Serum adiponectin concentrations during a 72-hour fast in over- and normal-weight humans

Article metrics



Adiponectin is currently considered an important link between obesity and insulin resistance, since circulating levels of this insulin sensitizing hormone have consistently been found to be reduced in obese subjects. However, until now it is not known how the secretion of adiponectin is regulated in response to acute metabolic changes. Here, we assessed the influence of complete fasting for 72 h on serum adiponectin levels.


Between group comparison of repeated measurements.


In total, 18 normal-weight (mean±s.e.m. BMI: 22.2±0.4 kg/m2; age: 39.2±4.4 y) and nine over-weight (BMI: 33.2±1.8 kg/m2; age: 36.9±4.5 y) subjects.


Serum adiponectin levels were measured every 4 h during a 72-h fasting period. Additionally, concentrations of plasma glucose and serum insulin and leptin were assessed at the beginning and in the end of the fasting experiment. Insulin resistance was estimated using the homeostasis model assessment (HOMA).


While concentrations of glucose, insulin, and leptin decreased across the fasting period by 31.0, 33.1 and 60.0%, respectively (all P<0.005), adiponectin levels remained unchanged (P=0.817). Overall, over-weight subjects exhibited slightly lower adiponectin levels than normal-weight subjects (P=0.092), but there was no difference in the time course of adiponectin levels during fasting between these two groups (P=0.970). Although, averaged adiponectin levels before and after fasting did not systematically differ, individual changes in adiponectin levels across fasting displayed a slight but significant inverse correlation with changes in plasma glucose concentration (r=−0.42, P=0.03).


The data show that serum adiponectin concentrations remain remarkably stable during 72 h of fasting in normal- and over-weight subjects. Thus, adiponectin appears to reflect primarily long-term changes in body weight with little evidence for a dependence on short-term regulatory influences.


Adiponectin, a hormone secreted by the adipose tissue, is regarded as a potential link between obesity and insulin resistance.1 Obese subjects have been consistently found to have lower adiponectin levels than normal-weight subjects.2 Diminished adiponectin levels were also shown to be associated with decreased insulin sensitivity, that is, insulin resistance, and to predict the onset of type 2 diabetes mellitus (T2DM).3 On this background, adiponectin is presently believed to play an important role in glucose metabolism.

To date, little is known about the acute regulation of adipose tissue adiponectin secretion. Exercise, for instance, did not to affect circulating adiponectin levels.4 The influence of food intake remains obscure since previous studies have reported decreased,5 unchanged,6 or increased7 levels of the hormone after ingestion of a meal. These inconsistencies may have been due to differences in the meal compositions5, 6 and the subjects' body weight (obese vs lean)7 in these studies subjects.

Fasting is well known to induce substantial changes in glucose metabolism and numerous endocrine functions.8 However, its influence on adiponectin levels have rarely been assessed thus far. Gavrila et al2 have recently reported unchanged adiponectin levels after 48 h of acute fasting in eight normal-weight men. However, responses in adiponectin levels to fasting may depend on body weight or may differ between sexes. Therefore, in the present study we measured adiponectin concentrations in 18 normal-weight and nine over-weight subjects of both sexes who fasted for 72 h. Blood was sampled every 4 h throughout the 72-h study period, since it has previously been shown that adiponectin levels follow a distinct circardian rhythm,9 and it is presently unclear whether this variation depends on regular food supply.

Subjects and methods


The fasting experiments were performed in 18 normal-weight (BMI <26.0 kg/m2) and nine over-weight (BMI ≥28.0 kg/m2) subjects. The normal-weight subjects (men/women: 5/13) had a mean±s.e.m. age of 39.2±4.4 y (range: 18–72 y) and a mean BMI of 22.2±0.4 kg/m2 (range: 19.2–25.7 kg/m2). The over-weight subjects (men/women: 3/6) had a mean age of 36.9±4.5 y (range: 22–62 y) and a mean BMI of 33.2±1.8 kg/m2 (range: 28.2–44.1 kg/m2). None of the subjects had a fasting glucose concentration above 7.0 mmol/l. All subjects gave informed consent.

Fasting procedure

Fasting started after subjects had a regular breakfast in the morning of the first day. The first blood sample was collected at noon of the first day; further samples were taken every 4 h until 0800 on day 4. Patients were allowed to drink water and unsweetened tea. They could walk around in the hospital area and had regular sleep times, except for minor disturbances due to nocturnal blood sampling.


Serum adiponectin levels were measured using an ELISA (Human Adiponectin ELISA Kit, B-Bridge International, Inc., Sunnyvale, CA, USA, intra-assay coefficient of variance (CV) 3.45%, interassay CV 5.2%). Plasma glucose concentration was determined using the glucose hexokinase method. Serum insulin and leptin concentration was determined using ELISAs (Insulin: DAKO Cytomation, Cambridgeshire, UK, inter-assay CV 7.5%, intra-assay CV 6.7%; leptin: Diagnostic Systems Laboratories, Sinsheim, Germany, interassay CV 4.4%, intra-assay CV 3.8%). Insulin resistance was estimated using the homeostasis model (HOMA).10

Statistical analysis

All values are presented as mean±s.e.m. Analysis of variance (ANOVA) was performed including the repeated measure factor ‘time’ (representing the multiple measurements during the fasting period) and the group factors ‘BMI’ (for effects of body weight) and ‘sex’ (for difference between men and women). Also, area under the curve (AUC) for adiponectin levels were calculated and compared between the different groups (over-weight vs normal-weight subjects; men vs women) by unpaired t-test. The study population was additionally divided into two groups accordingly to the adiponectin levels at the beginning of the fast and the AUC adiponectin values by the median split method. Pearson's and partial correlation analyses (using gender as correction factor) were performed to assess relationships between variables at the beginning and end of the fasting period. Also, changes across the fasting period (value at the end minus value at the beginning) were calculated for circulating glucose and hormone levels, and correlated with each other. A P-value <0.05 was considered significant.


Throughout the fasting experiment, over-weight subjects tended to have lower serum adiponectin levels than normal-weight subjects (P=0.092 for main effect of ‘BMI’), and the same was true for men compared to women (P=0.074 for main effect of ‘sex’; Figure 1). Analyses of AUC adiponectin values revealed similar results with levels tending to be lower in over- than normal-weight subjects (101 548±21 493 vs 154 823±16 020 μg/ml × h; P=0.072) and lower in men than in women (96 515±20 476 vs 154 139±16 730 μg/ml × h; P=0.059). However, serum adiponectin concentrations remained remarkably stable throughout the fasting period (beginning: 8.15±0.78 μg/ml, end: 7.95±0.77 μg/ml, P=0.817). This result was neither influenced by body weight (P=0.970 for the ‘time by BMI’ interaction term) nor by gender (P=0.270 for the ‘time by sex’ interaction term). There was no evidence for a circadian variation in adiponectin concentrations in the entire study group as well as in a single subgroup (Figure 1).

Figure 1

Mean±s.e.m. adiponectin concentrations for men (left) and women (right) in normal-weight (open circles) and over-weight (black circles) subjects during 72 h of complete fasting. Normal-weight men (n=5), over-weight men (n=3), normal-weight women (n=13), over-weight women (n=6).

In contrast to adiponectin, plasma glucose concentration as well as serum insulin and leptin concentration dropped by 31.0, 33.1 and 60.0% during fasting, respectively (beginning vs end: glucose: 5.0±1.1 vs 3.4±1.1 mmol/l, P<0.001; insulin: 47.3±6.2 vs 31.6±4.2 pmol/l, P=0.003; leptin: 36.8±6.8 vs 14.7±2.6 ng/ml, P<0.001). HOMA also decreased from 1.5±0.2 to 0.7±0.1 (P<0.001). This result should be interpreted with caution since HOMA has not been validated for the prolonged fasting state known to increase rather than decrease insulin resistance.11

When the study population was divided into two groups accordingly to the adiponectin concentration at the beginning of the fast, analyses revealed that subjects with low basal adiponectin levels showed a distinctly greater reduction in circulating insulin during the fasting period than subjects with high basal adiponectin levels (−3.7±1.0 vs −1.0±0.6 pmol/l, P=0.031). In contrast, the extent of the decease in plasma glucose concentrations during fasting did not differ between the two groups (−1.6±0.2 vs −1.5±0.5, P=0.753). Identical results were found when the study population was divided accordingly to the AUC adiponectin values.

At the beginning of the fasting period, adiponectin levels were significantly correlated with HOMA (r=−0.42, P=0.03) and insulin concentrations (r=−0.40, P=0.04), but not with BMI (r=−0.17, P=0.39), and concentrations of glucose (r=−0.27, P=0.17) and leptin (r=0.18, P=0.36). These results did not change after controlling for gender, with HOMA (r=−0.45, P=0.04) and insulin concentrations (r=−0.46, P=0.03) being significantly related to adiponectin levels. In the end of the fasting period, adiponectin levels were still inversely correlated with HOMA (r=−0.44, P=0.04) and insulin concentrations (r=−0.42, P=0.05), but with none of the other variables (all P>0.45).

Although in the entire study population adiponectin levels did not systematically change across the fasting period, individual changes in adiponectin levels were correlated in an inverse manner with changes in plasma glucose concentration (r=−0.42, P=0.03; Figure 2). No such correlation was found between changes in adiponection and changes in any other variable (all P>0.22).

Figure 2

Correlation between changes (delta) in serum adiponection and plasma glucose concentrations during a 72-h fasting period. Delta=value at the end of fast−value at the beginning of fast.


The present data show for the first time that in both normal- and over-weight subjects serum adiponectin levels remain remarkably stable during a 72-h fasting period. Thus, unlike secretion of other hormones, for example, insulin and leptin, adiponectin secretion does not appear to be subjected to strong acute regulatory influences upon fasting.

In line with previous results,2, 6, 7 adiponectin levels tended to be lower in over-weight than normal-weight subjects and also in men than women. The previously reported inverse correlation of adiponectin levels with HOMA and insulin levels was likewise found in the present study.2 Together, these confirmatory findings provide support for the reliability of our adiponectin data.

The finding of unchanged adiponectin levels during fasting agrees with results of a foregoing study showing unaltered adiponectin levels over a 48-h period of acute fasting in eight normal-weight men.2 Weight reduction has been found to increase serum adiponectin levels in some12, 13, 14 but not all studies.15 The present result of stable adiponectin levels during prolonged (72 h) total fasting suggests that the changes in circulating adiponectin concentrations observed after weight reduction do not result from a short-term negative energy balance. Rather they may reflect decreased body fat content.

Contrasting with the results of a previous study by Gavrila et al,9 we did not find any hint at a circadian variation in serum adiponectin levels. In that foregoing study, which was performed in six lean healthy men, adiponectin levels showed a maximum diurnal variation of 37.5±6.8% across the 24-h cycle with nadir levels at about 0315 and peak levels at about 1100. The difference between these and the present results cannot be explained on the basis of our data. However, considering that in the present study samples were collected only every 4 h, which contrasts the much more frequent sampling carried out by Gavrila et al (every 15 min), it might be possible that here a more discrete circadian variation of adiponectin has simply been missed. Also, it could be possible that meal intake is implicated in the circadian variations of the hormone since in the study by Gavrila et al, the subjects were fed with an isocaloric diet. The expression of a circadian rhythm in adiponectin levels may also depend on gender and body weight. Thus, slight differences in these variables in the subject samples of the two studies might also have contributed to the inconsistent outcomes. In this context, it should be mentioned that our results fit with another study,12 which likewise failed to detect any systematic circadian variation in adiponectin levels in seven nondiabetic (3 men/4 women; mean BMI 31.1±2.4 kg/m2) and six diabetic (4 men/2 women; mean BMI 33.0±3.5 kg/m2) subjects.

Despite the fact that adiponectin levels did not systematically change during fasting, individual changes in adiponectin levels were inversely correlated with changes in plasma glucose concentration during the experiment. Although one should keep in mind that this association cannot prove a causal relationship, it temps to speculate that even slight changes in adiponectin levels exert a discrete influence on plasma glucose levels probably by modulating hepatic glucose output,1 which is the main source of circulating glucose during fasting.


  1. 1

    Chandran M, Phillips SA, Ciaraldi T, Henry RR . Adiponectin: more than just another fat cell hormone? Diabetes Care 2003; 26: 2442–2450.

  2. 2

    Gavrila A, Chan JL, Yiannakouris N, Kontogianni M, Miller LC, Orlova C, Mantzoros CS . Serum adiponectin levels are inversely associated with overall and central fat distribution but are not directly regulated by acute fasting or leptin administration in humans: cross-sectional and interventional studies. J Clin Endocrinol Metab 2003; 88: 4823–4831.

  3. 3

    Spranger J, Kroke A, Mohlig M, Bergmann MM, Ristow M, Boeing H, Pfeiffer AF . Adiponectin and protection against type 2 diabetes mellitus. Lancet 2003; 361: 226–228.

  4. 4

    Kraemer RR, Aboudehen KS, Carruth AK, Durand RT, Accvcdo EO, Hebert EP, Johnson LG, Castracanc VD . Adiponectin responses to continuous and progressively intense intermittent exercise. Med Sci Sports Exerc 2003; 35: 1320–1325.

  5. 5

    Esposito K, Nappo F, Giugliano F, Di Palo C, Ciotola M, Barbieri M, Paolisso G, Giugliano D . Meal modulation of circulating interleukin 18 and adiponectin concentrations in healthy subjects and in patients with type 2 diabetes mellitus. Am J Clin Nutr 2003; 78: 1135–1140.

  6. 6

    Peake PW, Kriketos AD, Denyer GS, Campbell LV, Charlesworth JA . The postprandial response of adiponectin to a high-fat meal in normal and insulin-resistant subjects. Int J Obes Relat Metab Disord 2003; 27: 657–662.

  7. 7

    English PJ, Coughlin SR, Hayden K, Malik IA, Wilding JP . Plasma adiponectin increases postprandially in obese, but not in lean, subjects. Obes Res 2003; 11: 839–844.

  8. 8

    Schwartz MW, Seeley RJ . Neuroendocrine responses to starvation and weight loss. N Engl J Med 1997; 336: 1802–1811.

  9. 9

    Gavrila A, Peng CK, Chan JL, Mietus JE, Goldberger AL, Mantzoros CS . Diurnal and ultradian dynamics of serum adiponectin in healthy men: comparison with leptin, circulating soluble leptin receptor, and cortisol patterns. J Clin Endocrinol Metab 2003; 88: 2838–2843.

  10. 10

    Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC . Homeostasis model assessment: insulin resistanceand beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 412–419.

  11. 11

    Newman WP, Brodows RG . Insulin action during acute starvation: evidence for selective insulin resistance in normal man. Metabolism 1983; 32: 590–596.

  12. 12

    Hotta K, Funahashi T, Arita Y, Takahashi M, Matsuda M, Okamoto Y, Iwahashi H, Kuriyama H, Ouchi N, Maeda K, Nishida M, Kihara S, Sakai N, Nakajima T, Hascgawa K, Muraguchi M, Ohmoto Y, Nakamura T, Yamashita S, Hanafusa T, Matsuzawa Y . Plasma concentrations of a novel, adipose-specific protein, adiponectin, in type 2 diabetic patients. Arterioscler Thromb Vasc Biol 2000; 20: 1595–1599.

  13. 13

    Yang WS, Lee WJ, Funahashi T, Tanaka S, Matsuzawa Y, Chao CL, Chen CL, Tai TY, Chuang LM . Weight reduction increases plasma levels of an adipose-derived anti-inflammatory protein, adiponectin. J Clin Endocrinol Metab 2001; 86: 3815–3819.

  14. 14

    Monzillo LU, Hamdy O, Horton ES, Ledbury S, Mullooly C, Jarema C, Porter S, Ovalle K, Moussa A, Mantzoros CS . Effect of lifestyle modification on adipokine levels in obese subjects with insulin resistance. Obes Res 2003; 11: 1048–1054.

  15. 15

    Ryan AS, Nicklas BJ, Berman DM, Elahi D . Adiponectin levels do not change with moderate dietary induced weight loss and exercise in obese postmenopausal women. Int J Obes Relat Metab Disord 2003; 27: 1066–1071.

Download references


We would like to thank Christiane Otten and Ingrid von Lützau for their technical assistance and Cornelia Jaworski for her organizational work.

Author information

Correspondence to B Schultes.

Rights and permissions

Reprints and Permissions

About this article


  • starvation
  • insulin resistance
  • HOMA
  • glucose metabolism
  • adiponectin

Further reading