Holding carbohydrate high during four days of low energy availability halved the drop in IGF-1

Both conditions ran at 25 kcal per kilogram of fat free mass per day, with fully controlled diets and daily exercise expending 20 kcal per kilogram of fat free mass. The only thing that changed was carbohydrate availability, 1.5 versus 4.0 grams per kilogram per day.

IGF-1 fell in both conditions but by different amounts. The drop was 22.9 plus or minus 29.6 ng/mL on high carbohydrate against 48.5 plus or minus 26.2 ng/mL on low carbohydrate (p=0.018).

Body mass fell about 1.3 percent regardless of condition (p<0.001). Insulin, leptin and testosterone all fell further on low carbohydrate, but none of those condition by time interactions reached significance.

High carbohydrate availability also softened the reduction in submaximal and maximal blood lactate. Peak oxygen uptake and maximal power output were unaffected by either the energy restriction or the carbohydrate manipulation.

The numbers

IGF-1 change, high carbohydrate availability	-22.9 plus or minus 29.6 ng/mL
IGF-1 change, low carbohydrate availability	-48.5 plus or minus 26.2 ng/mL
Difference between conditions	p=0.018
Body mass change, both conditions	-1.3%, p<0.001
Peak oxygen uptake and maximal power	unaffected by either condition

Why this might happen

Proposed by the authors This is the explanation the authors offer in their discussion. This study did not test it.

The authors' framing is that low energy availability is usually confounded with low carbohydrate availability, and that separating the two splits the response in half.

Their New and Noteworthy statement says that maintaining high carbohydrate availability during short term low energy availability attenuates reductions in IGF-1, and that the modulatory effect of carbohydrate on substrate use differs between sexes.

They read the preserved lactate response as partial preservation of glycolytic function, and they state that shifts in substrate utilisation under low energy availability are largely independent of carbohydrate availability.

So the proposal is that carbohydrate availability decouples the endocrine arm of the low energy availability response from the metabolic arm. The step itself was not tested here.

Drawn from New and Noteworthy statement, full text paywalled.

What this does not show

It does not show that protecting IGF-1 protects anything a person would notice. Four days is far shorter than the months over which the bone, menstrual and immune consequences of low energy availability appear. A hormone moving in a better direction is not an outcome.
The endocrine claim rests on one hormone. Insulin, leptin and testosterone all moved in the expected direction, but their interaction terms did not reach significance. IGF-1 is doing the work here on its own.
The reported sex difference is underpowered. Nine female and seven male participants split into a subgroup comparison, with no menstrual phase control described in the abstract.
These were not athletes. Recreationally active adults on a controlled feeding protocol, not a training population in a competitive season.

Where this leaves us

Some of what has been attributed to energy deficit alone looks like it belongs to carbohydrate deficit specifically, which refines a decade of low energy availability work that did not separate the two.

Recreationally active adults, nine women and seven men, across four days of controlled underfeeding with prescribed exercise.

A longer trial carrying the same separation into weeks, with bone turnover or menstrual outcomes attached, would show whether the IGF-1 difference matters.

Caveats worth holding

Sixteen participants across two four day conditions.
Insulin, leptin and testosterone are reported as moving further on low carbohydrate, but without significant interaction terms those differences are not established.
The full text is paywalled. The mechanistic reading here comes from the journal's New and Noteworthy statement rather than the discussion.