Researchers from Harvard’s School of Public Health have shown that Pokemon Go does get people moving more but that effect doesn’t last.

Shortly after the release of Pokemon Go, there was a lot of enthusiasm that it would be a healthy addiction by getting users more active. In this study, Howe et al looked at activity data from the phones of more than 1,000 users from Amazon Mechanical Turk (MTurk). MTurk is a platform for recruiting workers worldwide to complete all kinds of tasks; in this case, that was completing a survey and sharing daily activity data.

Of the 1,182 participants, 560 played Pokemon Go. Among Pokemon Go players, steps went up from an average of 4256 steps per day (in the four weeks preceding download) to 5123 steps in the first week of playing the game. After that, though, there was a gradual decline back to where they started by the sixth week. That pattern was consistent across various subgroups based on neighborhood walkability, income level, age, and more.

Making health apps stick is not a new problem. The drop off seen in Pokemon Go was also well demonstrated in several ResearchKit apps. In MyHeart Counts, researchers reported a significant drop off in participation that began at 7 days. Similarly, in Asthma Health, a steady decline over weeks was shown.

It is somewhat surprising that even with a game as engaging as Pokemon Go, the benefits on activity level were so transient. That could have implications for the design of other digital health interventions directed at activity level – while there was a lot of enthusiasm about gamification, thats probably not going to be enough.