Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Leptin suppresses semi-starvation induced hyperactivity in rats: implications for anorexia nervosa


Semi-starvation induced hyperactivity (SIH) occurs in rodents upon caloric restriction. We hypothesized that SIH is triggered by the decline in leptin secretion associated with food restriction. To test this hypothesis, rats, which had established a stable level of activity, were treated with leptin or vehicle via implanted minipumps concomitantly to initiation of food restriction for 7 days. In a second experiment treatment was initiated after SIH had already set in. In contrast to the vehicle-treated rats, which increased their baseline activity level by 300%, the development of SIH was suppressed by leptin. Furthermore, leptin was able to stop SIH, after it had set in. These results underscore the assumed major role of leptin in the adaptation to semi-starvation. Because SIH has been viewed as a model for anorexia nervosa, we also assessed subjective ratings of motor restlessness in 30 patients with this eating disorder in the emaciated state associated with hypoleptinemia and after increments in leptin secretion brought upon by therapeutically induced weight gain. Hypoleptinemic patients ranked their motor restlessness higher than upon attainment of their maximal leptin level during inpatient treatment. Thus, hypoleptinemia might also contribute to the hyperactivity frequently associated with anorexia nervosa.


Hyperactivity is observed in 40–80% of patients with anorexia nervosa (AN).1, 2 Caloric deprivation severe enough to result in significant weight loss possibly provokes sensations of behavioral arousal and activation in individuals with an innate vulnerability to develop AN.3 Indeed, reduced food intake and enhanced activity have been viewed as the core symptomatology in AN, because only these behavioral measures consistently distinguish AN from other disorders.4

Semi-starvation induced hyperactivity (SIH) is a well characterized phenomenon in laboratory animals.5, 6 Rats supplied with food for only 1 h per day manage to survive, but die within a short period of time when exposed to a running wheel. The enhancement of activity is related to the severity of food restriction; total food deprivation results in a disruption of the nocturnal activity pattern.6

Leptin is assumed to be the major hormone underlying the regulatory phenomena that lead to an adaptation of an organism to reduced energy supplies.7 Thus, in mice exogenously applied leptin has been shown to blunt the semi-starvation-induced downregulation of the hypothalamic-pituitary gonadal and thyroid axes as well as the upregulation of the hypothalamic-pituitary adrenal axis.7 The rapid decline in leptin secretion associated with caloric restriction and weight loss7, 8 could represent the initial trigger underlying the hyperactivity observed in SIH and AN. As an initial step towards testing this hypothesis we investigated the effect of continuous leptin treatment via minipumps on SIH in rats. In parallel, we describe the subjective experience of motor restlessness in 30 patients with AN, both during initial hypoleptinemia and after weight gain at times of maximal leptin secretion during inpatient treatment.

In the first series of animal studies, a total of 30 rats which had previously established a stable activity level, were subjected to a 7-day long continuous infusion of either leptin (n = 15) or vehicle (n = 15; controls) via subcutaneously implanted minipumps. After implantations an equal number of leptin-treated and control rats were allocated either to an ad libitum diet (n = 16) or a food restriction amounting to 60% of the previous ad libitum intake (n = 14), respectively. Implantations, irrespective of treatment, caused an intermittent decrease of activity (Figure 1a and b).

Figure 1

Suppression of semi-starvation induced hyperactivity (SIH) by leptin. Running wheel activity was recorded in rats, which were either restricted to 60% of their ad libitum food intake for 1 week (a, n = 7 in each treatment group) or had free access to food (b, n = 8 in each treatment group). Rats were treated with leptin (open symbols) or vehicle (closed symbols) via implanted minipumps beginning on the first day of food restriction. The duration of leptin application is indicated by the black bars and the period of food restriction is highlighted by the shaded arrow. The recording of running wheel activity started 2 weeks prior to treatment. Activity is presented as a percentage ± SEM of mean activity during the second pretreatment week, to adjust for individual variation in baseline activity. Asterisks denote a significant effect of leptin treatment on activity (P < 0.05).

Running wheel activity amongst the seven food-restricted controls bounced back to the pre-surgery level within 1 day (Figure 1a). During the following week food restriction continuously promoted the development of SIH up to a mean maximum of 4800 revolutions on day 7 corresponding to 300% of the presurgery baseline. In marked contrast and in accordance with our hypothesis, the activity of the seven food-restricted rats infused with leptin only gradually increased after the implantation, but even at day 7 did not significantly exceed the presurgery baseline. In ad libitum-fed animals, leptin treatment apparently facilitated the post-surgery recovery to baseline activity levels as compared to controls (Figure 1b), which might be related to the previously detected stress-reducing effect of leptin in rodents.9

Irrespective of mode of treatment, all food-restricted rats displayed a clear nocturnal pattern of activity with only 7–11% of total activity occurring during the light phase. In addition to the inhibitory effect on SIH, leptin treatment also altered the diurnal activity pattern during initial refeeding, during which minipumps no longer delivered leptin. Thus, in rats previously treated with leptin, daytime activity remained low, whereas in control animals the nocturnal activity pattern was disrupted during the first 3 days of refeeding, with daytime activity increasing by 27–44%. On the first day of refeeding, leptin-treated rats consumed less food as compared to controls (28.4 ± 1.4 g vs 34.7 ± 1.9 g; P < 0.05). Possibly the strong refeeding response in controls caused an altered circadian activity pattern, which was not observed in the leptin-treated animals, perhaps as a consequence of a lowered foraging drive.

In accordance with previous results in other rodents,10, 11 leptin treatment of ad libitum-fed rats caused reduced weight gain (5.8 ± 2.3 g per 7 days vs 23.6 ± 2.2 g per 7 days, P < 0.001) and reduced food intake (19.5 ± 0.78 g day−1 vs 22.37 ± 0.93 g day−1; P < 0.05) in comparison to controls. Surprisingly, despite the large difference in locomotor activity, hyperactive controls did not lose more body weight during semi-starvation for 1 week than leptin-treated rats (mean body weights ± SE at postsurgery day 7: 250.2 ± 8.2 g vs 252.4 ± 10.5 g). This could possibly be explained by findings in rodents that exogenous leptin blunts the semi-starvation-induced decrease of resting metabolic rate.12, 13

A second experiment was designed to test whether leptin treatment can ‘rescue’ hyperactive rats after SIH has set in. After food restriction for 5 days had induced SIH in all 14 rats, implantations of the pumps again led to a transient drop in activity. As expected, the seven control rats again increased activity during the following week of continued food restriction. In marked contrast, activity levels of the seven leptin-treated rats only slowly increased after implantations, but did not exceed the levels observed prior to food restriction (Figure 2). Measurements of serum leptin levels upon termination of the experiment 7 days after implantations revealed an approximately two-fold elevation in the leptin-treated rats as compared to controls (1406 ± 161 pg ml−1 vs 745 ± 151 pg ml−1, P < 0.05), thus indicating that minipumps were still releasing leptin.

Figure 2

Leptin inhibition of running wheel activity after semi-starvation induced hyperactivity (SIH) had set in. Rats were food restricted for 12 days to 60% of the ad libitum food intake, beginning 5 days prior to the start of either leptin (open symbols) or vehicle (closed symbols) treatment via implanted minipumps (n = 7 in each treatment group). The duration of leptin application is indicated by the black bar and the period of food restriction is highlighted by the shaded arrow. Activity is presented as a percentage ± SEM of mean activity during the week before the beginning of food restriction, to adjust for individual variation in baseline activity. Asterisks denote a significant effect of leptin treatment on activity (P < 0.05).

Serum leptin levels rapidly decrease in response to semi-starvation7, 8 and in addition, exercise in itself potentially may contribute to hypoleptinemia.14 Our results clearly suggest that the decrease of serum leptin levels in food-restricted rats is the major signal which triggers the expression of SIH, thus further extending the assumed key role of this hormone in semi-starvation7 to the behavioral level. A prolonged reduction of caloric intake in humans and rodents is accompanied by several adaptive mechanisms which counterregulate the development of negative energy balance. In this context, SIH appears to be a paradoxical reaction. However, if SIH leads to increased foraging behavior in a free-ranging animal, this may enhance fitness and can be viewed as a successful strategy to cope with limited energy resources.

Physical activity and reduced food intake have previously been invoked to increase hypothalamic neuropeptide Y (NPY) levels15 and to activate the hypothalamic-pituitary-adrenal axis. Elevated cortisol levels may sensitize the dopaminergic and noradrenergic neurons of the limbic system and reinforce self-starvation via reward mechanisms.4 Very recently it was demonstrated that leptin can modulate the brain reward circuitry.16 Leptin affects neuropeptidergic systems controlling energy metabolism including orexic NPY and anorexic corticotropin releasing factor (CRF). Leptin inhibits the semi-starvation-induced rise in NPY synthesis17 and prevents the stimulation of CRF synthesis and the activation of the adrenal axis.18 We hypothesize that the inhibitory action of exogenous leptin on SIH is most likely to be mediated through these hormones.

In an initial attempt to address the relationship between serum leptin levels and hyperactivity in patients with AN, the subjective experiences of motor restlessness in 30 patients (26 females, 4 males) with AN were assessed at three time points during therapeutically induced weight gain (Figure 3): (a) upon admission for inpatient treatment (T0) in the emaciated state associated with hypoleptinemia (0.65 ± 0.56 μg l−1; range: 0.015–1.62 μg l−1); (b) after serum leptin levels (2.38 ± 1.22 μg l−1) for the first time surpassed the 25th percentile of the gender-dependent reference range formed by healthy controls matched for BMI19; and (c) upon attainment of maximal leptin levels (Tmax) measured during inpatient treatment (5.86 ± 3.85 μg l−1; range 1.20–14.60 μg l−1). After leptin levels for the first time topped the 25th percentile of the reference range, patients reported a trend towards reduced motor restlessness (P = 0.13; one-sided) in comparison to baseline. Upon attainment of the maximal leptin levels motor restlessness was ranked lower as compared to baseline (P = 0.023; one-sided).

Figure 3

Subjective ratings of motor restlessness in 30 patients (four males) with anorexia nervosa at three time points during therapeutically induced weight gain: (1) upon admission for inpatient treatment in the emaciated state (TO); (2) upon attainment of the first serum leptin level at or greater than the 25th percentile of the reference range formed by healthy controls matched for gender and body mass index23 (T25th; four patients had already had a leptin level above this percentile at admission); and (3) upon attainment of the maximal leptin level during inpatient treatment (Tmax). Serum leptin levels were determined biweekly throughout treatment.

The higher subjective ratings of motor restlessness associated with hypoleptinemia upon initiation of treatment, in comparison to lower ratings observed upon attainment of maximal levels, need to be interpreted cautiously. This reduction could be caused by any one or a combination of several factors other than the increment in leptin secretion. Because the ranking of motor restlessness can theoretically also imply a subjective feeling with no overt behavior, objective measurements of energy expenditure and physical activity in parallel to monitoring of leptin secretion are clearly required to verify that motor activity indeed decreases. Nevertheless, our results in both rodents and humans are readily compatible with the hypothesis that hypoleptinemia contributes to the hyperactivity observed in patients with AN.

Whereas SIH has been discussed as a model for the hyperactivity observed in AN,4 other investigators have been concerned with obvious differences.20 Thus, AN in comparison with SIH is associated with a prolonged reduction in food intake. In clinical terms the hyperactivity in some patients seems to be strongly motivated by their pursuit of thinness. However, in other patients the hyperactivity appears beyond the patient's control, thus potentially representing a more biologically driven form of hyperactivity. In our opinion, especially this latter form of hyperactivity could be driven by hypoleptinemia.

In AN, therapeutically induced weight gain leads to normalization of leptin secretion; during this process leptin levels can even intermittently exceed the reference range formed by BMI and gender-matched healthy controls.19, 23 Indeed, in 13 of the 30 patients included in the present study the maximal leptin level exceeded the 90th percentile of the reference range. Patients with AN are well suited to analyse the potential implications of both hypoleptinemia upon emaciation and the increment in leptin secretion during weight gain. Because leptin normalizes activity levels in food-restricted rats, leptin administration in acute AN can possibly reduce hyperactivity potentially rendering patients more amenable to treatment.

Materials and methods

Male Wistar ‘HsdCpb. WU’ rats (Harland Laboratories, Germany) at an age of 38 days, weighing 140–163 g, were housed in individual cages equipped with a running wheel (circumference 110 cm) in a climate chamber (23 ± 1°C). Lights were on from 6:00 am to 6:00 pm. Rats intitially had free access to rat chow (R/M-H, sniff Spezialdiäten GmbH, Germany) and water. On-line activity monitoring started at the age of 52 days, when average body mass had reached 232 g. Body weight and food intake were continuously monitored at 5:00 pm in 2–3 day intervals.

In two independent experiments, food-restricted or ad libitum-fed rats were treated with either recombinant mouse leptin (1.29 mg ml−1, supplied by Hoechst AG, Frankfurt, Germany) or with vehicle (phosphate-buffered saline, PBS) via subcutaneously implanted miniosmotic pumps (Alzet, Model 2001), which release the respective fluids at a rate of 1 μl h−1 for 7 days according to the manufacturer. Thus, rats received 31 μg recombinant mouse leptin per day.

For the first series of experiments a total of 30 rats were allocated to two equally sized vehicle- and leptin-treated groups, by matching for pretreatment activity and body weight. Within each treatment group rats were fed ad libitum (n = 8) or restricted to 60% of ad libitum food intake (n = 7) during the 7-day treatment period and (re)fed ad libitum during 14 days of post-treatment. Minipumps were removed 10 days after implantation. In the second experiment 14 rats were food restricted for 12 days to 60% of ad libitum food intake, with leptin treatment via implantation of minipumps starting on day 6. To verify ongoing leptin infusion at the end of the experiment (7 days after implantation), serum leptin levels were assayed as described previously19 in both leptin-treated and control rats.

The effect of leptin on running wheel activity was analyzed by a repeated-measures ANCOVA procedure adjusted for the activity level of individual rats 1 week before treatment (SPSS 8.0). For graphical presentation running wheel activity was expressed as a percentage of the mean activity during the week prior to food restriction. Mean serum leptin levels in control and leptin-treated rats were compared by Student's t-test. Data are presented as mean ± SEM.

For the purpose of our study we selected 30 patients (4 males) with AN (DSM-IV criteria; mean age ± SD: 16.44 ± 3.5 years; range: 12.3–30.7), who presented with hypoleptinemia (initial serum leptin levels <1.85 μg l−1)24 and whose serum leptin levels during therapeutically induced weight gain topped the 25th percentile of the reference range formed by healthy gender and BMI-matched controls.19, 23 The 30 patients were blood sampled at 8:00 am within 3 days after admission for inpatient treatment to allow determination of serum leptin levels and thereafter biweekly until discharge as reported previously.23 Their mean BMI (± SD) upon the initial sampling after admission was 14.50 ± 1.36 kg m−2 (range: 11.87–16.80 kg m−2). All patients and in case of minors their respective parents gave written informed consent for participation in the clinical study, which was approved by the Ethics Committee of the University of Marburg.

Concomitantly to the biweekly blood samplings patients were asked to rate their subjective experience of motor restlessness during the preceding week on a seven-point visual analogue scale with a score of 1 indicating no motor restlessness and a score of 7 indicating severe restlessness. The relationship between the patients’ serum leptin levels and their scores for motor restlessness were compared at three time points: (1) upon admission in the semi-starved condition associated with hypoleptinemia; (2) after initial normalisation of serum leptin levels determined on an individual basis as the first leptin level achieved during weight gain that topped the 25th percentile of the reference range; and (3) finally after attainment of the individual maximal leptin level determined biweekly during inpatient treatment. Upon attainment of these maximal leptin levels the patients’ mean BMI (± SD) had increased to 17.39 ± 1.02 kg m−2; range: 15.66–19.14 kg m−2). We hypothesized that patients would rate their subjective experience of motor restlessness higher upon admission.


  1. 1

    Beumont PJ, Arthur B, Russell JD, Touyz SW . Excessive physical activity in dieting disorder patients: proposals for a supervised exercise program Int J Eat Disord 1994; 15: 21–36

    CAS  Article  Google Scholar 

  2. 2

    Davis C . Eating disorders and hyperactivity: a psychobiological perspective Can J Psychiatry 1997; 42: 168–175

    CAS  Article  Google Scholar 

  3. 3

    Casper RC . Behavioral activation and lack of concern, core symptoms of anorexia nervosa? Int J Eat Disord 1998; 24: 381–393

    CAS  Article  Google Scholar 

  4. 4

    Bergh C, Sodersten P . Anorexia nervosa, self-starvation and the reward of stress Nat Med 1996; 2: 21–22

    CAS  Article  Google Scholar 

  5. 5

    Russell JC, Epling WF, Pierce D, Amy RM, Boer DP . Induction of voluntary prolonged running by rats J Appl Physiol 1987; 63: 2549–2553

    CAS  Article  Google Scholar 

  6. 6

    Morse AD, Russell JC, Hunt TW, Wood GO, Epling WF, Pierce WD . Diurnal variation of intensive running in food-deprived rats Can J Physiol Pharmacol 1995; 73: 1519–1523

    CAS  Article  Google Scholar 

  7. 7

    Ahima RS, Prabakaran D, Mantzoros C, Qu D, Lowell B, Maratos-Flier E et al. Role of leptin in the neuroendocrine response to fasting Nature 1996; 382: 250–252

    CAS  Article  Google Scholar 

  8. 8

    Kolaczynski JW, Considine RV, Ohannesian J, Marco C, Opentanova I, Nyce MR et al. Responses of leptin to short-term fasting and refeeding in humans: a link with ketogenesis but not ketones themselves Diabetes 1996; 45: 1511–1515

    CAS  Article  Google Scholar 

  9. 9

    Heiman ML, Ahima RS, Craft LS, Schoner B, Stephens TW, Flier JS . Leptin inhibition of the hypothalamic-pituitary-adrenal axis in response to stress Endocrinology 1997; 138: 3859–3863

    CAS  Article  Google Scholar 

  10. 10

    Halaas JL, Gajiwala KS, Maffei M, Cohen SL, Chait BT, Rabinowitz D et al. Weight-reducing effects of the plasma protein encoded by the obese gene Science 1995; 269: 543–546

    CAS  Article  Google Scholar 

  11. 11

    Klingenspor M, Niggemann H, Heldmaier G . Modulation of leptin sensitivity by short photoperiod acclimation in the Djungarian hamster Phodopus sungorus 2000; J Comp Physiol [B] 170: 37–43

    CAS  Article  Google Scholar 

  12. 12

    Scarpace PJ, Matheny M, Pollock BH, Tumer N . Leptin increases uncoupling protein expression and energy expenditure Am J Physiol 1997; 273: E226–E230

    CAS  PubMed  PubMed Central  Google Scholar 

  13. 13

    Döring H, Schwarzer K, Nuesslein-Hildesheim B, Schmidt I . Leptin selectively increases energy expenditure of food-restricted lean mice Int J Obes Relat Metab Disord 1998; 22: 83–88

    Article  Google Scholar 

  14. 14

    Zachwieja JJ, Hendry SL, Smith SR, Harris RB . Voluntary wheel running decreases adipose tissue mass and expression of leptin mRNA in Osborne–Mendel rats Diabetes 1997; 46: 1159–1166

    CAS  Article  Google Scholar 

  15. 15

    Lewis DE, Shellard L, Koeslag DG, Boer DE, McCarthy HD, McKibbin PE et al. Intense exercise and food restriction cause similar hypothalamic neuropeptide Y increases in rats Am J Physiol 1993; 264: E279–E284

    CAS  PubMed  Google Scholar 

  16. 16

    Fulton S, Woodside B, Shizgal P . Modulation of brain reward circuitry by leptin Science 2000; 287: 125–128

    CAS  Article  Google Scholar 

  17. 17

    Ahima RS, Kelly J, Elmquist JK, Flier JS . Distinct physiologic and neuronal responses to decreased leptin and mild hyperleptinemia Endocrinology 1999; 140: 4923–4931

    CAS  Article  Google Scholar 

  18. 18

    Huang Q, Rivest R, Richard D . Effects of leptin on corticotropin-releasing factor (CRF) synthesis and CRF neuron activation in the paraventricular hypothalamic nucleus of obese (ob/ob) mice Endocrinology 1998; 139: 1524–1532

    CAS  Article  Google Scholar 

  19. 19

    Blum WF, Englaro P, Hanitsch S, Juul A, Hertel NT, Muller J et al. Plasma leptin levels in healthy children and adolescents: dependence on body mass index, body fat mass, gender, pubertal stage, and testosterone J Clin Endocrinol Metab 1997; 82: 2904–2910

    CAS  Google Scholar 

  20. 20

    Pirke KM, Broocks A, Wilckens T, Marquard R, Schweiger U . Starvation-induced hyperactivity in the rat: the role of endocrine and neurotransmitter changes Neurosci Biobehav Rev 1993; 17: 287–294

    CAS  Article  Google Scholar 

  21. 21

    Ballauff A, Ziegler A, Emons G, Sturm G, Blum WF, Remschmidt H et al. Serum leptin and gonadotropin levels in patients with anorexia nervosa during weight gain Mol Psychiatry 1999; 4: 71–75

    CAS  Article  Google Scholar 

  22. 22

    Hebebrand J, van der Heyden J, Devos R, Kopp W, Herpertz S, Remschmidt H et al. Plasma concentrations of obese protein in anorexia nervosa Lancet 1995; 346: 1624–1625

    CAS  Article  Google Scholar 

  23. 23

    Hebebrand J, Blum WF, Barth N, Coners H, Englaro P, Juul A et al. Leptin levels in patients with anorexia nervosa are reduced in the acute stage and elevated upon short-term weight restoration Mol Psychiatry 1997; 2: 330–334

    CAS  Article  Google Scholar 

  24. 24

    Köpp W, Blum WF, von Prittwitz S, Ziegler A, Lubbert H, Emons G et al. Low leptin levels predict amenorrhea in underweight and eating disordered females Mol Psychiatry 1997; 2: 335–340

    Article  Google Scholar 

Download references


We thank Franziska Wollnik (Universität Stuttgart, Germany) for generous provision of the running wheels. We thank Tilman Görg for his technical assistance. These studies were supported by the Deutsche Forschungsgemeinschaft.

Author information



Corresponding author

Correspondence to M Klingenspor.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Exner, C., Hebebrand, J., Remschmidt, H. et al. Leptin suppresses semi-starvation induced hyperactivity in rats: implications for anorexia nervosa. Mol Psychiatry 5, 476–481 (2000).

Download citation


  • food intake
  • food restriction
  • energy metabolism
  • neuroendocrine control of appetite
  • locomotor activity
  • running wheels

Further reading


Quick links