It's not just insulin.  This is the most likely explanation for those
who do not respond as well to low carb as others, in terms of appetite
suppression and ease of weight loss:

http://jcem.endojournals.org/cgi/content/abstract/87/8/3984

Response of the Hypothalamic-Pituitary-Adrenocortical Axis to
High-Protein/Fat and High-Carbohydrate Meals in Women with Different
Obesity Phenotypes

Valentina Vicennati, Luana Ceroni, Lorenza Gagliardi, Alessandra
Gambineri and Renato Pasquali
Endocrine Unit, Department of Internal Medicine and Gastroenterology, S.
Orsola-Malpighi Hospital, University Alma Mater of Bologna, 40138
Bologna, Italy

Address all correspondence and requests for reprints to: Renato
Pasquali, M.D., Unità di Endocrinologia, Dipartimento di Medicina
Interna e Gastroenterologia, Policlinico S. Orsola-Malpighi, Via
Massarenti 9, 40138, Bologna, Italy. E-mail: . rpasqual@almadns.unibo.it

Abstract

Subjects with abdominal obesity are characterized by hyperactivation of
the hypothalamic-pituitary-adrenal (HPA) axis. Food intake, particularly
at noon, is a well-known inducer of HPA axis activation. Whether obese
subjects present an abnormal response to meals containing different
macronutrient proportions is at present unknown. Therefore, this study
was carried out to investigate the effect of a high-lipid/protein meal
(HLP-meal) and a high-carbohydrate meal (HCHO-meal) on the HPA axis
activity in women with different obesity phenotypes. Nondepressed,
noncomplicated obese (body mass index greater than 28 kg/m2) women with
abdominal (A-BFD) (n = 10) and peripheral body fat distribution (P-BFD)
(n = 9) and a group of 11 normal-weight controls were investigated in
the follicular phase of the menstrual cycle. They were randomly given an
800-kcal HCHO-meal (containing 89% carbohydrates, 11% proteins, 0%
lipids), and an 800-kcal HLP-meal (containing 53% lipids, 43% proteins,
4% carbohydrates), which were eaten within 15 min at noon, with an
interval of 2 d between each meal. Blood samples for ACTH, cortisol,
glucose, and insulin were obtained at 15-min intervals before and after
each meal. Baseline hormone and glucose concentrations in the three
groups were similar.

*******After the HLP-meal, ACTH tended to similarly but insignificantly
increase in all groups, whereas cortisol increased significantly (P <
0.05) in the P-BFD group and insignificantly in the other groups.
Conversely, both ACTH and cortisol significantly (P < 0.05) increased
only in the A-BFD group, without any significant changes in both
controls and P-BFD women.***************

The analysis of the interaction between meals and groups clearly
indicated that the cortisol response to the HLP-meal and the HCHO-meal
was significantly different (P < 0.025) between the two obese groups,
the A-BFD group being characterized by a significantly lower response to
the HLP-meal and a significantly higher response to the HCHO-meal,
compared with the P-BFD group. Considering all groups together and after
adjusting for body mass index, a highly significant relationship was
found between cortisol-area under the curve and ACTH-area under the
curve after each meal test. However, no relationships were found between
changes in ACTH and cortisol and those of glucose, insulin, and the
glucose:insulin ratio after each meal.

Therefore, our data demonstrate that the response of the HPA axis to
meals containing different macronutrient proportions may depend on the
pattern of body fat distribution. We also suggest that the activation of
the HPA axis following the ingestion of large amounts of carbohydrates
may have some pathophysiological relevance, specifically in women with
the abdominal obesity phenotype.

--------------------