During the short-term intervention of four weeks, there was a marked increase of both fat free mass and of fat mass in both men and women [20, 21]. Body weight increased by 6.4 kg. After the intervention the subjects could go back to their usual eating- and physical activity habits and when followed-up 6 months later their body weight had decreased, however not to baseline values. There were individual differences of weight changes after the intervention but only one third of the participants returned to +0.5 kg or less of their baseline body weight at 6 or 12 months after the intervention and in average there was an increase of body weight of 1.5 kg one year after the intervention period. Our hypothesis was that body weight should return to baseline after one year and we therefore did not measure body weight of the controls neither at 6 nor 12 months after their last visit. However, body weight was measured in both groups after 2.5 years and at that time there was a further increase in the subjects who participated in the intervention but body weight was unchanged compared to baseline in controls.
Siervo et al  overfed 6 subjects stepwise during 3 periods of 3 weeks each (+20, +40, +60% increase of their baseline energy intake) and found an increase of fat mass by 3.1 kg and of fat free mass by 2.7 kg during the last step. For comparison our subjects increased their energy intake with in average 70% during the whole intervention  and their increase of fat mass was 3.7 kg. Norgan & Durnin  overfed 6 healthy men (+1500 kcal/day) during 42 days and they found an average weight gain of 6 kg corresponding to a 10% weight gain, and the increase of fat mass was 3.7 kg. Both these studies included few participants but show similar changes in body composition as our study.
In our study body composition was followed-up 12 months after the intervention and at that time we found an increase of fat mass compared to baseline while the initial increase of fat free mass had returned to baseline levels. Previously, overeating experiments have been performed to describe effects of increased caloric intake in human beings [15–17] but only few long-term follow-up studies have been made [18, 19]. The increase of fat mass, on average 1.4 kg, shows that the increase of body weight found after 12 months was fully caused by an increase of fat mass. In a study 23 young men were overfed during 100 days and followed-up 4 months later . During the 100 days of overfeeding body weight increased with on average 8.1 kg; both fat mass and fat free mass increased. When those men were followed-up 4 months later they had, just like our subjects after 6 months, lost most of the body weight gain but not to origin. Effects of massive overfeeding by carbohydrates during 4-6 months, the Guru Walla session (a traditional fattening session in Cameroun), were described by Pasquet and Apfelbaum . There was an increase of body weight by 19 (3.2) kg and of fat mass by 11.8 (2.5). After 30 months body weight had returned to baseline and only 5% of the increase of body fat remained but 1 participant who kept much of the gained body weight was excluded from the analysis.
Our finding of an increased body weight of 1.5 kg and fat mass of 1.4 kg over just 12 months raises the question whether a short period of overeating can induce a subsequent increase of fat mass. The suggestion of such an effect is supported by the fact that we did not observe a weight gain in the controls when body weight was measured after 2.5 years. Increases in body weight in adult individuals are common with increasing age but the changes described are smaller than what we observed. In Sweden a 10-year increase of body weight of 3.8 (6.0) kg has been found in individuals aged 25-64 years  which is clearly less than the increase we found in just one year if evenly distributed during the 10-year period. In the US obesity is more prevalent than in Sweden and increases of approximately 9.1 kg between the ages of 25-55 years have been described  and for most young adults there is a yearly increase of body weight of 0.2-0.8 kg . The subjects in the Canadian study by Tremblay et al  were re-investigated 5 years later  and were found to have increased their body weight by 5 kg. Although this was a very large increase of body weight, it was concluded that there was no persistent effects of exposure to the overfeeding protocol over the expected age-associated increases in body mass, body fat, upper-body fat, abdominal visceral fat, and metabolic variables.
In a Swedish study lean healthy young adults free from health problems have been found to be the most likely group to gain weight  which might indicate that young individuals being overweight or obese have increased the awareness of obesity related problems and avoiding further weight gain. We lack knowledge of our participants' attitudes towards a healthy lifestyle, physical activity and eating habits before inclusion but on the other hand, most of them were medical students which is why we believe that they were aware of the potential risks by gaining body weight. The intervention was also clearly described to all participants. However, due to the demanding design we did not randomize subjects to the intervention or the control group, which might have biased the long-term results. Speculatively some of the subjects in the intervention group could be less cautious about gaining body weight as they volunteered to participate in the intervention.
The accumulation of trunk fat mass has been reported to be related to total body fatness in both genders  and in healthy non-obese men trunk fat mass measured with DXA has been found to have high correlation to visceral adipose tissue measured with MRI . In the overfeeding study by Siervo et al  an increase of trunk fat mass was found and when fat volume derived from DXA was compared with abdominal fat volume derived from MRI they found visceral fat mass to have increased in greater extent than subcutaneous fat mass. We have previously reported an increase of abdominal fat volumes derived from MRI during the intervention  and men were more likely to accumulate fat mass in the abdominal region as visceral fat mass than women. In this study we found a non-significant tendency towards a greater increase of android than gynoid fat mass at 12 months. In addition the change appeared to be greater in men than in women but this was not statistically significant. We did not perform MRI 12 months after the intervention and cannot discern subcutaneous fat mass but the changes of trunk fat mass determined by DXA show that fat mass had increased above baseline values and fat free mass decreased to baseline in that area.
Weight gain due to overeating has been reported to be less than theoretically expected due to an increase of BMR . The increase of BMR has however been found to be related to body weight [15, 16, 18]. We found an increase of BMR during the intervention [20, 21] which secured our subject from an even greater weight gain but at the follow-up after the intervention BMR had returned to baseline in agreement with reduction of fat free mass to baseline level. Strong associations between fat free mass and BMR were found on all occasions confirming that fat free mass is of importance for BMR and can by itself reduce the risk for gaining extra body weight.
Interestingly we also found a small but significant increase of BMC during the intervention, which could be seen as an effect of increased body weight on bone. Measurements of BMC may on the other hand not completely correct for body size, especially when body size is changing dramatically  as in our study.
We have previously described an increase of serum ALT associated with increase of body weight and intake of energy from carbohydrates during the intervention . This long-term follow-up shows no remaining effect on these liver enzymes in spite of the increase of fat mass found one year after the intervention. On the other hand a small deterioration of the lipid profile was found.