Sage Journals: Discover world-class research

Abstract

Fasting of mice is a common procedure performed in association with many different types of experiments mainly in order to reduce variability in investigatory parameters or to facilitate surgical procedures. However, the effects of fasting not directly related to the investigatory parameters are often ignored. The aim of this review is to present and summarize knowledge about the effects of fasting of mice to facilitate optimization of the fasting procedure for any given study and thereby maximize the scientific outcome and minimize the discomfort for the mice and hence ensure high animal welfare. The results are presented from a number of experimental studies, providing evidence for fasting-induced changes in hormone balance, body weight, metabolism, hepatic enzymes, cardiovascular parameters, body temperature and toxicological responses. A description of relevant normal behaviour and standard physiological parameters is given, concluding that mice are primarily nocturnal and consume two-thirds of their total food intake during the night. It is argued that overnight fasting of mice is not comparable with overnight fasting of humans because the mouse has a nocturnal circadian rhythm and a higher metabolic rate. It is suggested that because many physiological parameters are regulated by circadian rhythms, fasting initiated at different points in the circadian rhythm has different impacts and produces different results.

Keywords

fasting mice refinement nutrition torpor

Fasting of mice is considered to be a standard procedure used in association with many different types of experiments.^1–3 Fasting is primarily used as a way of standardizing tests or before surgery.

Motivations for fasting mice include: Overnight fasting before measuring blood glucose to reduce variability in basal blood glucose, to assure a more uniform drug absorption by preventing mixing test drug with food; overnight fasting for toxicology testing; fasting-induced feeding to investigate anorexigenic compounds, in which case mice are usually fasted for 4 h up to overnight; and pharmacokinetic studies in general.^1,2,4–7

In this review the term ‘fasting’ refers to a state where the animal is completely deprived of food but has access to water, which is similar to the study designs of most of the references. In order to facilitate comparison of quantitative data, values are given in international units, and have been converted where required.⁸

The aim of this review is to summarize present knowledge about the effects of fasting mice to facilitate an optimization of the fasting procedure and thereby secure better scientific data, minimize the discomfort of the animals, and hence increase animal welfare.

Physiology of non-fasted mice

Food and water intake

Circadian rhythms regulate many aspects of the body physiology, including cellular and physiological systems such as hormone secretion rhythms, body temperature, hepatic energy metabolism, and sleep–wake cycles. The circadian rhythms are primarily paced by the suprachiasmatic nucleus (SCN) of the anterior hypothalamus in mammals. External stimuli such as light and dark influence circadian rhythms.^9,10 Light is the most potent stimulus, entraining the SCN via the retinohypothalamic tract.¹¹ It has been shown in rats that food intake and feeding pattern are disrupted if the SCN is lesioned.¹²

The sleep–wake pattern of mice has a characteristic nocturnal pattern; mice sleep more during photophase (light period) than during scotophase (dark period).¹³

Food intake and metabolism are closely associated with the function of the hypothalamo–pituitary–adrenal system (HPA).^14,15 HPA activity is normally of low amplitude at the beginning of photophase and of high amplitude at the beginning of scotophase, corresponding with the high activity level of the mouse in scotophase and the low activity level in photophase.

Mice consume more food and water during scotophase than during photophase (Table 1), in accordance with the nocturnal pattern. Approximately two-thirds of the total food and water intake are consumed during scotophase, and one-third during photophase.^16–21 Petersen reported the eating rate to rise five-fold at dusk.²² Even when held under continuous light conditions, mice are known to have a circadian feeding pattern.²³ The circadian rhythm of glucose concentration – the most important energy source – is characterized by an increase from the beginning of the photophase to a peak in the middle of the photophase, decreasing to the beginning of the scotophase, and then staying at a base level during scotophase or with a lower peak before dawn (Table 2).^24–26

Table 1.

Food and water intake in mice.

Photophase (07.00–19.00 h)	Scotophase (19.00–07.00 h)
Food (g)	0.95 ± 0.19	2.97 ± 0.45
Water (g)	0.99 ± 0.32	3.52 ± 0.49

Total food and water intake (mean ± SD) in eight male, 12-week-old C57BL/6 J during photophase (light phase) and scotophase (dark phase). Food: powder CE-2 diet. Data from Kurokawa et al. (2000).²⁰

Table 2.

Glucose concentrations of non-fasted mice.

Reference	Strain/gender	n	Age/weight	Glucose concentration	Time of measurement
Tsuneki et al. 2002⁷⁸	ddY/male (saline injected)	6	8–10 weeks	9.6 ± 0.5 (mmol/L ± SE)	10.00*
Oosterveer et al. 2008⁴⁶	Sv129/OlaHsd C57Bl/6 J/male	4–6	–	8.8 ± 0.3 (mmol/L ± SE)	08.00 Photophase: 0700–1900
Echavell & Epeledev, 1972^2 8	C3H-S/male	59	84 days	6.05 (mmol/L)	24 h mean
Gagliardino & Hernández, 1971²⁶	C3H-S/female	240	10 weeks	8.3 (mmol/L)	24 h mean
Nowell 1970²⁴	Albino/male	9	30 g	6.3 ± 0.2 (mmol/L ± SE)	08.00 Photophase: 08.00–20.00
Mühlbauer et al. 2009²⁵	Wild type MelAABB/ male	104	–	8.75 ± 0.14 (mmol/L ± SE)	24 h mean

SE: standard error. *12 h of light and 12 h of dark, but information about timing are not available.

A significant (P < 0.05) increase in blood glucose was seen at 13.00 h, compared with 08.00, 20.00 and 01.00 h (photophase 08.00–20.00 h). This is in agreement with the rhythm described above. Male albino mice sampled at 01.00 h show a significantly (P < 0.01) greater tolerance to glucose than mice sampled at 08.00, 13.00 and 20.00 h.²⁴ These results suggest that glucose uptake is faster in the scotophase than in the photophase. Similarly, mice show a circadian rhythm regarding their body weight, which increases during scotophase and decreases during photophase.²⁷

Deviations from these diurnal rhythms in eating behaviour and glucose fluctuations have been found, though Petersen demonstrated a significant (P < 0.01) blood glucose fall at dusk, and a circadian rhythm with a peak in blood glucose 2 h before photophase and a lower peak 2 h before scotophase (12 h light, 12 h dark) has been found.^22,28 Ahrén found no evidence of a circadian rhythm of plasma glucose in male and female NMRI mice.²⁹ Also, a diurnal rhythm has been shown in a study in which the mice had reduced food intake during scotophase.³⁰ Another study demonstrated a higher feeding activity during a certain time interval in the photophase, arguing that this might be due to disturbances of the mice in the same time span.²⁴ This might be a plausible explanation for the results by Bailey et al. as well.³⁰

In a study with eight mice, it was shown that the feeding and drinking events were synchronized, but that the mice drank infrequently without eating.²⁰ Kurokawa et al. did not provide data of mean meal time, mean drinking time, mean inter-meal interval time or mean interval between drinking time.²⁰ Another study found that mice had an average of 36 food bouts and 32 water bouts during one day and night, lasting approximately 3.00 min and 0.88 min, respectively. The average interval between food bouts was 34 minutes and between water bouts it was 42 minutes. Drinking typically followed eating within 5 min.¹⁸ Meal patterns were analyzed for seven days in a study, where the mice were fasted for 5 h in the photophase and fed ad libitum for the remaining 19 h. The average number of meals per day and night was approximately 20, lasting approximately 5 min and with an approximately 40 min inter-meal interval.³¹

The different results in the above-mentioned studies might be due to the differences in experimental design, regarding strain and age of mice, diet, housing and method of measuring. Information about bedding and ambient temperature were not given by Gannon et al.¹⁸ Type of food was not reported by Kurokawa et al., although it has been shown to affect eating patterns.^20,31

Ad libitum fed mice kept at thermoneutrality at 33℃ ate about 50% less than mice kept at 23℃.³² In a study by Bronson, mice housed at 11℃ had a 37% higher food intake compared with mice housed at 22℃.³³ The mice housed at 11℃ had cotton nests with a core temperature of 18℃ and thus were only exposed to the cold when feeding. Similarly, Overton and Williams reported a reduction in food intake when ambient temperature increased from 23℃ to 30℃.³⁴ Mice housed at 20℃ had a significantly higher food intake than mice housed at 25℃, demonstrating that a 5℃ difference in ambient temperature can cause a significant difference in food intake.³⁵

Bachmanov et al. measured food and water intake of 323 male mice from 28 inbred strains, including C57BL/6 J and BALB/cByJ (Table 3).³⁶ Significant strain differences were found regarding food and water intake, with an at least two-fold difference between strains with the highest and lowest trait values.

Table 3.

Mean food and water intake of mice.

Overall strain mean (mean ± SE)	Lowest strain mean – highest strain mean (mean ± SE)
Daily food intake – unadjusted (g/mouse)	4.4 ± 0.1	3.1 ± 0.1 – 6.3 ± 0.3
Daily food intake – adjusted (g/30 g bodyweight)	5.7 ± 0.2	4.3 ± 1.1 – 8.8 ± 0.2
Daily water intake – unadjusted (mL/mouse)	5.8 ± 0.2	3.9 ± 0.2 – 8.2 ± 0.3
Daily water intake – adjusted (mL/30 g bodyweight)	7.7 ± 0.3	5.7 ± 0.2 – 11.4 ± 0.5

Mean food and water intake in 28 inbred strains of mice. Food: pelleted Teklad Rodent Diet 8604. Ambient temperature: 23℃. SE: standard error. Data from Bachmanov et al. (2002).³⁶

The gastric content in ad libitum fed mice shows a circadian rhythm, being highest in the middle of scotophase and lowest towards the end of photophase, which is analyzed using measurements every fourth hour.^37,38 The gastric content was similar for animals with and without bedding.³⁸

In conclusion, the food and water intake in mice thus seem to be following a pattern with small, frequent meals associated with drinking. Food intake is dependent on housing temperature and strain and two-thirds of the food are eaten during the dark phase.

Hormones

Adrenal corticosterone is produced when the adrenal gland is stimulated by the adrenocorticotropic hormone.³⁹ Several studies show that the serum corticosterone level peaks near the transition from photophase to scotophase, followed by a gradual decrease during scotophase and stays at base level during photophase until the transition to scotophase. The lower peak is near the transition from scotophase to photophase.^39–45 There are significant differences in corticosterone levels over 24 h.^40,44 Corticosterone leads the activity level of the mouse, increasing as the activity level increases in scotophase and decreasing as the activity level decreases in photophase.⁴⁰

Food intake stimulates the pancreatic release of insulin. In a study by Ahrén, plasma insulin displayed diurnal variation with a nocturnal peak in NMRI mice, and the plasma insulin level reached a significantly higher peak in males than in females, but there was no difference between gender regarding the relative increase.²⁹ Other investigators have found a circadian rhythm characterized by an insulin peak (6 ng/mL) at or just before the transition from photophase to scotophase.^25,26,43 The insulin level at approximately one hour into the photophase is 1.59 ng/mL.⁴⁶

The ghrelin hormone is mainly produced in the stomach and regulates feeding activity, being elevated during fasting and suppressed after food intake. Food intake is stimulated by activating neuropeptide Y-ergic neurons in the hypothalamus.¹³ Other hormones regulating metabolism include adrenalin, noradrenalin, glucagon, growth hormone, thyroid hormones, leptin and resistin.⁴⁷

Body temperature

The body temperature of mice follows a circadian rhythm, with a peak body temperature at the beginning of scotophase and the lowest body temperature in the middle of photophase.^13,48 It has been shown that the 24 h mean temperature is about 37.7℃, the lowest hourly mean 36.7℃ and the range of hourly means 1.9℃.^49,50 Corresponding with these results, Swoap reported a photophase mean body temperature of 36.7℃ and a scotophase mean body temperature of 37.8℃.⁵¹ As in other mammal species, the temperature is highest in the activity period and lowest in the resting period.

Fasting of mice: physiological changes

Non-food and water intake

Mice housed on solid flooring have the opportunity to consume faeces (coprophagia) during fasting periods, as well as bedding, environmental enrichment, biting sticks and nesting material if provided. Mice usually have access to water ad libitum during fasting, but since the water intake, as earlier described, is related to food intake, fasting also influences water intake.

In a study by Kutscher, mice fasted for up to five days (they could not survive any longer) displayed different drinking patterns: consistent polydipsia, sporadic polydipsia, or polydipsia restricted to the beginning or the end of the fast.¹⁶ In total 35% had polydipsia and 42% had decreased drinking activity (compared with their intake measured while not fasted). Five of the polydipsic mice showed this pattern on day 1. Both polydipsic and non-polydipsic mice kept their nocturnal drinking pattern during three days of fasting, except that non-polydipsic mice seemed to have a disturbed drinking pattern after three days of fasting.

Some strains of inbred mice show polydipsia during fasting, whereas other strains show hypodipsia. In a study where the circadian rhythm of water intake during fasting was measured, the fasted mice showed a pronounced circadian rhythm in their water intake.⁵² Significant (P < 0.05) reduction of the water intake in 24 and 48 h fasted mice was reported by Williams et al., resulting in an approximately 66% reduction after 48 h of fasting.⁵³

Hormones

Among others, the corticosterone hormone is known to be stress responsive, and is therefore relevant when exploring the effects of fasting. Stress has been defined as a response of an organism to external stimuli or change.⁴⁵ It is characterized by the activation of the autonomic nervous system and HPA axis and thereby an increase of glucocorticoids, such as corticosterone.^45,54

Fasting also stimulates the sympathetic nervous system. In rats it was shown that increased noradrenalin turnover caused an elevated sympathetic activity, resulting in fatty acid release.⁵⁵

Mice of the C57BL/6 J strain, fasted for 2, 4, 6 and 12 h had significantly (P < 0.05) lower corticosterone levels than mice fasted for 16 h (overnight). The highest level was reached after 16 h of fasting, being 477 nmol/L for males and 1251 nmol/L for females.⁵⁶ Male mice, fasted for 24 h, had a significant (P < 0.05) increase in corticosterone, reaching 612.68 nmol/L; and fasting of male mice for 48 h, resulted in a significant (P < 0.05) increase in corticosterone, reaching 982.6 nmol/L.^8,57

By comparison, mice of the BALB/c strain, (gender not given), stressed by swimming until exhausted, had a significant (P < 0.05) increase in serum corticosterone at all times of the circadian rhythm compared with the control group. The highest concentration measured was approximately 1445 nmol/L.⁴⁵ Male Swiss mice exposed to a cage containing a rat (a predator) had significantly (P < 0.05) higher plasma corticosterone levels after 5 min of exposure, than mice exposed to an empty cage. The plasma corticosterone level reached 404.6 nmol/L after 15 min.⁵⁴

Corticosterone levels are complex to compare, since there are gender differences, as shown by Champy et al., and there might be strain differences.⁵⁶

Leptin is a hormone secreted from adipose tissue in proportion to fat storage reserves, regulating the whole-body energy balance. Mice fasted for 2, 4 and 6 h had a significantly (P < 0.05) higher leptin level than mice fasted for 16 h. There was no significant difference between 12 and 16 h of fasting.⁵⁶ Mice fasted for 24 h, had a significantly lower serum leptin level than did non-fasted mice.^58–61 This indicates that a stable, low leptin level can be reached after 6–12 h of fasting.

Neuropeptide Y (NPY) is down-regulated as a result of increase in adipose tissue and leptin secretion. NPY is appetite stimulating. Mice fasted for 48 h had a significant (P < 0.05) increase in NPY, corresponding with the lowered leptin level, decreasing the down-regulation of NPY.⁸

Ghrelin is elevated in response to fasting and hunger. The high ghrelin production during fasting might be associated with an increase in glucagon. Ghrelin stimulates feeding behaviour via the nervous system and is antagonistic to leptin.⁶² Ghrelin gene expression in the stomach of mice was increased by 48 h of fasting.⁶³

Resistin is, like leptin, a hormone secreted by adipocytes. Resistin increases with high-fat feeding and causes decreased sensitivity to the effects of insulin. It is decreased by 48 h of fasting.⁶⁴

An insulin drop is considered to drive the metabolic adaptation to fasting.⁶⁵ There was no significant difference in plasma insulin levels between mice fasted for 16 h and mice fasted for 4 h. When measured under hyperglycaemic conditions, 16 h fasted mice had a significantly (P < 0.01) lower plasma insulin level, suggesting an increased insulin clearing. The whole-body insulin sensitivity index was significantly (P < 0.01) higher in 16 h fasted mice than in 4 h fasted mice.⁶⁶ In another study, there was a significant (P < 0.05) lower insulin level after 16 h of fasting compared with 2, 4 and 6, but not 12 h of fasting.⁵⁶ Mice fasted for 18 h had a significantly (P < 0.05) lower insulin level than mice fasted for 5 h.²¹ Mice fasted for 24 h had significantly (P ≤ 0.001) lower insulin levels than non-fasted mice.⁶⁵

Thyroxine (T4) and triiodothyronine (T3) are regulators of basal metabolic rate. Both were significantly (P < 0.05) decreased after 24 h of fasting and after 16 h of fasting compared with 2 h of fasting.^56,57 T4 was also significantly (P < 0.05) decreased after 48 h of fasting, but there was no significant change in thyroid-stimulating hormone.^8,57 A reduction of T3 in relation to fasting can contribute toward conserving energy, because T3 stimulates the breakdown of carbohydrates, proteins and lipids.⁶⁷

Testosterone was significantly (P < 0.05) decreased after 24 h and 48 h of fasting, and after 16 h of fasting compared with 4 h.^8,56,57 The weight of the testis was not affected by fasting.⁵⁷

In summary, corticosterone, NPY, ghrelin and resistin increase and T4, T3, testosterone, insulin and leptin decrease in response to fasting.

Weight loss

Several authors report differing weight loss in mice after different durations of fasting (Table 4). Mice experience a significant weight loss after 12 h of fasting, 18 h of fasting, 24 h of fasting, as well as after 48 and 72 h of fasting.^{53,65,68–70}

Table 4.

Relative (%) body weight loss of fasted mice.

Reference	Strain/gender	n	Age/weight	Duration of fast (h)	Weight loss (%)
Prior et al. 2009⁷¹	CD-1/male	8	35–51 g	3	0^†
Ayala et al. 2006²¹	C57BL/6 J/ male	6	9 weeks	5	5^†
Prior et al. 2009⁷¹	CD-1/male	8	35–51 g	6	6^†
Sokolović et al. 2007⁶⁸	FVB/male	8	6 weeks	12	12*
Ayala et al. 2006²¹	C57BL/6 J/ male	6	9 weeks	18	16^†
Prior et al. 2009⁷¹	CD-1/male	8	35–51 g	18	14^†
Warwick et al. 1985⁶⁹	Swiss albino/male	7	15–18 g	18	17*
Furner & Feller 1971⁷⁰	–/male	20	39.3 g (mean)	24	11*
Strubelt et al. 1981¹	NMRI/male	20	25–32 g	24	20^†
Kanizsai et al. 2009⁸⁴	C57BL/6/–	–	–	48	21^†
Furner & Feller 1971⁷⁰	–/male	16	39.3 g (mean)	48	21*
Furner & Feller 1971⁷⁰	–/male	17	39.3 g (mean)	72	23*
Sokolović et al. 2007⁶⁸	FVB/male	8	6 weeks	72	30*

Significantly different from non-fasted mice. ^†No data about significance available.

Mice lost significantly more weight after 16 h of fasting compared with 4 h, and a significantly (P < 0.05) larger weight loss was reported after 18 h of fasting compared with 5 h of fasting, 16% and 5%, respectively.^21,66 Sokolović et al. found that after 12 h of fasting the weight loss was steady at a rate of approximately 7% per day, based on a total of 72 h of recordings.⁶⁸

Mice are typically used for experiments at a relatively young age, and normally they would gain weight through natural growth, thus the relative weight loss is more severe if corrected for expected weight gain, depending on the duration of the fast.^71,72

Mice fasted for 18 h had significantly (P < 0.05) lower gastrointestinal tract weights than non-fasted mice, but there were no differences in brain weight. In this case, 40% of the weight loss was due to gastrointestinal reduction and 60% was due to apparent body mass loss. Liver, kidney and heart weights were also significantly reduced.⁶⁹ In agreement with this, liver weight was significantly (P < 0.05) decreased by 31% in 24 h fasted mice, as well as after 48 (33%) and 72 (32%) h of fasting, compared with non-fasted mice.^1,70 No change in hepatic water content or liver to body weight ratio was seen after 24 h.¹ Liver weight/100 g body weight ratio was significantly (P < 0.05) decreased in 24 and 48 h fasted mice, but not after 72 h of fasting.⁷⁰

Fat storages were dramatically decreased after 28 h of fasting, and not much different from the decreases seen in mice dead from starvation.³³

Gastrointestinal emptying

An empty stomach, small intestine and/or large intestine might be required in some studies including surgery or investigation of oral/gastrointestinal absorption. In a study by Prior et al., mice were fasted for 0, 3, 6 and 18 h.⁷¹ There was no significant difference in the gastric content weight between 6 and 18 h fasted mice, and no difference between 0 and 3 h fasted mice, but there was a significant increase between 0 or 3 h, compared with 6 or 18 h. The same pattern was observed for intestinal transit time of charcoal. It was concluded that fasting for 6 h gave similar results to fasting for 18 h regarding gastric emptying and intestinal transit time of charcoal.

Another author showed that mice fasted for 4 and 8 h, with the start of fasting from 4 h before photophase, had significantly lower gastric contents than non-fasted mice. Interestingly, mice fasted for 12 h, reported as a complete emptying of the stomach, did not have a significantly lower content than non-fasted mice. Non-fasted mice had a spontaneous decline in their feeding activity, resulting in a low gastric content at a point coinciding with the mice fasted for 12 h, and therefore no significant difference was seen after 12 h of fasting.³⁷ Bronson reported that most of the food eaten in the scotophase cleared the stomach within 6 h of fasting in the photophase, and after 14 h of fasting in the photophase, the small intestines were essentially empty.³³

During three days of fasting, mice excreted boluses of their bedding material, sterolit (a granular clay material), mean weight being 1.8, 1.1 and 0.5 g, respectively, for the first three days of fasting, indicating intake of bedding.¹⁶ Yeung et al. found that mice fasted for 24 h on grid floor had a significantly (P < 0.01) lower gastric content than mice kept on a solid polypropylene floor, thus indicating coprophagia.⁷³

The mean residual gastric content after 18 h of fasting was 0.076 g.⁷⁴

Fasting for 29 h resulted in a significantly lower gastric content compared with non-fasted mice. The gastric content of fasted mice showed a circadian rhythm similar to that of non-fasted animals. Mice that had bedding had a 72% mean drop in gastric content, while mice without bedding only had a 50% drop.³⁸ It was emphasized that the relative variation was identical under both fasting circumstances, but no explanation of the difference in the total decrease in the mean gastric content was given. The gastric content of the 29 h fasted mice might be secretions from the stomach.

Gastric emptying after administration of a simple, oral solution was faster in 24 h fasted mice than in non-fasted mice, although no information about significance was given.⁷⁵ Gastric emptying time is associated with stomach volume and body size.⁷⁶

In conclusion, gastric content of mice follows a circadian rhythm associated with food intake. Shorter periods of fasting do not necessarily lead to less gastric content than non-fasted mice would experience at certain points in the circadian rhythm. There is still some form of gastric content after 29 h of fasting.

Metabolism

Blood glucose decreases during fasting due to lack of food intake and thereby to lack of glucose absorbed from the intestines.

Mice fasted for 9 h had significantly (P < 0.05) lower blood glucose than non-fasted mice and a further significant (P < 0.05) decrease from 9 to 24 h of fasting.⁴⁶ Mice fasted for 16 h had significantly (P < 0.05) lower plasma glucose concentrations than mice fasted for 4 h.^56,66 Fasting mice for 48 h caused a 50% decrease in plasma glucose levels. There was no further decrease after 72 h.⁷⁷ Fasting for 29 h (with or without bedding) caused a significantly (P < 0.001) lower serum blood glucose than no fasting, being 26% lower for mice provided with bedding and 6% lower for mice without bedding.³⁸ Assuming that removal from one cage to another and/or absence of bedding can cause an increased blood glucose level provide a possible explanation for the latter difference.

In mice given saline injections every other hour during 6 h of fasting, a gradual decrease of approximately 40% in the blood glucose was seen. The decrease during 6 h of fasting, preceded by 10–14 h of fasting, was 12–17%.⁷⁸ This illustrates that the mice had a more rapid decrease in the first 6 h of fasting than 6 h after being fasted for 16–20 h (Table 5). Significant (P < 0.01) variability among different strains of mice regarding mean blood glucose is found in mice fasted for 4–6 h.⁷⁶ When the 29 h fasted serum glucose level was measured every fourth hour for 24 h, a circadian rhythm could be seen, described to be of the same magnitude and shape as for controls.³⁸

Table 5.

Glucose concentrations of fasted mice.

Reference	Strain/gender	N	Age/ weight	Duration of fast (h)	Glucose concentration (mmol/L ± SE)	Time of measurement
Oosterveer et al. 2008⁴⁶	Sv129/OlaHsd C57Bl/6 J/male	4–6	–	9	5.2 ± 0.3	08.00 Photophase: 07.00–19.00
Oosterveer et al. 2008⁴⁶	Sv129/OlaHsd C57Bl/6 J/male	4–6	–	24	3.5 ± 0.4
Heijboer et al. 2005⁶⁶	C57Bl/6/male	≥ 9	12–16 weeks	4	5.9 ± 0.7	09.00*
Heijboer et al. 2005⁶⁶	C57Bl/6/male	≥ 9	12–16 weeks	16	4.9 ± 0.9
Pessacq et al. 1976³⁸	C3H-S/female	12	8 weeks	29 + bedding	5.8 ± 0.2	24 h mean
Pessacq et al. 1976³⁸	C3H-S/female	12	8 weeks	29 ÷ bedding	7.2 ± 0.2
Tsuneki et al. 2002⁷⁸	ddY/male (saline injected)	6	8–10 weeks	6	6.6 ± 0.3	10.00*
				10–14	4.7 ± 0.2	10.00*
				16–20	3.9 ± 0.0	16.00*
Nowell 1970²⁴	Albino/male	9	30 g	24	3.4 ± 0.4	08.00–09.00 Photophase: 08.00–20.00

SE: standard error. *12 h of light and 12 h of dark, but information about timing of lights are not available.

Hepatic glucose production was not significantly different from 4 to 16 h of fasting.⁶⁶

Muscle specific glucose uptake under hyperinsulinaemic conditions was significantly (P < 0.01) higher in 16 h fasted mice than in 4 h fasted mice. This indicates that muscle insulin sensitivity is increased by the fast.⁶⁶

Glucose can be stored as glycogen in the liver, kidneys, heart and muscles. This reservoir is mobilized by glycogenolysis during the early stage of fasting, in order to maintain blood glucose concentration. In mice, hepatic glycogen is significantly (P < 0.05) lower after 18 h of fasting than after 5 h of fasting.²¹ It is significantly (P < 0.05) lowered after 24 h of fasting, being only 7% of its original concentration.¹ After 48 and 72 h of fasting it is largely depleted.⁷⁷

There was no significant (P < 0.05) change in the glycogen content of the gastrocnemius muscle after 18 h of fasting compared with 5 h.²¹ In 24 h fasted mice the muscle glycogen was depleted to about half the level of that in non-fasted mice.⁵⁷ The glycogen depots of the muscles are not provided for maintaining blood glucose, but for muscle contraction.

Prolonged fasting induces an increase in plasma glucagon, plasma glucocorticoids, plasma (nor) epinephrine and a decrease in plasma insulin concentrations. This leads to hydrolysis of triglyceride (TRIG) in adipose tissue, which causes a rise in plasma free fatty acid (FFA) concentration.^65,79 There was a significantly (P < 0.05) higher fat mass loss in mice fasted for 18 h compared with mice fasted for 5 h.²¹ Strain differences exist regarding plasma FFA levels during fasting: BALB/c mice had a significant (P < 0.05) increase in plasma FFA after 6 h of fasting, and a further significant increase after 14 h of fasting. C57BL and SWR mice did not have any significant differences after 6 h of fasting, but after 14 h of fasting there was a significant increase.⁸⁰ There was no difference in plasma FFA concentrations in 16 h fasted mice compared with 4 h fasted mice.⁶⁶ After a 24 h fast, the plasma FFA concentration was significantly higher in fasted mice than in non-fasted mice.^65,79 After 48 h of fasting, there was a nearly two-fold increase in plasma FFA, and after 72 h of fasting, the level was nearly back to normal.⁷⁷ Carnitine transports FFAs from the cytosol to the mitochondria where the oxidation to acetyl-CoA takes place. A dramatic increase in hepatic carnitine was seen after 48 and 72 h of fasting.⁷⁷

The FFAs are taken up by the liver, and either re-esterified or oxidized (β-oxidation). Re-esterification leads to synthesis of TRIG. These are secreted as very low density lipoproteins, which can provide FFAs for other tissues. Hepatic steatosis is a pathological increase in hepatic TRIG. Hepatic TRIG was six-fold higher in 16 h fasted mice than in 4 h fasted mice (P < 0.01), significantly (P < 0.05) increased about three-fold in 24 h fasted mice and increased about 10-fold following 48 and 72 h of fasting.^1,66,77 There are differences between mouse strains regarding hepatic TRIG levels during fasting: BALB/c and C57BL mice had a significant (P < 0.05) increase in hepatic TRIG after 6 h of fasting and a further significant increase after 14 h of fasting, but SWR mice did not show any significant difference before 14 h of fasting.⁸⁰ There was a significant (P < 0.01) increase in liver mRNA expression levels for transcription factors and proteins involved in TRIG synthesis after 16 hours of fasting compared with 4 h.⁶⁶ Plasma TRIG levels decreased from 2 to 16 h of fasting and following 48 and 72 h of fasting.^56,77 There was no significant change in muscle TRIG content between 16 h and 4 h fasted mice.⁶⁶ No increase in kidney or heart TRIG content appeared, possibly due to a controlled uptake that does not exceed requirement.⁷⁷

The majority of the FFAs that are oxidized by the liver are only partially oxidized, to acetyl-CoA, which condenses itself into ketone bodies, providing an energy source for the brain.^65,79

Plasma ketone bodies were significantly (P < 0.01) higher in 16 h fasted mice than in 4 h fasted mice.⁶⁶ After 6 h of fasting no significant (P < 0.05) difference in plasma β-hydroxybutyrate (former 3-hydroxybutyrate) were seen in three mouse strains (BALB/c, C57BL and SWR), but after 14 h of fasting there was a significant increase.⁸⁰ Fasting for 24 h significantly increased plasma β-hydroxybutyrate in mice, and this increased 21-fold in a study by Gavrilova et al.^57,65,79 Fasting for 48 and 72 h also produced a dramatic increase in β-hydroxybutyrate.⁷⁷ Genes encoding enzymes used in β-oxidation of fatty acids were highly up-regulated after 24 and 48 h of fasting, because the body starts to break down fatty tissue to provide energy.⁸¹ By contrast, genes encoding fat synthesis enzymes were down-regulated.^66,81

The FFAs fully oxidized by mitochondria, provide energy for hepatic gluconeogenesis, using substrates as glycerol from hydrolysis of TRIG in adipose tissue, lactate from glycogenolysis in skeletal muscle and amino acids.⁷⁹ Plasma lactate levels decreased gradually during 72 h of fasting.⁷⁷ Hepatic mRNA expression levels of transcription factors and related proteins involved in gluconeogenesis were elevated, while those for glycogenolysis were lowered, in 16 h fasted mice compared with 4 h fasted mice.⁶⁶ The latter is because the liver glycogen depots are decreasing, thereby depleting the substrate for glycogenolysis. Total hepatic glucose production was not altered.⁶⁶ No change in urea (a product of protein catabolism) was seen after 24 h of fasting or during 2 to 16 h of fasting.^56,57 These results indicate that glycogen is rapidly utilized, lactate and glycerol are utilized for gluconeogenesis, but protein breakdown is retained after 24 h of fasting.

In conclusion, the metabolic state of fasting can be described as catabolic or post-absorptive characterized by mobilization of glycogen providing glucose and increased mobilization of adipose tissue, again providing FFAs, utilized for production of ketone bodies and TRIG and providing energy for gluconeogenesis using lactate, amino acids and glycerol as substrates.

Liver parameters

Hepatic steatosis is seen after 16 h of fasting in mice.⁶⁶ After 48–72 h of fasting subtle steatosis is seen in the liver in centrizonal hepatocytes. After 66–72 h, the fatty change was minimal with few microvesicular lipid droplets. After 96 h, no fatty change was seen, and the hepatocytes revealed features of cytoplasmic atrophy.⁷⁷

No significant difference in cytochrome P-450 level was seen after 12 h of fasting, but was significantly (P < 0.05) lower (about 33%) after 24 h of fasting in mice.⁸² This is interesting since cytochrome P-450 contributes to the detoxification of xenobiotics. Aniline hydroxylase activity and aminopyrine N-demethylase activity were significantly (P < 0.05) increased as was glucuronyl transferase activity (P < 0.01).^1,82

Reduced glutathione is an antioxidant which is significantly decreased after 17 h and 24 h of fasting.^1,83

Body temperature

Mice fasted for 24 h had a significantly (P ≤ 0.01) lower body temperature than non-fasted mice. The temperature drop was approximately 3℃.⁶¹ Fasting for two days provoked progressive temperature falls during photophase, with temperatures down to approximately 27℃ at the end of the second day. Temperature rose back to normal in the scotophase.⁸⁴

Torpor has been defined as a suppression of metabolic rate to a level as low as 30% of the basal metabolic rate for a period of several hours by the lowering of the body temperature. Torpor has also been defined as a body temperature of less than 31℃.^85,86

Non-fasted mice do not normally exhibit torpor.^87–89 Mice can become torpid after about 7 h of fasting, with a body temperature drop of up to 15℃.^60,89,90 The mean time from initiation of fasting to first torpor bout was reported to be 13.3 h in a study by Brown and Staples.⁸⁹ It has been stated that torpor typically occurs after 8 h of fasting, and/or at the end of the dark cycle which is coincident with the peak of circulating ghrelin. Ghrelin is able to deepen torpor.⁹¹ A food restriction of 50% for three days also induced torpor.⁵¹

Fasting-induced torpor has been recognized in both sexes, young and adult, individually- and group-housed animals and in several strains of mice, both inbred and outbred.^86,88 Free-living mice in nests have also been found torpid during the cold months.⁹² After 24 h of fasting 35% of mice were torpid, and after 48 h of fasting 100% were torpid.⁸⁸ Group-housed mice had a significantly greater torpor incidence than singly-housed mice. Mice that were fasted several times all experienced torpor at some point.⁸⁶

The minimum possible body temperature in mice is 16–19℃.⁸⁷ During three days of fasting, mice typically undergo at least one bout of torpor each day, usually beginning at the end of scotophase. Swoap et al. found the rate of temperature decrease to be 0.16℃ per minute.⁶⁰ Some mice become hypothermic, characterized by an apparent inability to spontaneously arouse without assistance, and are likely not able to survive as long as those undergoing a daily torpor.⁸⁹ Emergence from torpor is typically finished within 30–45 min of the initiation from recovery, maximally gaining 0.26℃ per minute.⁹⁰ Handling of mice for subcutaneous injection or insertion of a rectal probe causes immediate arousal from the torpid state.^88,90 Mice are also reported to arouse spontaneously and as a result of noise or moving of the cage.^86,87

It is suggested that entry into torpor is a consequence of a cold environment or fasting. Norepinephrine release induces β3-mediated sympathetic activation of white adipocytes, causing suppression of leptin secretion and an elevation of lipolysis, resulting in a liberation of FFAs to the blood.⁹⁰ Torpor is likely regulated by a complex interaction among ghrelin, leptin, insulin and hypothalamic neurons.⁹¹ Mitochondrial respiration is actively suppressed during daily torpor in fasting mice. Mitochondrial respiration rate is a measure for the activity of the components of oxidative phosphorylation that control it.⁸⁹

In conclusion, torpor seems to be a way of reducing the energy expenditure during times of actual or predicted food shortage, and mice can enter torpor when temperatures are low or during times of fasting.

Cardiovascular parameters

A negative energy balance leads to decreased heart rate and blood pressure.⁵¹ Mean arterial pressure is significantly (P < 0.05) decreased during scotophase, but not during photophase of 24 and 48 h fasted mice.⁵³ Heart rate is significantly (P < 0.05) decreased after 24 and 48 h of fasting in mice with the fast starting 2 h before the scotophase.⁵³ After 4 h in the scotophase, the heart rate started to decrease, reaching the lowest point in the transition from scotophase to photophase. Oxygen consumption is significantly (P < 0.05) decreased after 48 h of fasting in mice, but not after 24 h.⁵³

Bradycardia and hypotension are suggested to be part of the torpor response. The bradycardia patterns were noticed within 6 h of fasting.⁵³ The gene that was mostly up-regulated after 24 and 48 h of fasting was cytochrome P-450 4A14. Disruption of this gene has been associated with hypertension.⁸¹

The obtainable blood volume decreased significantly after 24 h of fasting and a further significant decrease was reported after five days of fasting.⁹³

White and red blood cells, haemoglobin, mean cell volume and platelets were significantly decreased after 2 h of fasting compared with 16 h of fasting, whereas haematocrit was significantly (P < 0.05) increased. No difference was seen in mean corpuscular haemoglobin and mean cell haemoglobin concentration.⁵⁶

Toxicology

In tests with bunitrolol or pindolol (β-adrenoceptor blockers), mice fasted for 17–20 h developed clinical signs comparable with non-fasted mice about 10 min earlier, acute oral toxicity was significantly higher and death was hastened compared with non-fasted mice.⁹⁴ The blue metabolite of pindolol (β-adrenoceptor blocker) appeared in the urine of non-fasted mice after one hour, but in the urine of 17–20 h fasted mice, after half an hour.⁹⁴

In a toxicity study testing the hepatotoxic response to carbon tetrachloride, no significant difference in aspartate aminotransferase (AST), alanine aminotransferase (ALT) and sorbitol dehydrogenase (SDH) was seen after 6 h of fasting, but after 12, 24 and 36 h of fasting AST, ALT and SDH were significantly increased in fasted mice compared with non-fasted mice. The increase was enhanced after 24 h but not after 36 h. Similarly, in 24 h tests with thioacetamide, paracetamol, bromobenzene, allyl alcohol and phalloidin, a significant increase was seen in at least one of the parameters AST, ALT and SDH compared with non-fasted mice.¹

Carbon tetrachloride caused massive hepatic centrilobular necroses in 24 h fasted mice, where non-fasted mice only showed moderate toxic alterations. Fasting did not influence enzyme concentrations after saline, alpha-amanitin and praseodymium treatment. In vitro protein binding of both carbon tetrachloride and paracetamol to hepatic microsomal protein remained unaffected upon 24 h of fasting. Total hepatic amount of carbon tetrachloride was significantly (P < 0.05) increased (about 30%) in 24 h fasted animals. Regarding paracetamol, fasting had no influence on the total hepatic concentration or the amount covalently bound to hepatic proteins. Administration of carbon tetrachloride produced a similar increase in lipid peroxidation in 24 h fasted and non-fasted mice.¹

Ethylmorphine, p-nitroanisole and aniline are drugs which undergo oxidative enzymatic degradation by hepatic microsomes. The metabolism of ethylmorphine and p-nitroanisole was significantly elevated after two and three days of fasting, but not after one day. Aniline metabolism was significantly elevated after one, two and three days of fasting.⁷⁰

Mice fasted for 12 h overnight prior to oral administration of paracetamol (e.g. acetaminophen), a drug that can cause hypothermia, liver necrosis and hepatic congestion in mice, produced a significantly (P < 0.01) higher covalent binding of paracetamol or its metabolite to liver protein 3 h after administration, but not 0.75, 1.5 or 6 h after administration. The fasted mice had a higher prevalence of hepatic centrilobular congestion, congestion and necrosis 1.5, 3 and 6 h after administration. Mortality (20%) only occurred in fasted animals, 4–6 h after administration. Fasted mice developed a significant (P < 0.001) increased liver weight/body weight ratio after paracetamol treatment compared with non-fasted mice.⁸²

In a study examining intravenously administered digitoxigenin toxicity mortality was significantly (P < 0.05) lower in 18 h fasted mice than in non-fasted mice. The distribution of the drug was also altered by fasting: In the cortex cerebrum of fasted mice there was significantly (P < 0.05) less drug content, cerebellum, heart and kidney did not differ, and in the liver and blood the drug content was significantly higher. Although these organs also lost weight, the relative concentration was higher than the weight loss. The fasted mice may benefit from the fact that they were dosed after body weight, and they had a lower body weight. In an attempt to correct this difference, a significant difference in the toxicity between non-fasted and fasted mice were still observed.⁶⁹

In conclusion, fasting is capable of influencing the toxicity of drugs, either by increasing or decreasing toxicity.

Discussion

Fasting of mice is often used for evaluation of blood glucose and insulin in association with metabolic diseases, such as diabetes. Glucose tolerance tests, insulin tolerance tests as well as hyperglycaemic, hyperinsulinaemic–euglycaemic and hyperinsulinaemic–hypoglycaemic clamps include commonly used tests in which a fasting duration of 5–6 h has been recommended. Fasting for 5–6 h instead of overnight (16–18 h) might offer a better comparison to humans, who are usually fasted overnight before similar tests. Blood glucose and insulin levels in humans after an overnight fast represent ‘a basal steady-state’, whereas overnight fasting of mice cannot be considered a routine event, due to their nocturnal eating pattern.⁹⁵

Fasting of mice changes several physiological parameters and moreover, these changes are dependent on a variety of environmental factors.

Glucose, insulin and corticosterone among others, all fluctuate with circadian rhythms in mice. To compare these parameters between mice in the same study or between studies, it is necessary to have information about light/dark cycles (durations and timing of scotophase and photophase), and exactly when the samples are collected. If there is lack of information, reported differences might be completely or partially caused by normal fluctuations in circadian rhythm, or actual differences could be masked.

When animals are fasted in a study, it is necessary to report when the animals were fasted and for how long, otherwise comparison is difficult or almost impossible. Since the normal temperature in an animal room is 20–24℃, mice are housed under less than their thermoneutral temperature of approximately 30℃.^96–98 Ambient temperature affects the food intake of mice, and the food intake affects other aspects of the metabolism, hence knowledge about ambient temperature is also important.

Some parameters oscillate during fasting, therefore, it cannot be assumed that, for example, an increase from one point to another is equal to a gradual increase.⁵⁶

It is complex to assess whether animal welfare is influenced by fasting. Duration and timing of the fast are important factors. It seems difficult to prove when the mouse recognizes that it is fasting, since this is probably individualistic and associated with habits and external factors. The longest voluntary inter-meal interval reported in this review was of 5 h.²⁰ This offers an idea of when the mouse starts to experience that it is being fasted, but this interval was registered in the photophase, and cannot be used as an argument for fasting mice for 5 h during scotophase. Secondly, strain differences must be expected. It has been stated that many wild mice only feed and forage in the night, or that they often encounter food restriction or deprivation.^33,99 If fasting is considered a part of the mouse’s daily cycle, one can argue that the changes resulting from fasting can be considered to be physiologically normal. This can only be claimed for fasting in the photophase, there is no evidence that mice would omit to eat in the scotophase. Experiments are mostly done in the photophase, which means that the mice will usually be fasting during scotophase because it is easy in practice to remove the food the afternoon before, except when the fasting period is only for a couple of hours.

Mice held under an opposite light/dark cycle can easily be fasted during photophase. Keeping the mice on an opposite light/dark cycle could also provide advantages regarding the fact that cleaning, feeding and activities in general would be done during the activity period of the mice. Disadvantages include trouble and discomfort working in the dark: in cases of large animal rooms everybody needs to agree upon having opposite light/dark cycles and extra time spent when mice have to adapt to the opposite light/dark cycle; for example corticosterone levels take about six days to normalize.⁴² If the mice have to be transferred to a lightened room for experimental work, and then put back into the scotophase, it must be considered whether these changes between light and dark may be more disturbing than being disturbed in the photophase. A previous investigation assessed that up to 24 h of fasting is minimally stressful, but reported that data on mice are lacking.¹⁰⁰

Feeding sucrose cubes to rats and mice has been presented as an alternative to fasting them. This feeding can produce marked changes in gastrointestinal tract weight and pancreatic and hepatic structure and function. Using sucrose feeding for toxicology studies or any studies examining hepatic function is therefore not recommended.¹⁰¹ It is obviously not a valid alternative if the purpose of fasting is to evaluate blood glucose level.

Placing the mice on a grid floor during fasting to avoid intake of bedding material has been recommended.¹⁰² It also precludes the mice from coprophagia. The animal welfare implications of changing the environment by placing the mice on grid floor must be assessed as well as the consequences of the possibility of mice eating bedding or faeces. In a study by Llanos and Nash, gastric content after fasting for 12 h was no different from gastric content of non-fasted mice at a particular time of day (at the end of photophase), and the authors discussed that fasting is not always necessary.³⁷ The feeding activity of the mice was restricted to the scotophase, thus these results may not be applicable for mice eating in the photophase as well. There was no description of how the feeding activity was measured. To optimize animal welfare, an alternative to feed the mice a hydrated diet, which increases the gastric emptying rate has been provided.⁷³

There is little evidence of how the mouse behaves during a fast, but stress and aggression requiring single-housing have been mentioned.⁷¹ One study showed that fasted mice increased their general activity significantly compared with non-fasted mice and discussed that the increase might be due to motivation to find food.⁹⁹ Another study found increased activity of fasted mice during scotophase, and activity levels similar to non-fasted mice during photophase, while other investigators found photophase hyperactivity during a 24 h fast.^34,84 Williams et al. reported activity during fasting to be increased, but not significantly increased, during photophase.⁵³ It has been suggested that an increase in locomotor activity associated with fasting can be interpreted as food foraging behaviour or exercise-induced thermogenesis.¹⁰³

Conclusion

The implications of fasting are comprehensive (Table 6). There is reason to investigate whether fasting compromises animal welfare or produces any variability or significant changes in the planned scientific outcome due to the effects of fasting, before using fasting as a routine procedure. Fasting leads to: body weight loss, significant changes in multiple cardiovascular, hormonal and metabolic parameters and changes in hepatic metabolism, and the possibility of mice entering torpor after about 7 h of fasting. In many instances it is not known when the significant changes occur soonest. Fasting for 18 h (overnight) is a standard procedure.^{56,71,73,74,104} Some studies show that gastric content is reduced as much after 4–6 h of fasting as after 12 or 18 h.^37,71 Some mice died after fasting for two days and nights, whereas others can survive for up to five days and nights.^16,53,93 This might be associated with strain differences and differences in ability to save calories by entering torpor. It seems appropriate, in regard to both animal welfare and the validity of a study, to apply the shortest possible fast, and preferably to fast the mice in the photophase, to comply with the behaviour of the species. Hence further research should seek to establish the shortest fasting time needed to obtain a stable-state blood glucose in mice, including the possible effect of time of day for the fasting. Moreover, the fluctuations in gastric content during fasting should be assessed to establish whether fasting of mice to obtain an empty stomach is at all possible without having to muzzle the animals.

Table 6.

Effects of fasting.

Parameter	Effect	References
Adipose tissue	↓	Bronson 1987;³³ Ayala et al. 2006²¹
Aminopyrine N-demethylase activity	↑	Strubelt et al. 1981¹
Aniline hydroxylase activity	↑	Strubelt et al. 1981¹
Aryl hydrocarbon hydroxylase activity	↔	Walker et al. 1982⁸²
Blood glucagon	↑	Moitra et al. 1998;⁶⁵ Kersten et al. 1999⁷⁹
Body temperature	↓	Hudson & Scott 1979;⁸⁷ Webb et al. 1980;⁸⁸ Webb et al. 1982;⁸⁶ Kanizsai et al. 2009;⁸⁴ Brown & Staples 2010;⁸⁹ Wankhade et al. 2010⁶¹
Blood volume	↓	Hodge & MacLachlan 1947⁹³
Body weight	↓	Furner & Feller 1971;⁷⁰ Strubelt et al. 1981;¹ Warwick et al. 1985;⁶⁹ Williams et al. 2002;⁵³ Ayala et al. 2006;²¹ Sokolović et al. 2007;⁶⁸ Prior et al. 2009⁷¹
Brain weight	↔	Warwick et al. 1985⁶⁹
Carnitine	↑	Hashimoto et al. 2000⁷⁷
Corticosterone	↑	Ahima et al. 1996;⁸ Gavrilova et al. 1999;⁵⁷ Champy et al. 2004⁵⁶
Cytochrome P-450	↕	Strubelt et al. 1981;¹ Walker et al. 1982⁸²
Gastric content	↓	Llanos & Nash 1971;³⁷ Pessacq et al. 1976;³⁸ Yeung et al. 2001;⁷³ Osinski et al. 2002;⁷⁴ Prior et al. 2009⁷¹
Gastric emptying	↑	Watanabe et al. 1977⁷⁵
Gastrointestinal tract weight	↓	Warwick et al. 1985⁶⁹
Ghrelin	↑	Asakawa et al. 2001;⁶³ Kojima & Kangawa 2005⁶²
Glucose, blood	↓	Pessacq et al. 1976;³⁸ Hashimoto et al. 2000;⁷⁷ Tsuneki et al. 2002;⁷⁸ Champy et al. 2004;⁵⁶ Oosterveer et al. 2008⁴⁶
Glucuronyl transferase activity	↑	Walker et al. 1982⁸²
Heart rate	↓	Williams et al. 2002⁵³
Heart triglyceride	↔	Hashimoto et al. 2000⁷⁷
Heart weight	↓	Warwick et al. 1985⁶⁹
Haematocrit	↓	Champy et al. 2004⁵⁶
Haemoglobin	↑	Champy et al. 2004⁵⁶
Hepatic alanine transaminase	↔	Strubelt et al. 1981;¹ Champy et al. 2004⁵⁶
Hepatic aspartate transaminase	↔	Strubelt et al. 1981;¹ Champy et al. 2004⁵⁶
Hepatic glycogen	↓	Strubelt et al. 1981;¹ Hashimoto et al. 2000;⁷⁷ Yeung et al. 2001;⁷³ Ayala et al. 2006²¹
Hepatic sorbitol dehydrogenase	↔	Strubelt et al. 1981¹
Hepatic triglyceride	↑	Strubelt et al. 1981;¹ Hashimoto et al. 2000;⁷⁷ Heijboer et al. 2005;⁶⁶ Lin et al. 2005⁸⁰
Hepatic water content	↔	Strubelt et al. 1981¹
Intestinal transit time of charcoal	↓	Prior et al. 2009⁷¹
Kidney triglyceride	↔	Hashimoto et al. 2000⁷⁷
Kidney weight	↓	Warwick et al. 1985⁶⁹
Leptin	↓	Rayner et al. 1998;⁵⁸ Trayhurn et al. 1998;⁵⁹ Champy et al. 2004;⁵⁶ Swoap et al. 2006;⁶⁰ Wankhade et al. 2010⁶¹
Liver weight	↓	Furner & Feller 1971;⁷⁰ Strubelt et al. 1981;¹ Warwick et al. 1985⁶⁹
Liver weight/100 g body weight ratio	↓	Furner & Feller 1971⁷⁰
Liver to body weight ratio	↔	Strubelt et al. 1981¹
Mean arterial pressure	↓	Williams et al. 2002⁵³
Mean cell haemoglobin conc.	↔	Champy et al. 2004⁵⁶
Mean cell volume	↑	Champy et al. 2004⁵⁶
Mean corpuscular haemoglobin	↔	Champy et al. 2004⁵⁶
Mitochondrial HMG-CoA synthase	↑	Bauer et al. 2004⁸¹
Muscle glycogen	↓	Gavrilova et al. 1999⁵⁷
Muscle specific glucose uptake	↑	Heijboer et al. 2005⁶⁶
Muscle triglyceride	↔	Heijboer et al. 2005⁶⁶
NADPH-cytochrome c-reductase activity	↔	Strubelt et al. 1981¹
Neuropeptide Y	↑	Ahima et al. 1996⁸
Oxygen consumption	↓	Williams et al. 2002⁵³
Plasma β-hydroxybutyrate	↑	Moitra et al. 1998;⁶⁵ Kersten et al. 1999;⁷⁹ Gavrilova et al. 1999;⁵⁷ Hashimoto et al. 2000;⁷⁷ Lin et al. 2005⁸⁰
Plasma free fatty acids	↑	Moitra et al. 1998;⁶⁵ Kersten et al. 1999;⁷⁹ Hashimoto et al. 2000;⁷⁷ Lin et al. 2005⁸⁰
Plasma insulin	↓	Moitra et al. 1998;⁶⁵ Kersten et al. 1999;⁷⁹ Champy et al. 2004⁵⁶
Plasma ketone bodies	↑	Heijboer et al. 2005⁶⁶
Plasma lactate	↓	Hashimoto et al. 2000⁷⁷
Plasma (nor) epinephrine	↑	Moitra et al. 1998;⁶⁵ Kersten et al. 1999⁷⁹
Plasma triglyceride	↓	Hashimoto et al. 2000;⁷⁷ Champy et al. 2004⁵⁶
Platelets	↑	Champy et al. 2004⁵⁶
Red blood cells	↑	Champy et al. 2004⁵⁶
Resistin	↑	Steppan et al. 2001⁶⁴
Testis weight	↔	Gavrilova et al. 1999⁵⁷
Testosterone	↓	Ahima et al. 1996;⁸ Gavrilova et al. 1999;⁵⁷ Champy et al. 2004⁵⁶
Thyroid stimulating hormone	↔	Gavrilova et al. 1999⁵⁷
Thyroxine	↓	Ahima et al. 1996;⁸ Gavrilova et al. 1999;⁵⁷ Champy et al. 2004⁵⁶
Total hepatic glucose production	↔	Heijboer et al. 2005⁶⁶
Total protein content of liver microsomes	↔	Strubelt et al. 1981;¹ Walker et al. 1982⁸²
Triiodothyronine	↓	Gavrilova et al. 1999;⁵⁷ Champy et al. 2004⁵⁶
Urea	↔	Gavrilova et al. 1999;⁵⁷ Champy et al. 2004⁵⁶
Water intake	↕	Kutscher 1971;¹⁶ Kutscher 1974;⁵² Williams et al. 2002⁵³
White blood cells	↑	Champy et al. 2004⁵⁶

The arrows indicate whether the parameters increase, decrease, do both or stay level during fasting.

↓

= decrease, ↑ = increase, ↕ = decrease and increase, ↔ = no change. See references for information about duration of fasting. Genetics and toxicology are excluded.

Footnotes

Declaration of Conflicting Interests

The authors declare no conflict of interest.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Strubelt

Dost-Kempf

Siegers

. The influence of fasting on the susceptibility of mice to hepatotoxic injury. Toxicol Appl Pharmacol 1981; 60: 66–77.

Hedric

Bullock

. The laboratory mouse, Amsterdam: Elsevier Academic Press, 2004.

Ebert

. Automation of rodent fasting. J Am Assoc Lab Anim Sci 2009; 48: 537–538.

Kinoshita

Igarashi

Kume

Saito

Arakawa

. Fasting induces impairment of gastric mucosal integrity in non-insulin-dependent diabetic (db/db) mice. Aliment Pharmacol Ther 2000; 14: 359–366.

Kim

Hyun

Choung

. Anti-diabetic effect of cinnamon extract on blood glucose in db/db mice. J Ethnopharmacol 2006; 104: 119–123.

Organization for Economic Cooperation and Development (OECD). Guidelines for testing of chemicals. No. 407–409, Section 4: health effects, Paris: OECD, 1995.

Ellacott

KLJ

Morton

Woods

Tso

Schwartz

. Assessment of feeding behavior in laboratory mice. Cell Metab 2010; 12: 10–17.

Ahima

Prabakaran

Mantzoros

. Role of leptin in the neuroendocrine response to fasting. Nature 1996; 382: 250–252.

Froy

Miskin

. Interrelations among feeding, circadian rhythms and ageing. Prog Neurobiol 2007; 82: 142–150.

10.

Noguchi

Michihata

Nakamura

. Dual-color luciferase mouse directly demonstrates coupled expression of two clock genes. Biochemistry 2010; 49: 8053–8061.

11.

Reppert

Weaver

. Coordination of circadian timing in mammals. Nature 2002; 418: 935–941.

12.

Strubbe

Prins

AJA

Bruggink

Steffens

. Daily variation of food-induced changes in blood glucose and insulin in the rat and the control by the suprachiasmatic nucleus and the vagus nerve. J Auton Nerv Syst 1987; 20: 113–119.

13.

Szentirmai

Kapas

Sun

Smith

Krueger

. Restricted feeding-induced sleep, activity, and body temperature changes in normal and preproghrelin-deficient mice. Am J Physiol Regul Integr Comp Physiol 2010; 298: R467–R477.

14.

Dallman

Akana

Strack

Hanson

Sebastian

. The neural network that regulates energy balance is responsive to glucocorticoids and insulin and also regulates HPA axis responsivity at a site proximal to CRF neurons. Ann N Y Acad Sci 1995; 771: 730–742.

15.

Djordjevic

Jasnic

Vujovic

Djurasevic

Djordjevic

Cvijic

. The effect of fasting on the diurnal rhythm of rat ACTH and corticosterone secretion. Arch Biol Sci 2008; 60: 541–546.

16.

Kutscher

. Incidence of food-deprivation polydipsia in white Swiss mouse. Physiol Behav 1971; 7: 395–399.

17.

Kavaliers

Hirst

. The influence of opiate agonists on day–night feeding rhythms in young and old mice. Brain Res 1985; 326: 160–167.

18.

Gannon

Smith

Henderson

Hendrick

. A system for studying the microstructure of ingestive behavior in mice. Physiol Behav 1992; 51: 515–521.

19.

Schlingmann

Van de Weerd

Baumans

Remie

Van Zutphen

LFM

. A balance device for the analysis of behavioural patterns of the mouse. Anim Welf 1998; 7: 177–188.

20.

Kurokawa

Akino

Kanda

. A new apparatus for studying feeding and drinking in the mouse. Physiol Behav 2000; 70: 105–112.

21.

Ayala

Bracy

McGuinness

Wasserman

. Considerations in the design of hyperinsulinemic–euglycemic clamps in the conscious mouse. Diabetes 2006; 55: 390–397.

22.

Petersen

. Feeding, blood glucose and plasma insulin of mice at dusk. Nature 1978; 275: 647–649.

23.

Wiepkema

Deruiter

Reddingh

. Circadian rhythms in feeding behaviour of CBA mice. Nature 1966; 209: 935–936.

24.

Nowell

. Circadian rhythm of glucose tolerance in laboratory mice. Diabetologia 1970; 6: 488–492.

25.

Mühlbauer

Gross

Labucay

Wolgast

Peschke

. Loss of melatonin signalling and its impact on circadian rhythms in mouse organs regulating blood glucose. Eur J Pharmacol 2009; 606: 61–71.

26.

Gagliardino

Hernández

. Circadian variation of the serum glucose and immunoreactive insulin levels. Endocrinology 1971; 88: 1532–1534.

27.

Tucci

Hardy

Nolan

. A comparison of physiological and behavioural parameters in C57BL/6J mice undergoing food or water restriction regimes. Behav Brain Res 2006; 173: 22–29.

28.

Echavell

Epeledev

. Circadian rhythm in blood glucose levels in normal and hepatectomized mice. J Interdiscipl Cycle 1972; 3: 25–32.

29.

Ahrén

. Diurnal variation in circulating leptin is dependent on gender, food intake and circulating insulin in mice. Acta Physiol Scand 2000; 169: 325–331.

30.

Bailey

Atkins

Conner

Manley

Matty

. Diurnal variations of food consumption, plasma glucose and plasma insulin concentrations in lean and obese hyperglycemic mice. Horm Res 1975; 6: 380–386.

31.

South

Huang

. Inter-meal interval is increased in mice fed a high whey, as opposed to soy and gluten, protein diets. Appetite 2009; 52: 372–379.

32.

Boss

Samec

Kuhne

. Uncoupling protein-3 expression in rodent skeletal muscle is modulated by food intake but not by changes in environmental temperature. J Biol Chem 1998; 273: 5–8.

33.

Bronson

. Susceptibility of the fat reserves of mice to natural challenges. J Comp Physiol B 1987; 157: 551–554.

34.

Overton

Williams

. Behavioral and physiologic responses to caloric restriction in mice. Physiol Behav 2004; 81: 749–754.

35.

Uchida

Shiuchi

Inada

Minokoshi

Tominaga

. Metabolic adaptation of mice in a cool environment. Pflugers Arch 2010; 459: 765–774.

36.

Bachmanov

Reed

Beauchamp

Tordoff

. Food intake, water intake, and drinking spout side preference of 28 mouse strains. Behav Genet 2002; 32: 435–443.

37.

Llanos

JME

Nash

. Circadian rhythm in the gastric content of normal and hepatectomized mice. J Interdiscipl Cycle 1971; 2: 445–454.

38.

Pessacq

Mercer

Rebolledo

Gagliardino

. Effect of fasting on circadian-rhythm of serum insulin levels. Chronobiologia 1976; 3: 20–26.

39.

Ungar

Halberg

. Circadian rhythm in in vitro response of mouse adrenal to adrenocorticotropic hormone. Science 1962; 137: 1058–1060.

40.

Halberg

Peterson

Silber

. Phase relations of 24-hour periodicities in blood corticosterone, mitoses in cortical adrenal parenchyma, and total body activity. Endocrinology 1959; 64: 222–230.

41.

Shimizu

Amagaya

Ogihara

. Analysis of corticosterone in the serum of mice and rats using high performance liquid chromatography. J Chromatogr 1983; 272: 170–175.

42.

Hayashi

Kikuchi

. The influence of phase shift in the light–dark cycle on humoral immune responses of mice to sheep red blood cells and polyvinylpyrrolidone. J Immunol 1985; 134: 1455–1461.

43.

Harris

Gunion

Rosenthal

Walford

. Serum glucose, glucose tolerance, corticosterone and free fatty acids during aging in energy restricted mice. Mech Ageing Dev 1994; 73: 209–221.

44.

Weinert

Eimert

Erkert

Schneyer

. Resynchronization of the circadian corticosterone rhythm after a light–dark shift in juvenile and adult mice. Chronobiol Int 1994; 11: 222–231.

45.

Barriga

Martin

Tabla

Ortega

Rodriguez

. Circadian rhythm of melatonin, corticosterone and phagocytosis: effect of stress. J Pineal Res 2001; 30: 180–187.

46.

Oosterveer

Van Dijk

Grefhorst

. LXR alpha deficiency hampers the hepatic adaptive response to fasting in mice. J Biol Chem 2008; 283: 25437–25445.

47.

Awad

Constantin-Teodosiu

Macdonald

Lobo

. Short-term starvation and mitochondrial dysfunction – a possible mechanism leading to postoperative insulin resistance. Clin Nutr 2009; 28: 497–509.

48.

Connolly

Lynch

. Circadian variation of strain differences in body temperature and activity in mice. Physiol Behav 1981; 27: 1045–1049.

49.

Lacy

Lynch

. Quantitative genetic-analysis of temperature regulation in Mus musculus .1. Partitioning of variance. Genetics 1979; 91: 743–753.

50.

Nelson

Halberg

. Meal-timing, circadian rhythms and life span of mice. J Nutr 1986; 116: 2244–2253.

51.

Swoap

. Altered leptin signaling is sufficient, but not required, for hypotension associated with caloric restriction. Am J Physiol Heart Circ Physiol 2001; 281: H2473–H2479.

52.

Kutscher

. Strain differences in drinking in inbred mice during ad libitum feeding and food deprivation. Physiol Behav 1974; 13: 63–70.

53.

Williams

Chambers

Henderson

Rashotte

Overton

. Cardiovascular responses to caloric restriction and thermoneutrality in C57BL/6J mice. Am J Physiol Regul Integr Comp Physiol 2002; 282: R1459–R1467.

54.

Amaral

VCS

Gomes

Nunes-de-Souza

. Increased corticosterone levels in mice subjected to the rat exposure test. Horm Behav 2010; 57: 128–133.

55.

Rayner

. The sympathetic nervous system in white adipose tissue regulation. Proc Nutr Soc 2001; 60: 357–364.

56.

Champy

Selloum

Piard

. Mouse functional genomics requires standardization of mouse handling and housing conditions. Mamm Genome 2004; 15: 768–783.

57.

Gavrilova

Leon

Marcus-Samuels

. Torpor in mice is induced by both leptin-dependent and -independent mechanisms. Proc Natl Acad Sci USA 1999; 96: 14623–14628.

58.

Rayner

Simon

Duncan

Trayhurn

. Hyperleptinaemia in mice induced by administration of the tyrosine hydroxylase inhibitor alpha-methyl-p-tyrosine. FEBS Lett 1998; 429: 395–398.

59.

Trayhurn

Duncan

Hoggard

Rayner

. Regulation of leptin production: a dominant role for the sympathetic nervous system? Proc Nutr Soc 1998; 57: 413–419.

60.

Swoap

Gutilla

Liles

Smith

Weinshenker

. The full expression of fasting-induced torpor requires beta 3-adrenergic receptor signaling. J Neurosci 2006; 26: 241–245.

61.

Wankhade

Vella

Fox

Good

. Deletion of Nhlh2 results in a defective torpor response and reduced beta adrenergic receptor expression in adipose tissue. PLoS One 2010; 5: e12324–e12324.

62.

Kojima

Kangawa

. Ghrelin: structure and function. Physiol Rev 2005; 85: 495–522.

63.

Asakawa

Inui

Kaga

. Ghrelin is an appetite-stimulatory signal from stomach with structural resemblance to motilin. Gastroenterology 2001; 120: 337–345.

64.

Steppan

Bailey

Bhat

. The hormone resistin links obesity to diabetes. Nature 2001; 409: 307–312.

65.

Moitra

Mason

Olive

. Life without white fat: a transgenic mouse. Genes Dev 1998; 12: 3168–3181.

66.

Heijboer

Donga

Voshol

. Sixteen hours of fasting differentially affects hepatic and muscle insulin sensitivity in mice. J Lipid Res 2005; 46: 582–588.

67.

Quimby

. The clinical chemistry of laboratory animals, Philadelphia: Taylor & Francis, 1999.

68.

Sokolović

Wehkamp

Sokolović

. Fasting induces a biphasic adaptive metabolic response in murine small intestine. BMC Genomics 2007; 8: 361–361.

69.

Warwick

Smith

Graffy

Lage

. Altered distribution and toxicity of digitoxgenin in fasted mice. Life Sci 1985; 37: 775–782.

70.

Furner

Feller

. Influence of starvation upon hepatic drug metabolism in rats, mice, and guinea pigs. Proc Soc Exp Biol Med 1971; 137: 816–819.

71.

Prior

Ewart

Fantham

Pimlott

Cowell

Valentin

. Refinement of the charcoal meal GI study: reduction in the fasting period required. J Pharmacol Toxicol Methods 2009; 58: 174–174.

72.

Mano

Matsuura

Nakatani

Shimizu

Wada

. Mechanisms of blood glucose-lowering effect of aqueous extract from stems of Kothala himbutu (Salacia reticulata) in the mouse. J Ethnopharmacol 2009; 121: 234–240.

73.

Yeung

McCurrie

Wood

. A simple method to investigate the inhibitory effects of drugs on gastric emptying in the mouse in vivo. J Pharmacol Toxicol Methods 2001; 45: 235–240.

74.

Osinski

Seifert

Cox

Gintant

. An improved method of evaluation of drug-evoked changes in gastric emptying in mice. J Pharmacol Toxicol Methods 2002; 47: 115–120.

75.

Watanabe

Okabe

Ichihashi

Mizojiri

Yamada

Yamamoto

. Gastric emptying rate constants after oral administration of drug solution to mice, rats, and rabbits. Chem Pharm Bull 1977; 25: 2147–2155.

76.

Lin

. Species similarities and differences in pharmacokinetics. Drug Metab Dispos 1995; 23: 1008–1021.

77.

Hashimoto

Cook

Yeldandi

Reddy

Rao

. Defect in peroxisome proliferator-activated receptor alpha-inducible fatty acid oxidation determines the severity of hepatic steatosis in response to fasting. J Biol Chem 2000; 275: 28918–28928.

78.

Tsuneki

Sugihara

Honda

Wada

Sasaoka

Kimura

. Reduction of blood glucose level by orexins in fasting normal and streptozotocin-diabetic mice. Eur J Pharmacol 2002; 448: 245–252.

79.

Kersten

Seydoux

Peters

Gonzalez

Desvergne

Wrahli

. Peroxisome proliferator-activated receptor alpha mediates the adaptive response to fasting. J Clin Invest 1999; 103: 1489–1498.

80.

Lin

Yue

Chen

Schonfeld

. Hepatic triglyceride contents are genetically determined in mice: results of a strain survey. Am J Physiol Gastrointest Liver Physiol 2005; 288: G1179–G1189.

81.

Bauer

Hamm

Bonaus

. Starvation response in mouse liver shows strong correlation with life-span-prolonging processes. Physiol Genomics 2004; 17: 230–244.

82.

Walker

Massey

Mcelligott

Racz

. Acetaminophen toxicity in fed and fasted mice. Can J Physiol Pharmacol 1982; 60: 399–404.

83.

Brooks

Pong

. Effects of fasting, body weight, methylcellulose, and carboxymethylcellulose on hepatic glutathione levels in mice and hamsters. Biochem Pharmacol 1981; 30: 589–594.

84.

Kanizsai

Garami

Solymar

Szolcsanyi

Szelenyi

. Energetics of fasting heterothermia in TRPV1-KO and wild type mice. Physiol Behav 2009; 96: 149–154.

85.

Hudson

. Shallow, daily torpor – thermoregulatory adaptation. J Therm Biol 1978; 3: 87–87.

86.

Webb

Jagot

Jakobson

. Fasting-induced torpor in Mus musculus and its implications in the use of murine models for human obesity studies. Comp Biochem Physiol A Comp Physiol 1982; 72: 211–219.

87.

Hudson

Scott

. Daily torpor in the laboratory mouse, Mus musculus var albino. Physiol Zool 1979; 52: 205–218.

88.

Webb

Jagot

Rogers

Jakobson

. Effects of fasting on thermoregulation in normal and obese mice. IRCS Med Sci Biochem 1980; 8: 163–164.

89.

Brown

JCL

Staples

. Mitochondrial metabolism during fasting-induced daily torpor in mice. Biochim Biophys Acta 2010; 1797: 476–486.

90.

Swoap

Weinshenker

. Norepinephrine controls both torpor initiation and emergence via distinct mechanisms in the mouse. PLoS One 2008; 3: e4038–e4038.

91.

Gluck

Stephens

Swoap

. Peripheral ghrelin deepens torpor bouts in mice through the arcuate nucleus neuropeptide Y signaling pathway. Am J Physiol Regul Integr Comp Physiol 2006; 291: R1303–R1309.

92.

Morton

. Torpor and nest-sharing in free-living Sminthopsis crassicaudata (Marsupialia) and Mus musculus (Rodentia). J Mammal 1978; 59: 569–575.

93.

Hodge

MacLachlan

. Lipides of the fasting mouse; phospholipide to neutral fat relationship in the blood lipides. J Biol Chem 1947; 169: 707–711.

94.

Kast

Nishikawa

. The effect of fasting on oral acute toxicity of drugs in rats and mice. Lab Anim 1981; 15: 359–364.

95.

Ayala

Samuel

Morton

. Standard operating procedures for describing and performing metabolic tests of glucose homeostasis in mice. Dis Model Mech 2010; 3: 525–534.

96.

McGuinness

Ayala

Laughlin

Wasserman

. NIH experiment in centralized mouse phenotyping: the Vanderbilt experience and recommendations for evaluating glucose homeostasis in the mouse. Am J Physiol Endocrinol Metab 2009; 297: E849–E855.

97.

Gaskill

Rohr

Pajor

Lucas

Garner

. Some like it hot: mouse temperature preferences in laboratory housing. Appl Anim Behav Sci 2009; 116: 279–285.

98.

Gaskill

Rohr

Pajor

Lucas

Garner

. Working with what you've got: changes in thermal preference and behavior in mice with or without nesting material. J Therm Biol 2011; 36: 193–199.

99.

Cao

Zhang

L-N

Zhao

Z-J

. Trade-off between energy budget, thermogenesis and behavior in Swiss mice under stochastic food deprivation. J Therm Biol 2009; 34: 290–298.

100.

Rowland

. Food or fluid restriction in common laboratory animals: balancing welfare considerations with scientific inquiry. Comp Med 2007; 57: 149–160.

101.

Turner

Albassam

Walker

. The effects of overnight fasting, feeding, or sucrose supplementation prior to necropsy in rats. Contemp Top Lab Anim Sci 2001; 40: 36–40.

102.

Hau

Van Hoosier

Jr . Handbook of Laboratory animal Science, 2nd ed. Boca Raton: CRC Press, 2003.

103.

Gelegen

Collier

Campbell

Oppelaar

Kas

. Behavioral, physiological, and molecular differences in response to dietary restriction in three inbred mouse strains. Am J Physiol Endocrinol Metab 2006; 291: E574–E581.

104.

Roda

Mezzanotte

Aldini

Michelini

Cevenini

. A new gastric-emptying mouse model based on in vivo non-invasive bioluminescence imaging. Neurogastroenterol Motil 2010; 22: 1117–e288.

Fasting of mice: a review

Abstract

Keywords

Physiology of non-fasted mice

Food and water intake

Hormones

Body temperature

Fasting of mice: physiological changes

Non-food and water intake

Hormones

Weight loss

Gastrointestinal emptying

Metabolism

Liver parameters

Body temperature

Cardiovascular parameters

Toxicology

Discussion

Conclusion

Footnotes

Declaration of Conflicting Interests

Funding

References