New insights into the Mayer Hypothesis derived from a dynamic energy balance model

Abstract

BackgroundIn 1956 Jean Mayer and his colleagues proposed a provocative relationship between energy intake (EI), body weight and physical activity (PA) by examining the intake of over 200 sedentary to highly active individuals in West Bengal, India. The investigators found that high levels of EI were associated with high levels of PA. Surprisingly, sedentary individuals were also associated with high EI, resulting in a hyperbolic shaped curve of EI versus PA referred to as the Mayer curve or Mayer hypothesis. Challenges arising from objectively measuring EI and PA in free living humans have made experimentally testing the Mayer hypothesis difficult.MethodsA mathematical relationship between EI and PA levels (PAL) was derived from a newly formulated one parameter (α) dynamic energy balance equation. To test the validity of the lower end of the Mayer curve, the relationship between the parameter α and PAL was analyzed in subjects having Class 1, 2, and 3 obesity from the Institute of Medicine doubly‐labeled water (DLW) database (N=133, BMI=34.9±5.6 kg/m2). Similarly, the relationship between EI calculated by summing changed body energy stores with DLW measured energy expenditures versus PAL in elite athletes were applied to test the validity of the higher end of the Mayer curve (N=60, 21.7±2.0).ResultsThe steady state formulation from the model yielded the algebraic relationship, EI α×PAL×Weight confirming EI was an increasing function of higher values of PAL as hypothesized on the right side portion of the Mayer curve. The theoretical finding was confirmed in the elite athlete database (EI=39.5PAL+1057.6 kcal/d, R2=0.2). On the other hand, the steady state model did not yield the Mayer curve for lower PAL values unless the parameter, α, was inversely proportional to PAL. The inverse proportionality between α and PAL was nearly statistically significant in Institute of Medicine database.ConclusionsThe Mayer hypothesis is theoretically and experimentally validated in individuals with higher PAL. There is some evidence that certain subpopulations of sedentary individuals may be reflected in the Mayer curve, however, additional data needs to be analyzed to determine which subpopulations follow the Mayer curve and which deviate from the Mayer curve.