Actually, I have less issue with age in this study than other methodological and variable problems. The PD group and the Control group both had the exact same average age (approx. 73). But they only controlled for two other variables, sex and education.

I think a more comprehensive factor analysis would have been the appropriate statistical test. I also think that if this was done today, there may have been more focus on variables such as genetics; exposure to environmental hazards; medications; family history etc. Finally, they needed more assessment points. Every four years was too long of a delay and allowed for a significant number of drop-outs without data.

Link to the actual journal article:
http://bit.ly/16zPnp9