates, but right now we don't know what it is. The authors of the paper discuss possible mechanisms in great detail, and it is an interesting read.

Whether the message can be generalised to 'Drink coffee and avoid Parkinson's disease' is at least moot. Coffee addicts, though, will take some comfort from this report.


  1. GW Ross et al. Association of coffee and caffeine intake with the risk of Parkinson disease. JAMA 2000 293: 2674-2679.
previous story in this issue