This is a case report with tentative conclusions and speculative mechanisms of action. The observed outcome may prove to be transient or to have a different cause. The oral estradiol may have been one factor around several. The mechanism of action is plausible and fits with documented and understood principles, but put together in a slightly unusual way. There are other plausible mechanisms of action, which are a mess because the liver and first-pass metabolism of estrogens contain multitudes. Do not use anything here to make clinical decisions, but perhaps it may spark useful clinical questions.
The patient is a 45-year-old woman who is on a full replacement dose of estradiol (hereafter E2) taken orally. She has been on a well-formulated ketogenic diet for some time, and has experience being on ketogenic diets in the past. Her adherence to the ketogenic diet is confirmed with extensive logging. She is following a ketogenic diet with the intention of losing weight, and is losing weight. Her presenting concern is being hypoketotic, i.e. showing serum levels of betahydroxybutyrate (BHB) as measured through capillary blood of less than 0.5mmol/L despite adherence. She occasionally is ketotic (BHB greater than or equal to 0.5mmol/L), but surprisingly reports that she is more likely to be ketotic when fed than when fasted, and that she almost always wakes up hypoketotic.
The patient's hypothesis was that something was inhibiting lipolysis, which could have been checked by examining BHB and NEFA (free fatty acids) at the same time, however this testing was not performed. If lipolysis were impaired but ketogenesis were intact, we would expect to see low NEFA and consequently low BHB. We could also see defects in or downregulation of BDH1 causing a deranged AcAc:BHB ratio, with higher levels of acetoacetate (AcAc) in proportion to BHB, but this also was not tested. The patient additionally queried cortisol driving gluconeogenesis in a diurnal pattern, disrupting ketogenesis, and elevated insulin. No gross cortisol elevation was observed, however elevated fasting insulin was observed. Along with elevated fasting insulin, extremely low IGF-1 (z-score -2.5, if I recall correctly) was observed.
Low IGF-1 could be explained in part by adherence to a ketogenic diet itself, which is a known phenomenon, however the level was low enough to raise further questions. IGF-1 is produced in the liver in response to growth hormone (GH), and low GH would explain lower IGF-1. If there were a GH deficiency, that might also explain impaired ketogenesis, as GH pulses to the liver are one of the factors that controls ketogenesis, as observed in people with GH deficiency, and with increased rates of ketosis being associated with increases in GH treatment. There are many factors that contribute to ketogenesis, but this is one.
Before doing any investigation of potential pituitary or hypothalamic investigation, however, it was noticed that the patient was on oral E2. Oral estrogens going through first-pass hepatic metabolism go through a variety of different processes and have a variety of different effects compared to circulating estrogens. These effects are significant, and it is noted that weight gain, impaired fat oxidation, and lower IGF-1 are often caused by oral estrogens, and this effect is largely attributed to first-pass metabolism in the liver.
The underlying cause of depressed IGF-1 in the liver when oral estrogens are given seems to be that the GH receptors in the liver are antagonized, so that the liver becomes at least partially GH insensitive or resistant. This means that while GH pulses may occur as normal, the liver at least in part does not respond to them. There are many processes in the liver that occur in response to GH, including the secretion of IGF-1, and as already mentioned, some set of processes that I will not claim to understand that affect the rate of ketogenesis. When the liver produces IGF-1, this provides feedback to inhibit the release of GH, maintaining a homeostatic proportion of available IGF-1 and GH. What you observe with oral estrogens is that GH levels rise and IGF-1 levels fall.
My intuition is that through this mechanism or others, oral estrogens may impair ketogenesis. This may actually present as though it were a defect in ketogenesis proper, but without a genetic origin. You would expect to see, when fasting, elevated NEFA from functioning lipolysis, but lower than expected BHB. Depending on where the disruption is, you might see changes in AcAc, but I proclaim my ignorance of those processes.
This patient elected to switch from oral E2 to an injectable form, and after one or two weeks, we observed extremely significant improvements of ketogenesis. Where morning fasting levels of capillary BHB had averaged around 0.4mmol/L, they now averaged around 1.0mmol/L. Moreover, while the patient had suspected that fasting ketogenesis was particularly impaired, and that fed ketogenesis was intact, we observed changes in fed ketogenesis and daily maxima. Where daily highs had been ketotic, they were marginally so, averaging 0.6mmol/L, and with a maximum observed this year of 1.7mmol/L. After switching the E2 route, the patient observed an average daily maximum of almost 2.0mmol/L, and an overall maximum of 3.7mmol/L. This is a significant change, and the patient intends to remain with injectable E2.
Interestingly, the patient's weight loss had been relatively stable on a ketogenic diet, which fits with some elements of the presentation and hypothesis, but challenges others. Lipolysis is driven and inhibited by a complex array of factors, but among them are GH (drive) and insulin (inhibition). The pattern expected with oral estrogens would be to have elevated GH due to inhibited IGF-1, with the liver but not adipocytes being resistant to GH. We expect elevated fasting insulin, unless adipocytes are insulin resistant, to impair lipolysis. The patient may have, in fact, been experiencing elevated lipolysis but low energy availability due to impaired ketogenesis, leading to weight loss with additional malaise, and this is what the patient reports: steady weight loss, but malaise except on the occasions of being euketotic. The patient reports possibly slowed weight loss after changing the route of estrogen administration, but increased well-being and feelings of energy alongside increased ketosis. Any change in fasting insulin has not been observed, however IGF-1 and insulin typically balance one another to some degree, and it may be that insulin would fall either directly due to increased IGF-1, or due to functioning ketogenesis and downstream effects thereof, or other downstream effects of restored hepatic GH sensitivity.
It is already known that oral estrogens may be replaced with estrogens administered via another route in cases of GH deficiency, in order for the hepatic functions that respond to GH to proceed, and it is known that one hepatic process which responds to GH is ketogenesis. This case report raises the question of to what degree ketogenesis depends upon hepatic GH sensitivity, and to what degree oral estrogens impair ketogenesis through this and/or other pathways. I wonder, too, about what drives well-being on a ketogenic diet: total energy availability matters, and weight loss without the energy being available is probably hard to sustain; likewise, we know that GH is responsible for some feelings of bodily well-being, and I wonder how much of this is direct, and how much might be mediated through hepatic processes. Does this patient feel better on a ketogenic diet simply because of the rise in BHB, or does restored hepatic GH sensitivity play some other role, too?
I note that there are elements of the case presentation that have been changed and values that have been rounded both to not imply excess precision (no serious statistical analyses have been done) and to protect the patient's privacy. I will say that the patient had separately wondered whether oral E2 was reducing feelings of well-being and impairing adherence to and efficacy of a ketogenic diet previously, and produced logs to that effect going back several years. The intuitions we have are fascinating, and if the patient had had access to more comprehensive lab testing to confirm lipolysis, ketogenesis, etc., at an earlier date, this may have been found, and found conclusively, much sooner. Likewise, these outcomes are preliminary, and further investigation of this patient may present additional individual variations of relevance. Those genes that are known to be involved in these processes so far show no gross high-penetrance monogenic defects. As with all dietary interventions and weight loss, long-term results may differ, too. Still, I found this too tantalizing not to share as a preliminary head-scratcher case with an interesting convergence of known effects showing up as a distinctive clinical picture. If the hypotheses here are even approximately right, this may also not be a very uncommon clinical picture, differing perhaps in degree.
Some interesting and relevant literature:
Hypoketonemia and age-related fasting hypoglycemia in growth hormone deficiency
Oral estrogen antagonizes the metabolic actions of growth hormone in growth hormone-deficient women
No comments:
Post a Comment