All choices are not equal; effects of context on point‐of‐choice prompts for stair climbing

A recent article published in Obesity Reviews , Dolan et al . (1) set out to quantify the effects of point-of-choice interventions that encourage the use of the stairs at the expense of the escalator in community settings. While there is much to admire in the review, five of the 13 studies reviewed did not test stair use against escalators but rather against elevators. Unfortunately, the evidence for successful increases in stair climbing with point-of-choice prompts where the choice is between stairs and elevators is weak, albeit encouraging that change is possible (2). The net outcome is that the careful work by Dolan et al. (1). on a quantitative synthesis of the potential for calorific expenditure with stair climbing underestimates the public health gain possible from stair climbing interventions, combining as it does studies that involve different behavioural contexts. Table 1 summarizes the available studies for the choice between stairs and escalators in community settings over the period 1980–2004 covered in the review by Dolan et al . (1). The table includes some further studies overlooked by the review. In addition, only point-of-choice prompts that focused on health gain have been included (cf. 6). As is clear from the table, all studies report positive effects with only the use of a sign focusing on the health of the individual in an airport failing to produce a significant increase in stair climbing (3). The sample size weighted effect is 5.09%, somewhat larger than the 2.8% estimated by Dolan et al . (1). A further point is germane to any estimates of public health impact; point-of-choice prompts can generalize to other stair climbing opportunities. First, a number of studies demonstrate that stair climbing can remain elevated when the prompt is removed (4,5,3,6–8). Hence, the impact of point-of-choice prompts outlives their deployment in community settings. In addition, two recent studies, unavailable at the time of the review, have demonstrated that effects of point-of-choice interventions are not limited to the choice point at which they are placed (7,8). Thus, point-of-choice prompts can increase stair usage in situations where there is no prompt, e.g. a subsequent staircase (7), and as such the effects of prompting generalize to other stair use choices. This result indicates that any estimate of the population effects based on results from experimental staircases alone underestimates the public health impact. At this stage, it is impossible to quantify the amount of underestimation as we do not know the extent of the generalization possible with point-of-choice interventions. Finally, it is informative to consider why point-of-choice interventions have larger effects for escalators than for elevators (2). The discrepancy may reflect the behavioural context in which the choice is made. One can conceptualize stairs, escalators and elevators as barriers placed in the path of a pedestrian’s journey from their point of departure to their destination. For many such journeys, time to complete the journey is relevant and indeed, saving time is a common reason for using the stairs (9,10). When a pedestrian is faced by a choice between adjacent stairs and escalators, neither choice incurs a time penalty as both options are immediately available. In contrast, the choice between stairs and elevators is often influenced by time considerations (11). If the elevator is not immediately available, a pedestrian intending to take the elevator may choose the stairs as the quicker option. Alternatively, an individual aiming to take the stairs in response to an intervention may be seduced by the lure of an immediately available and Table 1 Percent change in stair use from health-related point-of-choice intervention signs for stairs vs. escalators