Our finding (Kirollos et al., 1993) a functional pre-clinical state in spouses of sufferers from idiopathic parkinsonism (IP), cohabiting for half a century, suggests a domestic causative influence and is compatible with transmission. There are epidemiological and biological links and similarities between IP and Helicobacter pylori infection (Dobbs et al, 2000). We sought proof-of-principle, for a pathogenic role of infection, by a randomized, double blind, placebo controlled study of the efficacy of H. pylori eradication in IP (calculated sample size 56 subjects). Our most discriminant measure of brady/hypokinesia, mean stride length (MSL) at free walking speed (Kirollos ibid) was primary outcome for this cardinal sign. Gastric pathology, other than gastritis, led to exclusion. Assessment, at a set time of day, included analysis of two walks. Two baseline assessments preceded randomization. Active treatment consisted of one week's omeprazole (20 mg 12 hourly) with 12 hourly clarithromycin 500 mg and amoxycillin 500 mg: 12 hourly metronidazole 400 mg and/or 6 hourly tetracycline 500 mg was substituted in light of in vitro insensitivity and/or suspected intolerance to first line antimicrobials. De-blinding was scheduled for 54 weeks post allocated treatment, sooner if life style threatened by progression of IP. Those who had received a matched placebo regimen were offered active treatment, and 6 weekly follow-up continued. Eradication was assessed in endoscopic gastric biopsies by light microscopy (6 biopsies), culture and PCR for Helicobacter DNA (2). A linear model was used to determine whether within-subject time trends in MSL were different after active and placebo treatments. A time variable was constructed where day 0 is end of active regimen, subsequent to baseline and any placebo phase. Within-subject, the knot for active and placebo splines was at day 0. Subject was fitted as a block to enable within-subject analysis. Interactions of placebo or active splines with between-subject covariates [blind/open active medication, presence/absence of anti-IP medication (constant throughout, evenly spaced, excluded levodopa) and mean baseline MSL] were examined. Initial analysis was recommended to Steering by Data Monitoring Committee after de-blinding of 21 subjects (10 men, 11 women, mean age 60 (range 50-80) years). In the 16, who had reached post-active endoscopy (mean 323 days), treatment had failed in 4 [by culture (1) or DNA (further 3) criteria]. Their decline in MSL after active (4 subjects) and placebo (2 subjects) was similar (5 (95% C.I. 2, 9) & 5 (1, 96) cm.year^-1, over mean 212 & -213 days, respectively). In the remaining 17, the main effects, irrespective of whether active treatment was blind or open, and of baseline MSL, were a significant (P<0.001) change in MSL on active, none on placebo. Presence (4 subjects) or absence (13) of anti-IP medication had significant effects on trends in MSL (P<0.001 & =0.01 after active and placebo). Improvement on active was 5 (95% C.I. 2, 7) cm.year^-1 in the 13 not receiving anti-IP medication, and 37 (29, 46) in the 4 receiving, over 345 & 150 days. On placebo, there was no significant change in the 7 not receiving anti-IP medicine (-221 days), but a decline of 14 (4, 24) cm.year^-1) in the 2 receiving (-217 days). That anti-microbial therapy might have eradicated organisms, other than H. pylori, still supports the principle of a pivotal role for infection in IP. Decline in MSL after placebo in those on anti-IP medication may reflect greater severity of disease. Their greater improvement after active treatment may be due to an additional effect of improved absorption of anti-IP medication or loss of an antagonistic effect of inflammatory products.

Dobbs S.M. et al. Med Hypotheses; 2000; 55: 91-6.
Kirollos C. et al. Age Ageing 1993; 22: 20-6.