OUP user menu

Reliability of leptin, but not adiponectin, as a biomarker for diet‐induced weight loss in humans

Monica C Klempel, Krista A Varady
DOI: http://dx.doi.org/10.1111/j.1753-4887.2011.00373.x 145-154 First published online: 1 March 2011

Abstract

Calorie restriction (CR)‐induced weight loss has been shown to lower the risk of chronic disease in obese individuals. Although the mechanisms that link weight loss to disease risk reduction remain unclear, evidence suggests adipokines may play a role. What has yet to be determined, however, is the dose‐response effect of body weight loss and visceral fat mass loss on adipokines. Accordingly, this review examines how varying degrees of CR‐induced weight loss (i.e., >10%, 5–10%, and <5% from baseline) impact plasma levels and expression of adiponectin, leptin, resistin, interleukin 6 (IL‐6), interleukin 8 (IL‐8), monocyte chemotactic protein 1 (MCP‐1), and retinol‐binding protein 4 (RBP‐4). The dose‐response relationship between visceral fat mass loss and adipokine profile improvement will also be explored. Results from this review demonstrate that even mild weight loss induced by CR may have beneficial effects on leptin levels, but it has no clear impact on adiponectin, resistin, IL‐6, IL‐8, MCP‐1, or RBP‐4 concentrations.

  • adipokines
  • calorie restriction
  • obesity
  • visceral fat
  • weight loss

INTRODUCTION

Overweight and obese individuals are at a greater risk for developing coronary heart disease and type 2 diabetes, compared to their normal‐weight counterparts.1 Weight loss has been shown to lower the risk of these diseases in obese individuals by modulating several key biomarkers.2 More specifically, losing 5–15% of initial body weight is associated with improved plasma lipid profile, decreased systolic and diastolic blood pressure, and increased insulin sensitivity.2,3 Although the mechanisms that link weight loss to disease risk reduction remain unclear, evidence suggests fat‐cell‐derived hormones (i.e., adipokines) may play a role.4,5 Adiponectin is a fat‐cell‐derived hormone shown to be inversely related to body weight and visceral fat mass, though the data are not consistent. This hormone is also currently thought to exert cardioprotective effects.6 Leptin, resistin, interleukin 6 (IL‐6), interleukin 8 (IL‐8), monocyte chemotactic protein 1 (MCP‐1), and retinol‐binding protein 4 (RBP‐4), in contrast, are adipokines that are augmented in overweight and obese individuals and exhibit proinflammatory properties.7,8 Thus, diet therapies that are able to increase adiponectin concentrations while decreasing leptin, resistin, IL‐6, IL‐8, MCP‐1, and RBP‐4 concentrations may confer protection against chronic disease in obese patients.

Calorie restriction (CR) regimens are the most commonly prescribed diet strategies to help overweight and obese individuals lose weight.9 CR regimens typically involve restricting energy intake from 15% to 40% of needs daily. More extreme CR regimens that require an individual to restrict energy intake by up to 75% of daily requirements (i.e., very‐low‐calorie diets) have also been implemented to induce rapid weight loss. Previous reports indicate that CR‐induced weight loss beneficially affects plasma levels and expression of key adipokines.10 What has yet to be determined, however, is the minimal degree of weight loss required to improve the expression and levels of each of these adipokines. Moreover, the dose‐response effect of body weight loss and visceral fat mass loss on adipokine concentrations also remains unclear.

Accordingly, this review examines how varying degrees of CR‐induced weight loss (i.e., >10% from baseline, 5–10% from baseline, and <5% from baseline) impacts adiponectin, leptin, resistin, IL‐6, IL‐8, MCP‐1, and RBP‐4 plasma levels and expression. The dose‐response relationship between visceral fat mass loss and adipokine profile improvement will also be explored. Only randomized clinical trials of CR (conducted from 1999 until present) performed in adults were included in the review. Trials that included diabetic individuals were excluded from the analysis.

ADIPONECTIN: RESPONSE TO DIET‐INDUCED WEIGHT LOSS

Adiponectin is a hormone secreted by adipocytes that is found in relatively high concentrations in plasma (0.01% of total plasma protein).11 Adiponectin exhibits both anti‐atherogenic and insulin‐sensitizing properties. The mechanisms by which adiponectin reduces atherosclerotic plaque formation are thought to include reduction in monocyte attachment to the vascular bed,12 inhibition of the transformation of macrophages into foam cells,13 and attenuation of growth‐factor‐induced proliferation of vascular smooth muscle cells.14 With regards to its insulin‐sensitizing actions, adiponectin has been shown to decrease circulating glucose levels by suppressing gluconeogenesis in the liver and enhancing insulin signaling in the skeletal muscle.15 Previous studies have indicated there is an inverse relationship between body weight and circulating adiponectin concentrations.16 Consequently, plasma adiponectin levels are decreased in overweight and obese adults (body mass index [BMI] > 25 kg/m2) and are elevated in individuals with normal body weight (BMI 18.5–25 kg/m2).16 The dose‐response effect of CR‐induced weight loss on adiponectin plasma levels and expression are reported in Table 1.

View this table:
Table 1

Effect of diet‐induced weight loss on adipokine concentrations and expression.

ReferenceSubjectsTrial lengthIntervention groupsΔ Body weight*Δ Fat mass*Δ Visceral fat mass*Δ Adiponectin*Δ Leptin*Δ Other adipokines*
PlasmamRNAPlasmamRNAPlasma
>10% Weight loss
 Torgerson (1999)39n = 69 MF

Obese

Age 44 ± 11 years
16 weeks1. 75% CR1. ↓18%1. ↓65%
 Kok (2005)38n = 11 F

Obese

Age 36 ± 2 years
16 weeks1. 70% CR

2. Control
1. ↓15%

2. Ø
1. ↓40%

2. Ø
1. ↓46%

2. Ø
 Heinonen (2009)17n = 35 MF

Obese

Age 53 ± 2 years
8 weeks1. 60% CR1. ↓15%1. ↓26%1. ↓27%1. ↑37%1. ↓62%IL‐6

1. ↓39%
 Esposito (2003)18n = 120 F

Obese

Age 35 ± 5 years
104 weeks1. 22% CR

2. Control
1. ↓15%

2. ↓3%
1. ↓9%

2. ↓2%
1. ↑9%

2. ↑9%
IL‐6

1. ↓32%

2. Ø
 Belza (2009)19n = 41 MF

Obese

Age 43 ± 11 years
20 weeks1. 55% CR1. ↓14%1. ↓33%1. ↓12%1. Ø1. ↓43%IL‐6

1. ↓21%
 Chan (2008)20n = 20 MF

Obese

Age 46 ± 8 years
16 weeks1. 30% CR1. ↓12%1. ↓22%1. ↓24%1. ØRBP‐4

1. ↓20%
 Larrouy (2008)21n = 10 F

Obese

Age 37 ± 7 years
14 weeks1. 40% CR1. ↓11%1. ↓19%1. Ø1. Ø1. ↓52%1. ØIL‐6

1. ↓42%
 Claessens (2009)22n = 48 MF

Obese

Age 46 ± 2 years
6 weeks1. 20% CR high protein

2. 20% CR high carb
1. ↓12%

2. ↓7%
1. ↓21%

2. ↓22%
1. ↓11%

2. ↓10%
1. Ø

2. Ø
1. ↓56%

2. ↓56%
 Christiansen (2010)23n = 59 MF

Obese

Age 37 ± 7 years
12 weeks1. 65% CR

2. 55% CR
1. ↓11%

2. ↓8%
1. ↓10%

2. ↓11%
1. ↑19%

2. ↑20%
1. ↑120%

2. ↑120%
 Browning (2008)53n = 54 F

Overweight

Age 45 ± 13 years
24 weeks1. 25% CR1. ↓11%1. ↓23%1. ↓30%IL‐6

1. ↓41%

MCP‐1

1. ↓29%
5–10% Weight loss
 Vitkova (2007)62n = 24 F

Obese

Age 38 ± 10 years
8 weeks1. 40% CR1. 10%1. ↓23%RBP‐4

1. ↓15%
 Keogh (2007)24n = 36 MF

Obese

Age 48 ± 2 years
12 weeks1. 20% CR high protein

2. 20% CR high carb
1. ↓9%

2. ↓8%
1. ↓5%

2. ↓7%
1. ↓6%

2. ↓6%
 Plat (2007)58n = 11 M

Obese

Age 59 ± 9 years
4 weeks1. 70%1. ↓9%MCP‐1

1. Ø
 Mahon (2007)25n = 54 F

Obese

Age 58 ± 2 years
9 weeks1. 40% CR1. ↓9%1. ↓9%1. Ø
 Kovacikova (2008)44n = 47 F

Overweight, obese

Age 42 ± 2 years
12 weeks1. 30% CR1. ↓8%1. ↓17%1. ↓8%1. ↓46%
 Arvidsson (2004)26n = 40 F

Obese

Age 35 ± 2 years
10 weeks1. 30% CR moderate fat

2. 30% CR low fat
1. ↓8%

2. ↓8%
1. Ø

2. Ø
1. Ø

2. Ø
1. Ø

2. Ø
1. ↓23%

2. ↓35%
1. Ø

2. Ø
IL‐6

1. ↓39%

2. ↓31%

IL‐8

1. Ø

2. Ø
 Garaulet (2004)27n = 33 F

Obese

Age 37 ± 6 years
4 weeks1. 50% CR1. ↓8%1. ↓10%1. ↓5%1. Ø1. Ø1. ↓56%
 Capel (2008)43n = 94 F

Obese

Age 39 ± 1 years
10 weeks1. 30% CR moderate fat

2. 30% CR low fat
1. ↓7%

2. ↓6%
1. ↓12%

2. ↓12%
1. ↓2%

2. ↓2%
1. ↓36%

2. ↓35%
 Ramel (2009)45n = 324 MF

Overweight

Age 30 ± 4 years
8 weeks1. 30% CR

2. 30% CR lean fish

3. 30% CR fatty fish

4. 30% CR fish oil
1. ↓5%

2. ↓6%

3. ↓6%

4. ↓6%
1. ↓4%

2. ↓5%

3. ↓6%

4. ↓5%
1. Ø

2. Ø

3. ↓23%

4. ↓20%
<5% Weight loss
 Kratz (2008)29n = 26 MF

Overweight,

obese

Age 38 ± 14 years
14 weeks1. 15% CR

2. 15% CR n‐3 PUFA
1.↓5%

2.↓4%
1. ↓9%

2. ↓6%
1. Ø

2. Ø
1. Ø

2. Ø
 de Luis (2007)30n = 90 MF

Obese

Age 43 ± 15 years
12 weeks1. 25% CR high protein

2. 25% CR high carb
1. ↓5%

2. ↓4%
1. ↓7%

2. ↓6%
1. Ø

2. Ø
1. ↓23%

2. ↓17%
Resistin

1. Ø

2. Ø
 Kabrnova‐Hlavata (2008)31n = 67 F

Overweight,

obese

Age 49 ± 12 years
4 weeks1. 35% CR

2. 35% CR Ca suppl.
1. ↓4%

2. ↓5%
1. ↓10%

2. ↓11%
1. Ø

2. Ø
1. Ø

2. Ø
Resistin

1. ↓7%

2. ↓14%
 Peairs (2008)31n = 19 MF

Overweight

Age 38 ± 10 years
1 weeks1. 30% CR high fat1. ↓4%IL‐6

1. Ø

MCP‐1

1. Ø
  • * Post‐treatment values significantly different (P < 0.05) from baseline values within an intervention group.

  • Abbreviations and symbols: BW, body weight, Ca suppl, calcium supplement; Carb, carbohydrate; CR, calorie restriction; F, female; High carb, high carbohydrate; IL‐6, interleukin 6; IL‐8, interleukin 8; M, male; MCP‐1, monocyte chemoattractant protein 1; PUFA, polyunsaturated fatty acids; RBP‐4, retinol binding protein 4; Ø, nonsignificant effect; ↑, increase; ↓, decrease.

Adiponectin changes associated with a greater than 10% weight loss

To date, seven clinical trials1723 have been performed to evaluate the effects of >10% weight loss on plasma adiponectin levels and expression. Two of these studies demonstrate increases in plasma adiponectin,17,23 while five studies reveal no effect.1822 For instance, in the study by Heinonen et al.,17 plasma adiponectin increased by 37% in response to a 15% reduction in body weight after 8 weeks of 60% CR in obese males and females.17 Fat mass and visceral fat mass were also significantly reduced by 26% and 27%, respectively, in this study. These beneficial effects of weight loss on adiponectin levels are supported by the findings of Christiansen et al.23 In this trial, obese subjects were randomized to partake in a 65% CR or a 55% CR intervention.23 Body weight loss was greater with 65% CR (11% weight loss) compared with 55% CR (8% weight loss).23 Despite the slightly greater weight loss in the 65% CR group, increases in plasma adiponectin (20% increase from baseline) and expression (120% increase from baseline) were similar between interventions.23 Thus, no dose‐response effect of weight loss on adiponectin levels was observed in this study.23 Although the trials by Heinonen et al.17 and Christiansen et al.23 reveal beneficial effects of a >10% weight loss on adiponectin levels, the majority of studies reviewed here show no effect.1822 The reason for these conflicting findings is not clear. However, it would appear as though the trials that were shorter in duration (i.e., 8–12 weeks) and that applied the greatest degree of CR (i.e., 60–65% CR) produced the most beneficial changes in adiponectin plasma levels and expression. As such, it is possible that the increases in adiponectin levels associated with weight loss peak within the first 12 weeks of treatment and then level off or diminish as the intervention continues. Future studies testing the time‐course of the effects of CR interventions on weight loss and changes in adiponectin concentrations would help clarify whether these associations hold true.

Adiponectin changes associated with a 5–10% weight loss

In examining the four trials2427 that tested this objective, it would appear as though this degree of weight loss has little or no effect on plasma adiponectin levels. In the study by Keogh et al.,24 which demonstrated a mild reduction in adiponectin, subjects were randomized to a 20% CR/high‐carbohydrate (HC) diet (20% of kcal as protein, 60% of kcal as carbohydrates, 20% of kcal as fat) or a 20% CR/high‐protein (HP) diet (40% of kcal as protein, 33% of kcal as carbohydrates, 27% of kcal as fat). No effect of background diet was noted, as both groups experienced similar reductions in plasma adiponectin levels (6% decrease from baseline) in response to an 8–9% weight loss (5–7% visceral fat mass loss).24 In contrast, the studies by Mahon et al.,25 Arvidsson et al.,26 and Garaulet et al.27 report no effect of an 8–9% weight loss and a 5–9% visceral fat mass loss on circulating levels of this adipokine. What set the study of Keogh et al.24 apart from these three other studies2527 was the inclusion of both men and women in the study population. These three studies,2527 which reported no effect of this degree of weight loss on adiponectin levels, included only women. These findings may suggest adiponectin levels in female participants are less responsive to CR‐induced weight loss than those in male participants. A study that directly compares changes in adiponectin levels between men and women in response to the same CR intervention would help clarify this assumption. Closer evaluation of these studies indicates another possible reason adiponectin did not respond to this degree of weight loss, i.e., because not enough visceral fat mass was lost over the course of the trial. Previous reports indicate that visceral fat mass is negatively correlated with circulating adiponectin, particularly in younger women (<40 years).28 Unfortunately, the precise degree of visceral fat mass loss required to elevate plasma adiponectin cannot be established from the studies reviewed here. It should be noted, however, that in both of the studies in which improvements in adiponectin levels were observed (i.e., Heinonen et al.17 and Christiansen et al.,23 discussed in the previous section), visceral fat mass decreased by >10% from baseline. Since this degree of visceral fat mass loss was not attained in these trials, this may partly explain the lack of effect on adiponectin levels.

Adiponectin changes associated with a lower than 5% weight loss

Since there was no impact of a 5–10% weight loss on adiponectin levels, it is not surprising that a <5% weight loss also has no effect.2931 For example, Kabrnova‐Hlavata et al.31 demonstrate that 4 weeks of 35% CR produced 4–5% decreases in body weight (10–11% decreases in fat mass), with no effect on circulating adiponectin. In support of these findings, de Luis et al.30 also reported no effect of a 4–5% weight loss (6–7% fat mass loss) on adiponectin levels after 12 weeks of dietary restriction. Participants in this study30 were randomized to either a 25% CR/HC diet (15% of kcal as protein, 60% of kcal as carbohydrates, 25% of kcal as fat) or a 25% CR/HP diet (35% of kcal as protein, 30% of kcal as carbohydrates, 35% of kcal as fat). The background diet appeared to have no effect on adiponectin levels, as neither intervention produced significant changes. These findings for dietary macronutrient composition are similar to those of Keogh et al.24 (reported in the 5–10% weight loss section of this review). Moreover, in the trial by Kratz et al.,29 it was shown that a CR diet high in n‐3 polyunsaturated fatty acids (PUFA) has very little effect on plasma adiponectin and expression with modest weight loss (4–5% from baseline). In this trial,29 subjects were randomized to either a 15% CR diet rich in n‐3 PUFA (3.5% of energy intake) or an approximately 15% CR control diet. The results reveal that a high level of n‐3 PUFA intake has no added benefit over a control diet low in n‐3 PUFA (0.5% of energy intake) on circulating adiponectin levels in overweight and obese adults.29 Taken together, these studies suggest a <5% body weight loss and a <10% fat mass loss are not sufficient to modulate circulating adiponectin levels.

LEPTIN: RESPONSE TO DIET‐INDUCED WEIGHT LOSS

Like adiponectin, leptin is a hormone released by fat cells. Leptin inhibits appetite by acting on receptors in the hypothalamus, where it counteracts the effects of feeding stimulants secreted by cells in the gut and in the hypothalamus.32 Variations in the leptin receptor, along with defects in leptin receptor signaling, have been associated with obesity.33 Therefore, leptin levels in obese individuals are typically elevated, which can lead to leptin resistance.34 Leptin also exhibits pro‐atherogenic properties. This adipokine promotes vascular proliferation and smooth muscle cell migration, which decreases arterial elasticity and consequently leads to obesity‐associated vascular diseases.35 The expression and secretion of leptin is positively correlated with body weight and fat mass.36 Circulating leptin levels are also directly proportional to the total amount of visceral fat in the body.37 Thus, individuals who are viscerally obese may have higher leptin concentrations than obese individuals with comparable amounts of adipose tissue stores located in the subcutaneous depots.37 Changes in leptin plasma levels and expression that occur with varying degrees of diet‐induced weight loss are shown in Table 1.

Leptin changes associated with a >10% weight loss

From the studies reviewed here, it would appear as though leptin is highly responsive to diet‐induced weight loss.17,19,21,22,38,39 More specifically, plasma leptin levels decreased by 43–65% with a >10% weight loss, though no clear dose‐response relationship could be noted.17,19,21,22,38,39 Torgerson et al.39 demonstrated the most potent leptin decrease (65% from baseline), which corresponded to the greatest degree of weight loss (18% from baseline). However, in the studies by Kok et al.,38 Heinonen et al.,17 and Belza et al.,19 similar reductions in circulating leptin (43–62% from baseline) were noted in response to a lesser degree of weight loss (14–15% from baseline). To further obscure any dose‐response effect, comparable decreases in leptin (52–56% from baseline, with no effect on leptin expression) were observed in relation to an 11–12% weight loss in the trials by Larrouy et al.21 and Claessens et al.22 Thus, a greater degree of weight loss may not always be required to observe significant reductions in leptin concentrations. Upon further investigation, it was also observed that no dose‐response relationship exists between visceral fat mass loss and change in leptin concentrations. Kok et al.38 demonstrated the greatest degree of visceral fat mass loss (40% from baseline) but found only a moderate decrease in leptin levels (46% from baseline). In comparison, Belza et al.19 reported the least amount of visceral fat mass loss (12% from baseline) but observed a reduction in leptin concentrations (43% from baseline) similar to that shown by Kok et al.38 Torgerson et al.39 did not measure visceral fat mass loss, and therefore it is difficult to conclude whether visceral fat or total body weight was the primary determinant of leptin change in this study. Complementary to the findings for adiponectin, macronutrient distribution of the background diet appears to have no effect on changes in leptin with CR. For instance, in the study by Claessens et al.,22 subjects were randomized to either a 20% CR/HC diet (15% of kcal as protein, 63% of kcal as carbohydrates, 22% of kcal as fat) or a 20% CR/HP diet (35% of kcal as protein, 40% of kcal as carbohydrates, 25% of kcal as fat). Despite differences in the macronutrient composition of the background diets, the same changes in plasma leptin levels were observed (56% decrease from baseline) with a 7–12% weight loss and a 10–11% visceral fat mass loss. It should also be noted that hypoleptinemia (leptin deficiency) can develop in response to extreme weight loss (i.e., >20% of initial body weight)40,41 and is also found in patients with anorexia nervosa.42

Leptin changes associated with a 5–10% weight loss

To date, five trials2645 have examined the effects of a 5–10% weight loss on leptin plasma levels and expression. These studies show that a decrease in body weight of 6–8% significantly lowered circulating leptin in range of 20–46% from baseline.2645 Similar to the findings reported for a >10% weight loss, no dose‐response relationship could be noted between body weight reduction and leptin decrease. For instance, Garaulet et al.27 and Kovacikova et al.44 reported the greatest decreases in plasma leptin (56% and 46% from baseline, respectively), in response to an 8% weight loss. Arvidsson et al.,26 Capel et al.,43 and Ramel et al.45 demonstrated slightly less potent decreases in leptin (20–35% from baseline) with a 6–8% weight loss. As for leptin expression, no changes were observed with this degree of weight loss.26 It should also be noted that these beneficial modulations in leptin in each of these studies occurred with very minimal decreases in visceral fat mass (i.e., 5–8% from baseline). These findings suggest changes in body weight may have more impact on circulating leptin concentrations than changes in visceral fat mass. Dietary fatty acid composition may also impact the response of leptin to CR. For example, in the study by Ramel et al.,45 leptin levels were reduced only in subjects supplementing their CR diet with fish oil or fatty fish. Although the precise mechanisms have yet to be determined, these preliminary findings suggest fish oil supplementation during CR may provide an added benefit by inducing a more potent leptin‐lowering effect.45

Leptin changes associated with a <5% weight loss

Only two studies have reported the effects of a <5% weight loss on leptin concentrations.7,31 Findings from these trials are contradictory, as one study7 reports mild decreases in leptin, while the other7,31 reports no effect. In the trial by de Luis et al.,7 obese men and women were randomized to a 25% CR/HP diet or a 25% CR/HC diet. After 12 weeks of treatment, subjects in both groups lost 4–5% of initial body weight and experienced a decrease in leptin levels ranging from 17–23%.7 In comparison, Kabrnova‐Hlavata et al.31 demonstrated no effect on plasma leptin with a 4–5% weight loss after 4 weeks of a 35% CR diet in overweight and obese men and women. The lack of effect of weight loss on leptin levels in the Kabrnova‐Hlavata et al.31 study may be partly explained by the study's small sample size (n = 67) and the inclusion of both obese and overweight subjects. It is possible that if the Kabrnova‐Hlavata et al.31 study had implemented a larger sample size and had limited the population group to obese subjects to reduce intersubject variation, changes in leptin levels associated with diet may have been significant. Whether changes in visceral fat mass play a role in modulating lipid levels with this degree of weight loss can, unfortunately, not be determined because abdominal fat was not assessed in either of these studies. Evidently, more studies are required in this area for solid conclusions to be reached.

OTHER KEY ADIPOKINES: RESPONSE TO DIET‐INDUCED WEIGHT LOSS

Resistin

Resistin is a proinflammatory mediator secreted predominantly by macrophages embedded in adipose tissue.46 The role of resistin in obesity and insulin resistance remains unclear.47 Higher levels of circulating resistin associated with obesity and insulin resistance have been shown in some, but not all, studies.47 Resistin may stimulate insulin resistance by increasing hepatic glucose production while decreasing the uptake of fatty acids in skeletal muscle.48 Resistin also displays potent proinflammatory properties.49 Specifically, resistin has been shown to accumulate at the site of inflammation and, once there, support the inflammatory process by triggering cytokine production while simultaneously upregulating its own expression.50 The effects of CR‐induced weight loss on plasma resistin concentrations are summarized in Table 1.

At present, only two studies have examined the effects of weight loss on circulating resistin levels. After 12 weeks of a 25% CR diet, de Luis et al.30 observed no change in plasma resistin concentrations in obese men and women. In contrast, Kabrnova‐Hlavata et al.31 reported mild decreases in resistin (7–14% from baseline) after 4 weeks of a 35% CR diet. Although weight loss in both of these studies30,31 was similar (4–5% decrease from baseline), the percent reduction in fat mass in the Kabrnova‐Hlavata et al.31 study (10–11% decrease from baseline) exceeded that in the de Luis et al.30 study (6–7% decrease from baseline). The greater fat mass reduction observed by Kabrnova‐Hlavata et al.31 may therefore explain why the subjects in this study experienced decreases in plasma resistin while subjects in the de Luis et al.30 study did not. Changes in visceral fat mass were not assessed in either study; thus, no comparisons can be made with regard to this anthropometric measure. Evidently, much more work is required in this area to determine the effects of varying degrees of weight loss on resistin levels. It will be of great interest to establish whether weight loss can reduce levels of this proinflammatory mediator and to determine how these changes in resistin relate to modulations in disease risk.

Interleukin 6 and interleukin 8

IL‐6 is a proinflammatory cytokine produced in substantial quantities by adipose tissue.51 Levels of this cytokine are strongly related to body weight and BMI.51 IL‐6 has been shown to act at multiple levels, both centrally and on peripheral tissues, to influence body weight, energy homeostasis, and insulin sensitivity.52 From the studies reviewed here, IL‐6 appears to be highly responsive to weight loss, though no clear dose‐response relationship is evident (Table 1). Specifically, plasma IL‐6 levels decrease by 21–42% with an 11–15% weight loss and a 9–27% visceral fat mass loss,1753 and by 31–39% with an 8% weight loss.26 No effect on IL‐6 was noted in response to a 4% weight loss, however.54 These data indicate a weight loss of >8% may reduce plasma IL‐6, while weight loss regimens that result in a loss of <5% weight may not alter circulating levels of this adipokine.

IL‐8 is a potent proinflammatory cytokine released by adipose tissue.51 Plasma IL‐8 is augmented in obese individuals compared with lean individuals and is correlated with measures of insulin resistance and the development of atherosclerosis.51 Only one study to date has tested the effect of CR‐induced weight loss on levels of this adipokine (Table 1). That study revealed that an 8% weight loss (with either a low‐fat or a moderate‐fat background diet) has no effect on plasma levels of IL‐8.26 Since the data are quite limited, much more research is required before a relationship between the degree of weight loss and changes in IL‐8 levels can be established.

Monocyte chemotactic protein 1

MCP‐1 is expressed in adipocytes and is considered an adipokine.55 MCP‐1 mediates the infiltration of macrophages into adipose tissue in obesity and may play a pivotal role in establishing the proinflammatory state that predisposes individuals to the development of insulin resistance.56 Obese individuals have higher circulating concentrations of MCP‐1 compared to their lean counterparts.57 Three studies to date have examined the ability of diet‐induced weight loss to lower MCP‐1 levels (Table 1). These studies demonstrate that reductions in body weight of >10% decrease MCP‐1 levels by 29%,53 while a 4–9% weight loss has no effect on circulating levels of this adipokine.54,58 Whether or not greater weight loss (i.e., 10–15%) significantly lowers levels of MCP‐1 is an important question that has yet to be tested.

Retinol‐binding protein 4

RBP‐4 is a newly discovered adipokine that is released from adipose tissue.59 In addition to its action as a carrier for retinol in the blood, RBP‐4 has been implicated in the pathophysiology of insulin resistance.5961 Moreover, RBP‐4 is strongly correlated with each component of the metabolic syndrome, including increased BMI and waist circumference.5961 Two clinical trials have examined the effects of CR‐induced weight loss on the plasma levels of this novel adipokine (Table 1). In the study by Chan et al.,20 a 12% weight loss (23% fat mass loss) significantly reduced RBP‐4 plasma levels by 20%. Complementary to these effects, Vitkova et al.62 demonstrated that a 10% weight loss (23% fat mass loss) decreased circulating RBP‐4 by 15%. Taken together, these findings suggest RBP‐4 may be responsive to moderate weight loss (10–12%) and fat mass loss (23%) induced by CR. More studies are needed to establish a dose‐response curve between these variables. Moreover, since RBP‐4 is strongly correlated with waist circumference, it will be of interest for future trials to establish how reductions in visceral fat mass affect levels of this adipokine and how these changes impact insulin resistance.

SUMMARY OF FINDINGS

While some studies report increases in plasma adiponectin (20–37% from baseline) and expression with a >10% weight loss,17,23 the majority of trials demonstrate no effect.1822 In line with these findings, CR regimens that induce a 4–9% weight loss also have no effect on circulating levels of this adipokine.17,23 Regarding visceral fat mass reduction, a decrease of >10% in visceral fat may be necessary to increase plasma adiponectin concentrations.17,23 However, this relationship requires further confirmation. Taking these findings together, clinicians who wish to design a weight loss intervention to increase plasma adiponectin in obese individuals should aim to achieve a >10% weight loss from baseline in conjunction with a >10% reduction in visceral fat mass.

In contrast to adiponectin, leptin appears to be highly responsive to CR‐induced weight loss. Reducing body weight by >10% from baseline produced significant declines in circulating leptin levels, ranging from 43% to 65%.17,19,21,22,38,39 Moderate weight loss (5–10% reduction from baseline) was also shown to have favorable effects on this adipokine (20–46% reduction from baseline).2645 With regard to mild weight loss (<5% from baseline), one study reports a reduction in leptin of approximately 20%,30 while another study reports no effect.31 In terms of visceral fat mass reduction, losing >10% of abdominal fat mass may be necessary to observe these beneficial changes in leptin concentrations.

One weight loss technique that appears to have consistent effects on adiponectin and leptin levels is bariatric surgery.6365 For instance, adiponectin levels have been shown to significantly increase by 30–140% with a 25–35% weight loss following surgery.63 A meta‐analysis of these findings also revealed the overall percentage increase of adiponectin levels strongly correlated with the percentage decrease in BMI and body weight.63 Leptin is also highly responsive to weight loss induced by bariatric surgery.64 On average, the levels of this hormone decrease by 70–95% with a 25–35% weight loss following this intervention.64 These findings suggest a >25% weight loss may be required in order to achieve consistent increases in adiponectin and consistent decreases in leptin.6365 Since CR‐induced weight loss does not generally produce such extreme weight loss, this may explain why the impact of CR on these adipokines is less consistent.

As for the effects of CR‐induced weight loss on plasma resistin, IL‐6, IL‐8, MCP‐1, and RBP‐4 levels, no solid conclusions can be reached, as the data are extremely limited. In examining the findings for resistin, one of the two studies in this area reported a decrease in resistin (7–14%) with a <5% weight loss,31 while another study reported no effect.30 Changes in visceral fat mass were not assessed in either study; thus, no conclusions can be reached with regard to the effect of fat distribution on resistin levels. In contrast, circulating IL‐6 concentrations appear to be highly responsive to weight loss. Specifically, a >10% weight loss resulted in a 21–42% decrease in levels of this adipokine,1753 while a 5–10% weight loss resulted in a 31–39% reduction.26 No effect of a <5% weight loss was noted, however.54 As for IL‐8, moderate weight loss (5–10%) has no effect on levels of this cytokine.26 MCP‐1 plasma levels decrease by 29% in response to a >10% weight loss53 but remain unchanged if weight loss falls below the 10% threshold.54,58 In terms of RBP‐4, results from one study indicate a >10% weight loss is able to reduce levels of this adipokine by 20% from baseline.20 It is important to note, however, that the evidence of effects on levels of resistin, IL‐6, IL‐8, MCP‐1, and RBP‐4 is very limited at present. Thus, much more research will be required to establish a relationship between diet‐induced weight loss and changes in each of these adipokines.

CONCLUSION

In summary, this review demonstrates that even mild weight loss induced by dietary restriction may have beneficial effects on leptin levels. In contrast, no clear relationship between degree of weight loss and changes in adiponectin, resistin, IL‐6, IL‐8, MCP‐1, and RBP‐4 could be established. More specifically, for leptin, a reduction in body weight as minimal as 4–5% from baseline may help to decrease circulating levels of leptin in overweight and obese individuals. However, in order to see the most favorable effects on leptin (including modulations in adiponectin plasma levels), a >10% weight loss, coupled with a >10% visceral fat mass loss, is most likely necessary. Weight‐loss‐induced enhancements in leptin have been linked to improvements in several chronic disease indicators, including plasma lipid levels, blood pressure, and insulin sensitivity. As such, clinicians should consider these weight loss guidelines when designing CR interventions to produce the most favorable changes in leptin and to confer maximal protection against chronic diseases in obese patients.

Acknowledgments

Funding

This work was funded by a departmental grant from the University of Illinois, Chicago, Department of Kinesiology and Nutrition.

Declaration of interest

The authors have no relevant interests to declare.

REFERENCES

View Abstract