We thank Ms. Horne et al. for the clarification of our misquoting of their paper (Wood et al., Reference Wood2016). They clarify that 21% of their overall sample of patients with Parkinson's disease (PD-MCI) converted to dementia in over four years, which we erroneously attributed to the mild cognitive impairment (MCI) group in our discussion (McDermott et al., Reference McDermott, Fisher, Bradford and Camicioli2017). This was virtually identical to our overall conversion rate of 20%. Their conversion rate of patients with PD-MCI, as defined by two cognitive tests impaired (1.5 SD) within a single cognitive domain, was 51%, whereas the conversion rate was 38% when the PD-MCI group included patients with impairment within and between cognitive domains. Their conversion rates are similar to our rate of 42% (as defined with 1.5 SD impairment within or across domains) and the rate of 39% in a study with five-years of follow-up of incident cases (Pedersen et al., Reference Pedersen, Larsen, Tysnes and Alves2017). Our overall conversion occurred over a slightly shorter time span. In addition to conversion rates, all the studies acknowledge that some patients can revert to normal cognitive status, which varies based on classification criteria and length of follow-up. Comparable conversion across studies using similar criteria is reassuring and can encourage planning of targeted interventions (Hoogland et al., Reference Hoogland2017).

We agree that MCI subtyping and careful definition is essential. A key finding of all the reported studies is that cognitive impairment is important to recognize patients with PD and that PD-MCI is a harbinger of dementia, potentially identifying a group of patients for secondary prevention of dementia.