The Hydration of Adipose Tissue — The Circuit Linking Water and Fat, and the Instrument That Cannot Measure It

1. Abstract

Body composition analyzers routinely report that total body water changes little during weight loss. This finding is more likely an artifact of the measurement model than a physiological fact, because standard bioelectrical impedance analysis (BIA) is built on the assumption that adipose tissue contains neither intracellular nor extracellular fluid. Recent physiology has described a circuit that contradicts that assumption. The adipocyte membrane carries a channel that conducts both water and glycerol (aquaporin-7, AQP7), and the more obstructed this water pathway becomes, the more lipid accumulates. As adipocytes enlarge, hypoxia and inflammation follow, interstitial fluid increases, and within that interstitium sodium is stored without accompanying water. Tonicity-responsive enhancer binding protein (TonEBP), which senses this local hypertonicity, epigenetically represses the β3-adrenergic receptor, the central switch for lipid oxidation. Water and fat therefore do not stand in a one-directional causal relationship but form a mutually reinforcing loop. This paper organizes that circuit into four layers and identifies the points at which clinical instruments fail to observe it.

2. The Question

A long-standing clinical observation frames the problem. Two people of identical weight and identical body fat percentage can inhabit entirely different bodies: one firm, the other heavy and prone to swelling. Clinicians have habitually invoked “water” to explain this difference, and many carry the impression that adipose tissue in an obese body holds a great deal of it.

Body composition analyzers do not support that impression. In studies of GLP-1 receptor agonists, body weight and fat mass fall markedly while total body water (TBW) and extracellular water (ECW) barely move. In a 30-day tirzepatide study (n=115), body weight decreased by 4.0 kg and fat mass by 3.6 kg, yet changes in TBW and ECW were negligible (sex-stratified subanalysis, p=0.733 / 0.240).

— PMC12256807

Longer-term administration shows the same pattern: body weight, fat mass, fat-free mass, skeletal muscle, TBW, ECW and ICW all decrease significantly, while the indices of water distribution themselves are preserved.

— Front Endocrinol, 2026

Read at face value, these numbers yield the conclusion that water does not move even as fat is lost. That conclusion, however, is reached without examining how the instrument sees water in the first place. This is precisely the point at which a limitation of the tool is mistaken for physiology.

What matters clinically is not which number is correct. It is how adipose tissue actually handles water, and how that process is entangled with lipid accumulation itself. If water and fat constitute a mutually reinforcing circuit, then a weight-loss strategy that watches only the scale and one that also watches water will not produce the same result.

3. What the Instrument Cannot See

The two-compartment model on which BIA and DXA rely rests on the assumption that the hydration of fat-free mass (FFM) is constant at approximately 0.73. That constant was derived in normal-weight subjects and does not hold in obesity. The fat-free component within adipose tissue is more highly hydrated, and the ECW/ICW ratio of the connective-tissue portion of adipose tissue is also higher. Consequently, as BMI rises, BIA and DXA systematically overestimate fat-free mass relative to three- and four-compartment models (p<0.001).

— PMC6696776

A more direct observation emerges after weight loss. The ECW/ICW ratio increased following weight reduction, and the investigators explicitly noted that if adipose tissue water had been the cause, FFM hydration should have fallen, whereas it moved in the opposite direction.

— J Appl Physiol, 1999

The root of the problem lies in the model’s premise. Standard BIA assumes that adipose tissue contains neither intracellular nor extracellular fluid, and the relevant patent literature explicitly describes this assumption as incorrect. Direct measurement tells a different story. Tissue water content of adipose tissue ranges from 17% to 84% by freeze-drying assay, a wide spread. The commonly cited figure of 10–20% refers to the adipocyte itself, not to the tissue as a whole.

In short, the instrument used to assess “water” excludes from its calculation the very tissue in which water exists most ambiguously. This is why the “no change in TBW” reported in the GLP-1 studies of the preceding section must not be read as a physiological conclusion.

4. The Circuit — Four Layers

4-1. The Water Pathway — AQP7

The glycerol channel of adipose tissue belongs to the aquaglyceroporin family and conducts water and glycerol together. AQP7 is the long-studied member, but in humans it is not the only one.

In mice, AQP7 has been treated as the sole aquaglyceroporin of adipose tissue. In human adipose tissue, however, two are present, and their distribution differs: AQP7 occurs in the plasma membranes of both adipocytes and capillaries, while AQP10 is expressed exclusively in adipocytes. Silencing AQP10 in differentiated human adipocytes reduced glycerol permeability by roughly 51% and osmotic water permeability by roughly 46%.

— PLOS One, 2013 / PMC3558521

The presence of two channels is held to be important for keeping intracellular glycerol content low, which the authors interpret as protection against obesity. The discussion that follows rests on AQP7 data while carrying the caveat that, in humans, the contribution of AQP10 has not been separated out.

Findings from 3T3-L1 adipocytes define this layer. AQP7 expression itself did not alter equilibrium cell volume, but loss of function raised triglyceride content. A negative correlation between water permeability and non-osmotic cell volume was reported for the first time, while water transport rate and glycerol transport rate showed a strong positive correlation.

— PMC3869813 / PMID 24376702

The implication is that cells less able to conduct water accumulate lipid more readily. Human data point the same way. Negative correlations are observed among methylation at the AQP7 locus, adipose AQP7 expression, and BMI. AQP7-null mice gain body weight and visceral fat through adipocyte hypertrophy and chronic inflammation, an effect aggravated by a high-glucose diet.

— npj Metab Health Dis, 2025 / PMC12595075

The direction of this layer is nonetheless inconsistent.

— PMC9963055

The authors themselves note the scarcity of functional studies of AQP7 in its physiological context. Drawing categorical conclusions at this layer would exceed the strength of the evidence.

4-2. The Interstitium — Where Fluid Collects

As adipocytes enlarge, hypoxia develops within the tissue, followed by inflammation and excessive collagen deposition. Inflammation raises microvascular permeability, and interstitial fluid increases as a result.

— PMC7719100

Lipedema shows this combination consolidated into pathology, with adipocyte hypertrophy, interstitial fibrosis and excess interstitial fluid progressing together by stage.

— PMC10417720

The relationship with the lymphatic system is bidirectional. Overnutrition impairs lymphatic function, producing local lymphedema, adipogenesis and insulin resistance; conversely, lymphatic injury accelerates lipid accumulation and insulin resistance.

— Nat Rev Endocrinol, 2026

Cellular patterns also diverge by obesity class. In Class I obesity adipocyte size is maximal and inflammation, insulin resistance and endoplasmic reticulum stress are highest, whereas in Class II and III adipogenesis proceeds through proliferation.

— PMC10289932

Inflammation at this layer does not arise solely within adipose tissue. A separate input arrives from the gut. The phenomenon in which gut microbiota-derived lipopolysaccharide (LPS) enters the circulation and generates low-grade systemic inflammation is termed metabolic endotoxemia.

Human data exist. When 33 patients scheduled for bariatric surgery were stratified by circulating LPS, the group with high metabolic endotoxemia showed lower expression of key genes for adipose function and lipogenesis (SREBP1, FABP4, FASN, LEP) and higher expression of inflammatory genes in both visceral and subcutaneous depots. In vitro work confirmed LPS as the direct cause of adipocyte inflammation and of downregulation of PPARG, SCD, FABP4 and LEP.

— Am J Physiol Endocrinol Metab, 2018

The causal direction has also been demonstrated experimentally. Subcutaneous infusion of LPS for four weeks into wild-type mice on normal chow increased whole-body, hepatic and adipose tissue weight and produced adipose and hepatic inflammation with fasting hyperglycemia and hyperinsulinemia, effects comparable to those induced by a high-fat diet.

— Cani et al., 2007, Diabetes

In adipose tissue, LPS induces adipogenesis, insulin resistance, macrophage infiltration, oxidative stress, and release of pro-inflammatory cytokines and chemokines.

— J Mol Endocrinol, 2013

Contrary findings exist regarding the route of translocation. In a Gambian cohort of women (lean n=48, obesity n=62, obesity-diabetes n=38), elevated LPS and IL-6 and adipose macrophage activation were confirmed, yet the lactulose-mannitol dual-sugar permeability test revealed no abnormality in gut permeability, and no evidence was found for chylomicron-associated postprandial translocation. Markedly lower EndoCAb IgM instead pointed to a failure of sequestration and detoxification.

— Int J Obes, 2022 / PMC9492538

The existence of metabolic endotoxemia and the attribution of its route to barrier disruption are therefore separate propositions. The problem may lie in processing capacity rather than in the quantity entering.

4-3. Sodium Without Water

This layer overturns conventional expectation.

Glycosaminoglycans in the dermal interstitium bind sodium through strong negative charge and store it in an osmotically inactive state. Sodium accumulates without accompanying water. As a result, sodium concentration and osmolality in the dermal interstitium exceed those of plasma. A human being carries roughly 5 kg of skin, and skin is the largest extracellular fluid reservoir in the body. One week of high-salt intake increases dermal sodium together with macrophages, VEGF-C and lymphatic vessels.

— Titze, PMID 14975935 and others

From the standpoint of clinical instrumentation, this layer is doubly invisible. Because it draws no water, it produces no change in volume; because it is osmotically inactive, it does not alter electrical resistance. Yet local hypertonicity is real and mobilizes immunity and lymphatics. Here lies the structural reason why a patient may appear swollen while body composition readings remain normal.

4-4. The Sensor Turns Off the Switch — TonEBP

The local hypertonicity generated in the preceding layer is not measured, but it is sensed.

TonEBP is a transcription factor responsive to hypertonicity. In subcutaneous adipocytes of mice fed a high-fat diet, TonEBP expression increased more than fiftyfold. Mice deficient or haplo-deficient in TonEBP resisted high-fat diet-induced obesity and metabolic derangement, showed increased oxygen consumption, tolerated hypothermia, and underwent browning of subcutaneous fat. The mechanism was recruitment of DNMT1 by TonEBP to epigenetically repress the promoter of the β3-adrenergic receptor.

— Nat Commun, 2019 / PMC6684655 (Kwon Hyug Moo group, UNIST)

β3 is a central regulator of lipolysis and thermogenesis. A sensor of osmotic stress is therefore switching off the machinery that burns fat: a structure in which state governs function.

5. Reading Across — Hydration as a Circuit

5-1. The Loop

The four layers are not independent observations; they close into a single loop.

   Obstructed water pathway (AQP7↓)
        ↓
   Lipid accumulation
        ↓
   Adipocyte hypertrophy
        ↓
   Hypoxia, inflammation, fibrosis → interstitial fluid↑
        ↓
   Local hypertonicity (GAG-bound Na storage)
        ↓
   TonEBP↑ → DNMT1 → repression of β3 promoter
        ↓
   Lipolysis and thermogenesis↓
        ↓
   Lipid accumulation  ──┐
        └────────────────┘

The evidentiary strength of each arrow is uneven. If the loop holds, however, the relationship between fat and water cannot be reduced to the one-directional statement that water is abundant because fat is abundant. The behavior of water sits upstream of lipid accumulation.

5-2. Where Instrument and Circuit Diverge

Of the four layers, BIA captures only one even partially, and it measures that one while excluding adipose tissue from the calculation. The “unchanged body water” cited in Section 2 does not mean that water failed to move; it means that the water falling within this instrument’s range failed to move.

5-3. Interventions Reach Different Layers

If the circuit holds, weight-loss interventions cannot be grouped by what they remove. The same operation described as “removing water” produces different outcomes depending on which layer it reaches.

Mapping fluid-related interventions onto the layers yields the following.

Three points emerge from this mapping.

First, free water and stored sodium are distinct layers requiring distinct operations. Sweating and diuresis move free water. Osmotically inactive sodium is scarcely removed by these routes. In health, sodium losses through sweat and feces are each approximately 5 mmol per day, roughly 0.12% of a total body sodium pool of about 4,200 mmol, and 90–95% of sodium regulation is handled by the kidney.

— Am J Physiol Regul Integr Comp Physiol, 2018

An important detail is that excretion of stored sodium proceeds in two stages. Sodium bound to glycosaminoglycans does not belong to the circulating pool and is therefore not directly filterable by the kidney. When restricted intake lowers circulating concentration, mobilization from the reservoir into the circulation must occur before renal excretion can follow. The excretory route is thus renal, while sodium restriction and maintained hydration constitute the conditions that create the mobilization step.

That mobilization does occur is established. Tissue sodium decreased both after treatment of primary aldosteronism and after administration of an SGLT-2 inhibitor. What remains unconfirmed is whether dietary restriction alone, without pharmacological or surgical intervention, generates a gradient of comparable magnitude, and what its time constant would be.

A study that imaged dermal sodium stores before and after aerobic exercise found sweat loss of only about 100 g, corresponding to an expected sodium loss of approximately 5 mmol, under 0.2% of total body sodium.

— Sports Med, 2021 (citing Hammon et al., 2015b)

The colon operates in the opposite direction. A normal colon can absorb up to 400 mEq of sodium per day against a substantial concentration gradient, and aldosterone strengthens reabsorption not only in the renal tubule but also in the colon, salivary glands and sweat glands. As sodium intake falls, the colon and sweat glands diminish as excretory routes, and excretion converges on the kidney.

— Gastroenterology, 1969 / Am J Physiol Regul Integr Comp Physiol, 2018

Second, dehydration and sodium depletion move in opposite directions within the circuit. Removing free water alone raises the relative sodium concentration of the remaining tissue, which from the standpoint of layer 4-4 constitutes a rise in local hypertonicity. Reducing sodium intake while maintaining hydration lowers it. In fluid-directed intervention, maintained hydration is not a safety measure but a necessary condition of the circuit. This prediction remains mechanistic and has not been verified in humans.

Third, the gut participates in the circuit as a route of inflammatory input, not as a route of excretion. Calculating intestinal evacuation as sodium excretion does not add up quantitatively. As established in Section 4-2, however, gut-derived LPS load is an independent input to adipose inflammation, and this is where the intestine actually meets the circuit. Because delayed intestinal transit during weight loss may increase that input, bowel management belongs to the management of inflammatory load rather than to ancillary care. No data confirm that any particular purgative formulation lowers circulating LPS or markers of adipose inflammation.

5-4. The Instrument May Conceal the Result

The most consequential practical conclusion of this circuit concerns measurement.

In patients with primary aldosteronism, muscle sodium was 29% higher than in normal subjects, and after successful treatment tissue sodium was mobilized without any loss of body weight.

— Hypertension, 2011

More direct evidence comes from a randomized controlled trial. In a double-blind crossover study of 59 patients with type 2 diabetes, six weeks of dapagliflozin 10 mg significantly reduced dermal tissue sodium from 24.1 ± 6.6 to 22.7 ± 6.4 A.U. (p=0.013), while tissue water content of skin and muscle showed no significant change.

— PMC5753452

This result confirms the character of layer 4-3 in humans. Sodium fell while water did not. The concept of osmotically inactive storage is supported by direct measurement, and at the same time it is confirmed that changes at this layer are not captured by indices of body water.

Two further points follow. First, the reduction amounted to only about 6%, the product of six weeks of a potent pharmacological intervention that blocks glucose and sodium reabsorption in the proximal tubule at a 1:1 ratio. The reservoir moves slowly. Second, in the same trial muscle sodium did not change significantly (20.5 ± 3.5 vs 20.4 ± 3.7, p=0.801). Response differs by tissue, and where adipose tissue falls has not been measured.

An intervention directed at layer 4-3 will therefore not register on the scale or the body composition analyzer even when it succeeds. The result is not absent; the instrument is measuring something else.

The converse also holds. A body water reading from a composition analyzer cannot serve as grounds for excluding a water problem involving adipose tissue. A normal reading alongside a complaint of swelling is not a contradiction but possibly a blind spot of the tool.

5-5. An Applied Case — The Rationale Behind a Detoxification Weight-Loss Program

Applying this layered analysis to an actual program clarifies why interventions must be positioned at different points in the circuit. The detoxification weight-loss program operated by the author comprises four components, each reaching a distinct layer.

Three matters are explained by this arrangement.

Why sodium restriction and hydration form a single axis, and at what intensity. Separated, the two operations reverse in direction. Only sodium reduction without water restriction presses the segment running from 4-3 to 4-4. Reducing sodium while also inducing dehydration raises local hypertonicity in the remaining tissue and may produce offsetting effects at 4-4.

Intensity has a lower bound. Sodium restriction in this program means discontinuing added salt in cooking and at the table; it does not extend to excluding the sodium intrinsic to foods, which would constitute deficiency-level restriction. The distinction is not arbitrary. According to a Cochrane review, severe sodium restriction significantly increases renin, aldosterone, noradrenaline, cholesterol and triglyceride, with the rise in cholesterol driven chiefly by LDL.

— Graudal et al., 2017, Cochrane Database Syst Rev

In animal work, ten weeks of a low-sodium diet impaired IRS1 and AKT phosphorylation, producing hepatic insulin resistance, and impaired lipoprotein lipase activity, elevating plasma triglyceride (PMC7949138). Since aldosterone is a sodium-conserving hormone, this response amounts to an intervention applying its own brake against the effort to deplete the 4-3 reservoir.

Interventions directed at 4-3 are therefore not better when stronger; there is a range. Restriction of added sodium returns habitual excess to the recommended range and differs from the deficiency range examined in the studies above. Water intake likewise has an upper bound. Under low sodium intake, excessive water carries a risk of dilutional hyponatremia, so intake guided by thirst is the operating principle and blanket instructions to consume large volumes are inappropriate.

Why the urine-and-perspiration axis of the herbal component is separate from the sodium axis. As established above, perspiration and intestinal evacuation are quantitatively negligible with respect to stored sodium. This axis targets free water and interstitial fluid, while sodium is handled by the restriction axis. Describing both as “removing body water” collapses operations that belong to different layers.

Why the purgative component is an inflammation axis rather than an excretory one. Calculated as a sodium route, intestinal evacuation amounts to roughly 5 mmol per day and is negligible. As established in Section 4-2, however, gut-derived LPS is an independent input to adipose inflammation. Sequence matters here. Restoring barrier integrity while inflammatory byproducts remain stagnant in the lumen confines that load. The load must first be evacuated, after which barrier recovery may proceed. Because caloric restriction during weight loss can itself delay intestinal transit and produce constipation, bowel management is not an ancillary measure serving adherence but a component of inflammatory load management.

This arrangement is derived from the circuit and does not constitute clinical validation of the program itself. As the table in Section 5-3 indicates, the evidentiary level of each axis is uneven.

6. What Remains Unknown

Limits of evidence level. The AQP7 and TonEBP data derive largely from cell lines and murine models. Human data remain correlational, with no intervention studies. The circuit presented here connects individually reported layers; no single study has verified the loop as a whole.

Scope of the Titze data. Sodium storage without water was reported in the dermal interstitium. Whether the same storage form occurs in adipose tissue has not been confirmed. The link between 4-3 and 4-4 is at present a hypothesis.

No evidence for dietary intervention alone. Reductions in stored sodium have been confirmed only for renal routes (SGLT-2 inhibition, treatment of aldosteronism); whether dietary restriction alone produces the same change is unconfirmed. Trials applying 23Na MRI to low-sodium intervention are under way in systemic lupus erythematosus (NCT02525835), rheumatoid arthritis (NCT03649178) and hemodialysis patients (NCT03189758), the last of which explicitly states that the impact of dietary sodium intake on tissue sodium is currently unknown.

Effect size of moderate restriction is particularly uncertain. The only available benchmark is the approximately 6% reduction in dermal sodium after six weeks of dapagliflozin, the result of a pharmacological intervention that directly blocks proximal tubular reabsorption. There is no basis for estimating the magnitude or time constant of dietary manipulation at the level of restricting added sodium. A moderate-intensity study lowering dialysate sodium from 138 to 135 mmol/L (NCT03525223) is under way, with results unpublished.

Response differs by tissue. In the same trial, dermal sodium decreased while muscle sodium did not. Storage behaves differently across tissues, and where adipose tissue falls has not been measured. Applying the discussion of Section 4-3 directly to adipose tissue is, at present, extrapolation.

Program-level validation is absent. The arrangement presented in Section 5-5 is a configuration derived from the circuit; no data measure whether the program actually alters tissue sodium, interstitial fluid or circulating LPS. The evidentiary level of the individual axes is likewise uneven.

Layer-specific effects of perspiration are unmeasured. That perspiration reduces free water is self-evident, but whether it actually reduces interstitial fluid (4-2) has not been measured. Tissue sodium rises rather than falls immediately after exercise and is highly sensitive to posture and prior activity (PMC11790297), so short-term measurement at this layer requires interpretive caution.

GLP-1 and this circuit. No study connects them directly. Three points of contact nonetheless exist.

1. They meet head-on at the β3 node. Liraglutide activated type 2 deiodinase (D2) in murine adipose tissue and enhanced β3-adrenergic thermogenesis. Intraperitoneal administration increased brown adipose tissue oxygen consumption and raised UCP-1 protein, the mechanism being intracellular thyroid hormone activation via increased D2 activity (Front Endocrinol, 2022 / PMC8771968). TonEBP epigenetically switches β3 off while liraglutide switches its downstream on. They meet in opposite directions at the same node, yet whether GLP-1 affects TonEBP itself has never been tested.

2. Directions diverge at AQP7. Downregulation of AQP7 has been observed one month after sleeve gastrectomy, but no study has examined the same with GLP-1.

3. An unexpected signal appears in the kidney. After single and repeated administration of liraglutide and semaglutide, renin (Ren1) expression was strongly induced in renal vascular smooth muscle (p<0.0001). Elevated renin may act, via aldosterone, in the direction of sodium retention, leaving open the possibility that this runs counter to the conventional account that GLP-1 promotes natriuresis. The authors themselves describe the downstream effects as unclear (PMC8279630).

What has not been measured. No data exist from human adipose tissue biopsy under GLP-1 administration measuring cell size and tissue water directly. Adipocytes are known to change in size rather than number during adult weight fluctuation (Endocr Rev, 2022), but how water participates in that size change is a separate question.

Further research questions

1. How do AQP7 expression and tissue water content change in human adipose tissue biopsy before and after GLP-1 administration?
2. Does dermal sodium (23Na MRI) move independently of body weight and fat mass during weight loss?
3. How do TonEBP activity and GLP-1-induced β3 signaling interact within the same tissue?
4. Does osmotically inactive sodium storage occur in the adipose interstitium as well?
5. Measured against four-compartment models or 23Na MRI, how biased are BIA body water values across obesity classes?
6. Does dietary sodium restriction alone reduce stored dermal and muscle sodium, and over what time constant?
7. Does bowel management during weight loss lower circulating LPS and markers of adipose inflammation?
8. Depending on whether hydration is maintained, does an identical reduction in body fluid act in opposite directions on local tissue osmolality?

References

Source 1

• Source: Fat-free mass hydration and multi-compartment model comparison in obesity
• Access: PMC6696776
• Reliability: high
• Key point: BIA and DXA systematically overestimate fat-free mass relative to 3- and 4-compartment models as BMI rises (p<0.001) Source 2
• Source: Body composition changes after weight loss — ECW/ICW ratio
• Author/Year: 1999, 『J Appl Physiol』
• Reliability: high
• Key point: ECW/ICW ratio increased after weight loss, opposite to the adipose-water hypothesis Source 3
• Source: AQP7 water permeability and non-osmotic volume in 3T3-L1 adipocytes
• Access: PMC3869813 / PMID 24376702
• Reliability: high
• Key point: First report of a negative correlation between water permeability and non-osmotic volume; water and glycerol transport rates strongly correlated Source 4
• Source: AQP7 methylation, expression and BMI in human adipose tissue
• Author/Year: 2025, 『npj Metab Health Dis』 / PMC12595075
• Reliability: high
• Key point: Negative correlation among AQP7 methylation, expression and BMI; obesity via hypertrophy and chronic inflammation in null mice Source 5
• Source: Aquaporin-7 in metabolic tissues — context-dependent regulation
• Access: PMC9963055
• Reliability: medium (inconsistent direction across studies)
• Key point: Upregulated in visceral fat, downregulated in subcutaneous fat; authors note the lack of studies in physiological context Source 6
• Source: Adipocyte hypertrophy, hypoxia, fibrosis and interstitial fluid
• Access: PMC7719100
• Reliability: high
• Key point: Hypertrophy → hypoxia → inflammation → collagen deposition; raised microvascular permeability increases interstitial fluid Source 7
• Source: Obesity and lymphatic dysfunction — bidirectional relationship
• Author/Year: 2026, 『Nat Rev Endocrinol』
• Reliability: high
• Key point: Bidirectional causation between impaired lymphatic function and lipid accumulation with insulin resistance Source 8
• Source: Sodium storage in skin interstitium without commensurate water retention
• Author/Year: Titze, PMID 14975935 and others
• Reliability: high
• Key point: Osmotically inactive storage of GAG-bound sodium; skin as the largest extracellular fluid reservoir Source 9
• Source: TonEBP-DNMT1 axis epigenetically represses β3-adrenergic receptor
• Author/Year: 2019, 『Nat Commun』 / PMC6684655
• Reliability: high
• Key point: TonEBP increased more than fiftyfold in subcutaneous fat under high-fat diet; deficiency yields browning and obesity resistance Source 10
• Source: Liraglutide activates D2 and enhances β3-adrenergic thermogenesis
• Author/Year: 2022, 『Front Endocrinol』 / PMC8771968
• Reliability: high
• Key point: Increased brown fat oxygen consumption and UCP-1; opposite direction from TonEBP at the same β3 node Source 11
• Source: Renin (Ren1) induction in renal vascular smooth muscle after GLP-1 RA
• Access: PMC8279630
• Reliability: medium (downstream effects unresolved)
• Key point: Strong Ren1 induction after liraglutide and semaglutide administration (p<0.0001) Source 12
• Source: Tirzepatide 30-day body composition by BIA (n=115)
• Access: PMC12256807
• Reliability: high (instrument limitations discussed separately in the text)
• Key point: Negligible change in TBW and ECW despite −4.0 kg body weight and −3.6 kg fat mass Source 13
• Source: Metabolic endotoxemia promotes adipose dysfunction and inflammation in human obesity
• Author/Year: 2018, 『Am J Physiol Endocrinol Metab』
• Reliability: high
• Key point: Human adipose biopsy; lipogenic genes downregulated and inflammatory genes upregulated in the high-LPS group Source 14
• Source: Metabolic endotoxemia initiates obesity and insulin resistance
• Author/Year: Cani et al., 2007, 『Diabetes』
• Reliability: high
• Key point: Four weeks of subcutaneous LPS alone produced adipose and hepatic inflammation and weight gain comparable to a high-fat diet Source 15
• Source: Possible mediators of metabolic endotoxemia in women with obesity
• Author/Year: 2022, 『Int J Obes』 / PMC9492538
• Reliability: high (contrary finding)
• Key point: Elevated LPS confirmed but no gut permeability abnormality; lower EndoCAb IgM indicates failure of processing capacity Source 16
• Source: Mechanisms of sodium balance — total body sodium and renal excretion
• Author/Year: 2018, 『Am J Physiol Regul Integr Comp Physiol』
• Reliability: high
• Key point: Total body sodium approx. 4,200 mmol; 90–95% regulated renally; sweat and feces approx. 5 mmol/day each Source 17
• Source: Absorption and secretion by the colon
• Author/Year: 1969, 『Gastroenterology』
• Reliability: high
• Key point: The colon absorbs up to 400 mEq sodium per day against the gradient; balance maintained under low intake Source 18
• Source: Exertional heat stress and sodium balance
• Author/Year: 2021, 『Sports Med』
• Reliability: high
• Key point: Dermal sodium imaged before and after exercise; loss at 100 g sweat is under 0.2% of total body sodium Source 19
• Source: ²³Na MRI of tissue sodium
• Author/Year: 2011, 『Hypertension』
• Reliability: high
• Key point: Muscle sodium 29% higher in aldosteronism; tissue sodium mobilized after treatment without weight loss Source 20
• Source: SGLT-2 inhibition with dapagliflozin reduces tissue sodium content (RCT)
• Access: PMC5753452
• Reliability: high
• Key point: Tissue sodium reduction confirmed via renal-route intervention Source 21
• Source: Adaptation to fasting by glycerol transport through aquaporin 7
• Author/Year: Hibuse et al., 2005, 『PNAS』 / PMID 15591341
• Reliability: high
• Key point: AQP7 as the glycerol exit route during fasting; β3-agonist response impaired in its absence Source 22
• Source: Effects of low sodium diet versus high sodium diet on blood pressure, renin, aldosterone, catecholamines, cholesterol, and triglyceride
• Author/Year: Graudal et al., 2017, 『Cochrane Database Syst Rev』
• Reliability: high
• Key point: Severe sodium restriction significantly increases renin, aldosterone, noradrenaline, cholesterol and triglyceride, driven by LDL Source 23
• Source: Dietary sodium restriction alters muscle lipidomics that relates to insulin resistance in mice
• Access: PMC7949138
• Reliability: medium (animal model)
• Key point: Ten weeks of low-sodium diet impaired IRS1/AKT phosphorylation and LPL activity, raising triglyceride Source 24
• Source: Modulation of Tissue Sodium in Hemodialysis Patients (NCT03525223)
• Reliability: unpublished (design only)
• Key point: Testing whether moderate reduction of dialysate sodium from 138 to 135 mmol/L lowers tissue sodium

Source 25

• Source: Aquaporin-10 represents an alternative pathway for glycerol efflux from human adipocytes
• Author/Year: Laforenza et al., 2013, PLOS One / PMC3558521
• Reliability: high
• Key point: Humans express both AQP7 and AQP10 in adipocytes; silencing AQP10 cut glycerol permeability ~51% and osmotic water permeability ~46%