Improving weight loss RCTs. Measuring the step weight change from a sustained change in frequency of a particular eating or exercise pattern

The public's trust in the science of avoiding unhealthy weight depends on a radical reform of the design and execution of weight loss programmes and their clinical trials. This Perspective reiterates the longstanding argument for measuring the effectiveness of each component of an intervention on obesity. Body energy content change results from a difference in rates between input and output. These rates are determined by the frequencies of specific patterns of dietary behaviour, physical activity and thermal comfort, plus the cost of resting metabolism. Since fat-free mass changes alongside fat mass, the amount of change in weight from a change in the frequency of a behaviour pattern comes to an asymptote. That step change in weight per unit of behaviour change is measured by regression from the change in frequency of the behaviour that has been maintained from baseline to follow-up. For hard evidence, weight loss programme participants' own words must be used to specify behaviour. In RCTs of multiple-component programmes, sequences of the behaviour patterns to be changed are randomised among groups. The resulting evidence on effective slimming practices can be delivered directly into therapeutic services and public health interventions for the culture investigated.