Response to Letter to the Editor: “Twice as High Diet-Induced Thermogenesis After Breakfast vs Dinner on High-Calorie as Well as Low-Calorie Meals”

We thank Melanson and Chen for their interest in our article (1) which gives us the opportunity to discuss the methodological peculiarities of our study.

In their first point, the authors rightly note that diet-induced thermogenesis (DIT) after lunch was probably not yet completed at the time of dinner, resulting in a higher resting metabolic rate value before dinner than before breakfast. However, following their proposal to calculate with pre-breakfast values would result in the addition of potentially remaining DIT from lunch to dinner DIT. Hence, the comparison between breakfast and dinner DIT would not have been conclusive. When designing our study, we decided to integrate lunch to avoid long periods of fasting during the day and, moreover, to create equal thermogenetic conditions prior to dinner in the attempt to exclude any potential influence of the respective morning calorie load. In this context, however, another study, designed with an 8-hour fasting period before dinner, showed results that are consistent with ours (2).

In their second comment on the duration of DIT measurements, the authors refer to a study that examined inter-individual variation (3). In our crossover study (1), intra-individual differences were investigated and potential influencing factors were strictly controlled. Furthermore, the main outcome of our study—twice as high DIT after breakfast vs dinner—would only have been biased by too short measurement periods if the curve progression, i.e., increase/decrease and the time point of the peak, would significantly differ between morning and evening measurements. Our data show no evidence for this assumption (Fig. 3C and 3D) (1). In addition, from a practical implementation point of view, an extension of the postprandial measurement periods to 6 hours each would have led to a prolongation of the experimental day by at least 4 hours. In turn, the resulting reduction in sleep time would have affected energy metabolism.

Regarding the final point, it first needs to be clarified that the factor 2.5 given in the abstract is not a conclusion but the result of the analyses. It refers to the mean DIT difference between breakfast and dinner comprising all data (Fig. 3A), which is described in more detail in the results section (1). In terms of the magnitude, our results are very similar to those reported by Morris et al (2), who found that early DIT was 44% lower in the evening than morning. However, when assessing the meaning of our finding regarding the impact on energy balance, previous insight dissents from the opinion of Melanson and Chen, as also a relatively small thermogenetic effect may cumulate over a longer period. A study by Jakubowicz et al demonstrated that a weight loss diet with high-calorie intake at breakfast, in overweight and obese women, results in a significantly greater weight loss (−8.7 kg) after 12 weeks than an isocaloric diet with high-calorie intake at dinner (−3.6 kg) (4).

Additional information

Disclosure Summary: The authors have nothing to disclose.

References