Table 2.

Likelihood ratios (LRs) and odds ratios (ORs) for PD risk factors collected in the PREDICT-PD pilot cohort using the PREDICT-PD risk algorithms and MDS prodromal criteria.

Factor	MDS (LRs)a	Basic PREDICT (ORs)	Enhanced PREDICT (LRs)
Factors in both MDS and PREDICT algorithms
Age	Categorical 5-year age intervals	Age-based equation	Age-based equation
Sex	Male 1.2, female 0.8	Female 0.67	Male 1.2, Female 0.8
Coffee use	LR+=0.88, LR−=1.35	0.67	LR+=0.88, LR−=1.35
Current smoker	LR+=0.51	0.44	LR+=0.51
Former smoker	LR+=0.91	0.78	LR+=0.91
1st degree relative	LR+=2.5	3.2	LR+=2.5
Constipation	LR+=2.5, LR−=0.82	2.3	LR+=2.5, LR−=0.82
Erectile Dysfunction	LR+=3.4, LR−=0.87	3.8	LR+=3.4, LR−=0.87
Depression/anxiety	LR+=1.6, LR−=0.88	1.86	LR+=1.6, LR−=0.87
Factors in one algorithm but not the other
Objective motor impairmentb	LR+=3.5, LR−=0.60	–	Bivariate Gaussian model based equationc
REM-sleep behaviour disorderb	LR+=2.8, LR−=0.89	–	LR+=2.8, LR−=0.89
Olfactory impairmentb	LR+=4.0, LR−=0.43	–	Logistic regression model based equationc
Pesticides exposure	LR+=1.5	–	LR+=1.5
Never smoked	LR+=1.2	–	LR+=1.25
Diabetes	LR+=1.50, LR−=0.97		LR+=1.50, LR−=0.97
Head injury	–	1.58	LR+=1.55, LR−=0.98
NSAID use	–	0.83	LR+=0.96,LR−=1.16
CCB use	–	0.9	LR+=0.94, LR−=1.04
Beta blocker use	–	1.28	LR+=1.19, LR−=0.93
Alcohol	–	0.9	LR+=0.98, LR−=1.09

^aBerg et al. ², Heinzel et al. ³

^bPreviously used as intermediate outcome markers and so did not feature in the basic PREDICT-PD algorithm.

^cBestwick et al. ⁸