Obesity has steadily increased world-wide [¹], resulting in the emergence of various chronic diseases [²]. Consequently, the interest in weight control has become a matter of concern, as increased numbers of people attempt weight loss [³]. Weight control is desirable when it is for the maintenance of a healthy body; however, the supremacy of physical appearance in the modern era has caused more people to desire to become leaner and thinner [⁴]. Many people are dissatisfied with their looks in spite of having a normal body weight, as this misperception often leads to unnecessary attempts to lose weight [⁵]. Moreover, the overwhelming weight loss attempts by teenagers and those in their twenties have been identified as a major problem that can cause potential long-term health risks [⁶]. While some people try to maintain weight by adequate amounts of exercise and food intake, some choose inappropriate methods of weight loss due to personal financial conditions, social environment, and misperceptions. Incorrect usage of diet products, laxative medications, diuretics, appetite-suppressants, vomiting, and smoking are common examples [⁷]. This is becoming a social problem, particularly for teenage girls struggling financially, because many of them choose to fast and starve in order to lose weight.

Fasting has harmful influences on body metabolism [⁸], especially leading to gorging and weight gain. This yoyo phenomenon [⁹] is more apparent due to the misperception of weight loss and idle management after losing weight. Despite recent research on this yoyo phenomenon proving its harmful effects on health, many people still choose to fast in order to lose weight. Furthermore, people seem to disregard this yoyo phenomenon as being serious, and continue to fast after seeing the immediate effects.

Therefore, from a health point of view, there is an urgent need to identify the various effects of fasting and re-feeding on the body. However, with there being limitations to conducting repeated fasting-refeeding experiments in humans, animal experiments have been conducted to investigate effects on lipid and carnitine profiles. In this research, we used five groups of mice: one fed a normal diet while the others were subjected to fasting, and fasting-refeeding repeated once, twice, and three times, respectively. Food intake amount, body weight, serum and hepatic lipids, and serum and hepatic carnitine concentrations were compared and analyzed.

Before the start of the experiment, there were no significant changes in amount of food intake or weight between each mouse group; however, a distinguishable change was noticed after repeating fasting-refeeding (Table 2 and Fig. 3). The overall weight of all trial groups tended to decrease significantly when compared to the non-fasting diet group. During fasting, weight decreased and then tended to increase again with refeeding, however, weight did not reach the level of the non-fasting diet group. This corresponds with the results of Lee and Kim [¹⁴] and Jen et al. [¹⁵] and Kochan et al. [¹⁶], who also performed experiments on fasting-refeeding. Epididymal fat showed a tendency to significantly decrease in all trial groups compared to the non-fasting diet group. Conversely, liver weight decreased significantly more in the fasting group compared to the non-fasting diet group; however, liver weight significantly increased when fasting-refeeding were repeated (Fig. 3 and 4). Kochan et al. (2001) reported that bodyfat and epididymal fat dropped remarkably in energy restricted rats compared to a non-fasting diet group without restrictions [¹⁷]. Horton et al. (1998) observed a decrease in liver weight during fasting, but reported that with refeeding, weight increased even more than the initial weight before fasting [¹⁸].

Epididymal fat weight did not show any significant changes even after refeeding, and its percentage of body weight also remained unchanged. This might be due to low levels of TG accumulation, and increased lipolysis induced by excessive fasting, which was necessary to produce energy being stored in the fat tissue during refeeding after fasting.

Björntorp et al. (1983) reported in his research that after fasting and refeeding, glucosidase inhibitors restrained the absorption of glucose which delayed the storing of TG within fat tissues [¹⁹]. Also, in Lee and Kim's research, they reported that fasting and refeeding had an affect on the consumption of leptin, adiponectin, and UCP1 more than it did on fat accumulation such as by LPL [¹⁴].

On the other hand, the increase in liver weight, which was higher than when a normal feeding protocol was used, is considered to have resulted from the tendency to store nutrients. The liver plays a role in glycogen synthesis and TG storage, and with fasting, excess nutritional loss occurs in the liver. Refeeding provokes active protein synthesis that helps with body component restoration [²⁰] and adaptation, which allows the body to fight against future emergency situations. With the liver being the largest known nutritional storehouse in the body, an expansion in glycogen storage capacity would likely be responsible for the result.

In this research, repeated fasting-refeeding resulted in weight loss and decreased body fat, but also caused hypertrophy of the liver (Fig. 3, 4 and Table 3). An enlarged liver is usually caused by functional damage. The liver primarily stores glycogen, whereas TG is stored in only small portions, as it is regulated by choline, methionine, and lecithine, an anti-fatty liver element. There is a need to take this cautiously because when a functional problem exists in the liver, the balance breaks down and fat begins to accumulate, resulting in possible fatty liver formation.

Serum TG concentrations were significantly lower in all repeated (once, twice, three times) fasting-refeeding groups when compared to the normally fed group; the one time fasting group, although without significance, also showed a slight decreasing tendency (Fig. 5). As is well known, fatty acid and ketone bodies are used during fasting, which accompanies decreases in blood glucose, TG, leptin, and insulin [²¹]. Conversely, serum TG concentrations during refeeding not only decreased when compared to the non-fasting diet group, but also the fasting group. This result corresponds with Björntorp et al. (1982) [²⁰], as it is considered that in order to expand the storage capacity of exhausted protein and energy, blood glucose and TG concentrations decrease.

In other research it has been reported that refeeding lead to a decrease in fatty acid concentrations and caused quick fat synthesis, resulting in increased blood glucose and TG concentrations [¹⁶,¹⁸,²²,²³]. Each of these studies varied in terms of fasting period, animals, and overall experimental design, which leads to many different interpretations. There are limits in accommodating all these factors; therefore, more diverse, detailed research is needed in order to compensate for previous problems.

Contrary to several reports on fasting-refeeding and its induction of liver fat synthesis [^24-30], there were no significant changes in liver TG or TC content. However, serum LDL-cholesterol was highest in the three time repeated fasting-refeeding group, and its value was significantly higher when compared to the fasting group. Earlier studies also found the same type of relationship between fasting and increased ApoB gene expression and LDL production, which may be due to the inverse relationship between insulin deprivation and LDL production and LDL receptor-binding [^31-33]. HDL-cholesterol showed contrasting results, as values were lowest in the three time repeated fasting-refeeding group. HDL/TC values, known to induce arteriosclerosis, were also lowest in the three time repeated fasting-refeeding group. Repeated fasting and refeeding did not affect TC content, however, it did increase LDL-cholesterol, which is a cardiovascular risk factor, and lowering HDL-cholesterol and HDL/TC percentage are believed to have adverse effects on the body.

In this research, NEC (free carnitine) concentrations did present significant changes when both the non-fasting diet group and fasting group were compared (Fig. 7). When compared to the non-fasting diet group, the fasting group had a tendency to have somewhat higher serum acyl-carnitine (ASAC, AIAC) concentration, but the results were insignificant. It was indicated that increases in acyl-carnitine (ASAC, AIAC) concentrations during fasting were highly related to increments in ketone bodies [³⁴,³⁵]. In the repeated (once, twice, three times) fasting-refeeding groups, NEC, ASAC, AIAC, and TCNE all had a significant lowering tendency in comparison with the non-fasting diet group and the fasting group, by having varying numbers of fasting-refeeding repetitions. For the liver, when the fasting group and non-fasting diet groups were compared, NEC, ASAC, and TCNE levels showed a significant increase in the fasting group, while AIAC decreased significantly (Fig. 8). As reported in the preceding research, during fasting, the oxidation of long-chain fatty acids increased while liver TG synthesis decreased. Following up on this research, we found that the carnitine transport system within the liver induces active fatty acid oxidation.

On the other hand, when fasting-refeeding were repeated once, twice, and three times, significant decreasing tendencies of NEC, ASAC, TCNE, and Acyl/Free were present in all groups. This shows that stored lipids decreased during fasting, which may have lowered the expression of beta-oxidation regulating genes.

In summary, our study shows that repeated fasting and refeeding in mice results in visible effects such as weight loss and body fat reduction. Repeated fasting increased hepatic hypertrophy conditions and decreased rates of fatty acid oxidation, which favored a recovery of body weight by energy conservation. It is also noted that LDL levels increased and HDL levels decreased, which may lead to cardiovascular disease risk. Further studies are required to assess cardiovascular disease markers during fasting and refeeding states. These findings support the yoyo dieting phenomenon's harmful effects on health. Therefore, we suggest further investigations and monitoring of the same parameters in clinically obese patients requiring treatment.