Study review: Dietary sugars, not lipids, drive hypothalamic inflammation

Study background

Today’s post briefly reviews a neat little study in mice from 2017 by Gao et al. [1]. It’s called “Dietary sugars, not lipids, drive hypothalamic inflammation”. Isn’t it nice to a have a clear study title giving away the ending?

More and more scientific papers now have a graphical and text abstract, as does this paper.

This graphical abstract illustrates how, in mice, dietary fat and carbohydrate individually and synergistically contribute to the production of advanced-glycation end-products (AGEs). In turn, these AGEs can make microglia ‘reactive’. This reactivity is synonymous to the inflammatory response observed in the hypothalamus of these mice made obese through dietary means.

Microglia make up about 10-15% of cells in the brain and are the resident macrophages, meaning that much of their job is to affect immune responses. AGEs are the normal result of metabolic processes in the body, like when metabolizing fat, carbohydrate or protein, but can contribute to disease when in excess.

Briefly, the study comes off the back of years and years of research pointing to a link between brain inflammation and diet-induced obesity. Specifically, the inflammation concerns a part of the brain called the hypothalamus, a region of the brain involved in controlling metabolism, the autonomic nervous system, hunger, body temperature, fatigue, thirst, circadian rhythms and sleep amongst other things.

The hypothalamus is of big interest to obesity researchers. This is no surprise knowing you can stop an animal from feeding by sectioning parts of the hypothalamus [2] or get it to overeat by stimulating others [3].

What this study is not about

The field of nutrition science is plagued with studies overreaching in their answers. A big reason for this is that they’re often designed to ask questions not suitable for the answers they portend to seek.

This study does not explore whether obesity results from an inflamed hypothalamus or if an inflamed hypothalamus causes obesity

What this study tried to test

Like all good studies, this one started out with a prediction. Let me paraphrase:

diets high in both fats and carbohydrates (sugars) create excessive AGEs (advanced-glycation end-products) that inflame the hypothalamus

As the study title suggests, their prediction panned out, the mice got an inflamed hypothalamus.

But was it the combination of fats and sugars themselves creating lots of AGEs that went on to inflame the hypothalamus? Or was it something else doing it, like obesity per se?

To make sure the authors weren’t fooling themselves, they tried to control variables relating to diet and inflammatory AGEs pathways.


The authors based the diet they gave their mice on the poetically named D12331 diet [4] containing 35.8% fat, 35.5% carb, 23% protein. The rats could eat ad lib, meaning as much as they wanted to for four weeks regardless of the diet they were on.

Based on a 5,558.5 kcal diet of D12331 there are

  • 700 kcals from sucrose (table sugar) and 780 kcals from maltodextrin (used in sport gels for e.g.)
  • 225 kcals from soybean oil (used in high-volume restaurant cooking) and 3,001 kcals from coconut oil (a highly saturated fat)
  • 912 kcals from casein (a dairy protein)

Below is a table taken from the study detailing the 5 diets

  • (low-fat) control diet with carbs coming from wheat and corn
  • 2 HCHF diets (one with lots of sucrose [HCHF1], the other with lots of starch [HCHF2])
  • 2 LCHF diets (one with moderate protein [LCHF1],one very low in protein [LCHF2])

AGEs and AGE receptors

The authors wanted to control the inflammatory AGE pathways to make sure it was responsible for the hypothalamic inflammation and not some other phenomenon.

The idea is simple. Your normal mice with all their AGE receptors intact eat a diet that inflames their hypothalamus.

To be sure the inflammation is mediated through AGE pathway (or mechanism), create mice lacking certain AGE receptors (mice-/-), feed them the same diet, and if their hypothalamus doesn’t get inflamed then you’ve identified AGEs as the relevant pathway.

However, if they do get inflamed despite the AGE receptors being absent, then the inflammation can’t be happening through the AGE pathway (since you’ve taken it out of the mice!).

  • standard C57BL/6 mice with the ALCAM and RAGE genes
  • C57BL/6 mice without 2 copies of the RAGE gene (RAGE-/-)
  • C57BL/6 mice without 2 copies of the ALCAM gene (ALCAM-/-)
  • and C57BL/6 mice without 2 copies of the RAGE and ALCAM genes, respectively (RAGE-ALCAM-/-)

The authors also injected the hypothalamus of the mice with 2 AGEs, CML (N(6)-carboxymethyl lysine) and MGH (methylglyoxal), in order to reproduce the inflammation mechanism they’d originally seen with the diet. It’s sort of like double-checking the mechanism.

Summary of results

The HCHF diet, but not LCHF diet, induced obesity and increased inflammation in the hypothalamus of those mice.

As you can see from the figure below, all mice eating ad lib on the 4 non-control diets ate more calories, but the mice eating the HCHF1 diet ate the most calories on average. HCHF1 and 2 mice gained more body-weight by week 4 than the mice eating diets LCHF1 and 2.

Compared to their low-fat control counterparts, the mice eating the LCHF2 diet lost more body-weight by week 4 but ended up with a higher percentage of body fat. The latter is probably due to the lower protein content in the LCHF2 diet.

What’s really interesting about these results is that the HCHF and LCHF groups ate very similar amounts of total fat, suggesting it was the absence of carbohydrates (not simply calories per se) that associates with a lack of hypothalamic inflammation. More specifically, when the carbs were lowered, the authors observed “quiescent microglia”, a kind of brain cell in a non-inflamed state.

Furthermore, the authors found that, yes, any of their high-carb diets was associated with more of the AGE called CML, but that the worse combination was a diet high in carbs and high in fat. Visible in the image below, mice on the HCHF diets had a greater expression of the AGE receptor genes RAGE and ALCAM.

Interestingly enough, there didn’t appear to be a difference in hypothalamic inflammation whether the carbs were derived from sucrose or starch.

To reiterate the main findings of this study I’ll quote the authors (with my emphasis)

“The present experiments provide clear evidence that the amount of carbohydrate in a diet determines whether a high-fat diet will induce obesity, or not


I want to end by quoting a sentence from the paper (with my emphasis)

“Restricting dietary fat is not the only factor that needs to be considered for body weight reduction in obese individuals and that dietary carbohydrates might substantially gate the efficiency of calorie restriction for body weight reduction, via a hypothalamic mechanism”

I like this sentence because it recognizes the patently obvious, that ‘calories matter’, whilst still paying attention to biological differences in how calories are handled.

The data from this study (and others), suggests that reversing obesity is probably easier on a diet with less carbohydrate in it rather than more. There’s a parallel argument to be made for protein but this is a discussion for another time.

5 comments On Study review: Dietary sugars, not lipids, drive hypothalamic inflammation

  • I’m not getting the correct study with the link provided. I’m using this one

  • Nice to hear from you again N John Appleton!

    You’re right on both points, casein and isolated soybean oil. Recall that this study look at inflammation via anti-body staining of certain AGEs only in the hypothalamus of mice – this means if inflammation manifested elsewhere in the brain or body then they would’ve missed it. The conclusions drawn by the authors regarding the inflammation of the hypothalamus seems pretty solid, but that doesn’t mean that the soybean oil or casein containing diets didn’t cause problems elsewhere (for e.g., they weren’t looking at hs-CRP, a systemic marker of inflammation)

  • Looking forward to “the parallel article on PROTEIN.” Thank you for your work and generous sharing. Now if you only had transcripts for your podcast…my “lay” brain would be her happy. You do an excellent job of making science accessible to people like me.

    • Hi Celeste,

      Yes I haven’t forgotten your request! 🙂 The protein/gluconeogenesis article and future podcast is in the works.

      I’m glad we’re making the raw science more digestible for listeners like yourself. It’s very encouraging to know you appreciate it.

Leave a reply:

Your email address will not be published.