6  The Adipose Tissue Expandability Hypothesis and Ectopic Fat

The idea that ties together everything so far is the adipose tissue expandability hypothesis (Virtue and Vidal-Puig 2010).

6.1 What the hypothesis says

Virtue and Vidal-Puig summarise the hypothesis as follows (Virtue and Vidal-Puig 2010).

  • The key link between energy excess and type 2 diabetes is not obesity itself but the limit of adipose tissue’s capacity to expand.
  • Everyone has an upper limit to their adipose tissue expansion capacity, determined by both genes and environment.
  • Once that limit is reached, adipose tissue can no longer store energy efficiently, and lipid begins to accumulate in other tissues.
  • Lipid accumulation in cells other than adipocytes causes “lipotoxicity” such as insulin resistance, cell death and inflammation.

flowchart TB
  E["Surplus energy"] --> Q{"Does adipose tissue<br/>have spare<br/>expansion capacity?"}
  Q -->|"Yes"| S["Stored in adipose tissue<br/>(especially subcutaneous,<br/>gluteofemoral)"]
  Q -->|"No<br/>(limit reached)"| O["Lipid spills over<br/>into other tissues"]
  O --> L["Liver"]
  O --> M["Skeletal muscle"]
  O --> X["Other organs"]
  L --> R["Lipotoxicity<br/>insulin resistance,<br/>cell death, inflammation"]
  M --> R
  X --> R

The content of Chapters 3–5 fits well with this hypothesis.

  • Subcutaneous progenitors accumulate lipid well and proliferate well (a large reservoir) (Tchkonia et al. 2005).
  • Gluteofemoral fat protects against the harm of ectopic fat by storing fatty acids long-term (Manolopoulos et al. 2010).
  • When progenitors’ capacity for renewal declines with ageing, fatty acids can no longer be sequestered (Tchkonia et al. 2010).
  • Japanese people have less subcutaneous fat and relatively more ectopic fat (菅波孝祥 2024).

6.2 Childhood obesity and the “number” of adipocytes

One of the factors governing expansion capacity is the number of adipocytes and the capacity to make new adipocytes. So, does it make a difference whether one was overweight as a child?

6.2.1 Cell number is set during childhood to adolescence

  • In a study of 288 individuals aged 4 months to 19 years (110 obese, 178 non-obese), obese children began to deviate from normal development from around 1 year of age, with both cell number and cell size increasing earlier and faster, and the difference was maintained throughout 4 years of follow-up. By age 11, obese children exceeded the mean cell number of non-obese adults (Knittle et al. 1979).
  • In adults, adipocyte number is nearly constant in both lean and obese people and does not decrease even with substantial weight loss. From this, cell number is thought to be set during childhood to adolescence (Spalding et al. 2008).

6.2.2 The “many cells” type is metabolically favourable

  • A study of subcutaneous fat in 764 people, which quantified for a given body fat mass whether a person had the “many small cells” type (hyperplastic) or the “few large cells” type (hypertrophic), reported that
    • this morphology was independent of sex and body weight, and correlated with fasting insulin and insulin sensitivity (the more hyperplastic, the better)
    • the more hyperplastic, the higher the total cell number, and in the hypertrophic type the number of new adipocytes generated per year was 70% lower
    • adipocyte death rate (about 10% per year) and mean age (about 10 years) were unrelated to morphology
    (Arner et al. 2010).
  • It has also been reported that people with hypertrophic morphology have higher fasting insulin and insulin resistance (HOMA-IR) at the same body fat mass, but show greater improvement in insulin sensitivity with weight loss (both moderate diet-induced weight loss and large weight loss after gastric bypass surgery) (Eriksson-Hogling et al. 2015).

6.2.3 So do “people who were overweight as children have greater expansion capacity”?

WarningThere is no direct evidence here

From the studies above, one can reason along the line “childhood obesity → more cells → hyperplastic type → larger reservoir,” but at the time this material was written, we could not find any study directly showing that a history of childhood obesity itself increases expansion capacity (metabolic protection) in adulthood.

Points to note:

  • Because cell number does not decrease even with substantial weight loss (Spalding et al. 2008), the increased cell number could also be said to remain as a foundation that makes it easier to store fat.
  • Whether a person is hyperplastic or hypertrophic was determined independently of sex and body weight and was strongly related to differences in the capacity to make new cells in adulthood (Arner et al. 2010). It is reasonable to think that expansion capacity depends not only on childhood cell number but also on adipogenic capacity in adulthood (determined by genes and environment (Virtue and Vidal-Puig 2010)).
  • Therefore, the conclusion that “it is better to have been overweight as a child” cannot be drawn from these studies.

6.3 An extreme example: lipodystrophy

Lipodystrophy is a group of diseases in which adipose tissue is lost, congenitally or acquired (Garg 2004).

  • Severe lipodystrophy is accompanied by leptin deficiency, insulin resistance, hypertriglyceridaemia and fatty liver (Oral et al. 2002).
  • In a study in which 9 women with lipodystrophy and leptin deficiency received recombinant leptin (now metreleptin) for 4 months, HbA1c fell by 1.9% (in the 8 with diabetes), triglycerides fell by 60% on average, and liver volume decreased by 28% on average (Oral et al. 2002).

Lipodystrophy shows that “metabolism also worsens when there is too little adipose tissue,” and together with the observation that loss of gluteofemoral fat increases metabolic risk (Manolopoulos et al. 2010), it is regarded as an example supporting the expandability hypothesis (Virtue and Vidal-Puig 2010).

6.4 Drugs can also change “where”: thiazolidinediones

  • Thiazolidinediones are high-affinity ligands (activators) of PPARγ, a nuclear receptor involved in adipocyte differentiation, and promote differentiation of progenitors into adipocytes (Lehmann et al. 1995).
  • In a study in which 13 patients with type 2 diabetes received pioglitazone for 16 weeks, body weight (82 → 85 kg) and fat mass (27 → 30 kg) increased, yet hepatic and peripheral insulin sensitivity improved. On MRI, subcutaneous fat area increased (301 → 342 cm²) and visceral fat area decreased (144 → 131 cm²), with fat distribution shifting from visceral to subcutaneous (Miyazaki et al. 2002).

This is an example consistent with the expandability hypothesis: “expanding the subcutaneous reservoir reduces visceral fat and improves metabolism.”

WarningA caution on interpretation

Thiazolidinediones have important adverse effects such as fluid retention and worsening of heart failure (Nesto et al. 2003) and increased fracture risk (Loke et al. 2009), so this does not mean “anything that increases subcutaneous fat is good.” Please understand it strictly as an example showing that the distribution of fat matters for metabolism.

6.5 What happens with weight loss?

A systematic review of weight-loss studies that measured visceral and abdominal subcutaneous fat by imaging (61 studies) (Chaston and Dixon 2008) showed that

  • with moderate weight loss, visceral fat is preferentially lost, but this tendency weakens as the amount of weight lost increases
  • there is little difference by method of weight loss (diet or exercise), and with very-low-calorie diets visceral fat is especially preferentially lost only in the short term (less than 4 weeks)

The authors note that this may explain why even modest weight loss improves metabolism (Chaston and Dixon 2008).

6.6 Summary

  • The key to metabolic abnormalities is not obesity itself but the limit of adipose tissue’s expansion capacity. Beyond that limit, lipid spills over into other tissues and causes lipotoxicity (Virtue and Vidal-Puig 2010).
  • Lipodystrophy (no adipose tissue → severe metabolic abnormalities, improved by leptin replacement) (Oral et al. 2002; Garg 2004) and thiazolidinediones (more subcutaneous fat, less visceral fat → improvement) (Miyazaki et al. 2002) are examples supporting the hypothesis from both sides.
  • With moderate weight loss, visceral fat is preferentially lost (Chaston and Dixon 2008).
  • Adipocyte number is set during childhood to adolescence and increases early in obese children (Knittle et al. 1979; Spalding et al. 2008). In adults, the more hyperplastic (“many small cells”) the morphology, the better the metabolic profile and the higher the capacity to make new cells (Arner et al. 2010). However, at the time this material was written, no direct evidence was found that a history of childhood obesity increases expansion capacity.