Available online at www.sciencedirect.

com

ScienceDirect
Human brain responses to gastrointestinal nutrients and
gut hormones
John T McLaughlin1 and Shane McKie2
Functional mapping of human brain activation has made it
possible to understand how different nutrients in the gut impact
on homeostatic and appetitive brain responses. Current data
are limited, but nutrient-specific effects are observed, with
differential responses to lipid and sugars. Responses are not a
simple function of calorie intake. Gut hormones such as CCK,
PYY, GLP-1 and ghrelin are implicated in these responses, but
may not exert effects directly on the brain. Research is now
addressing how these homeostatic signalling states (fasting/
fed) interact with hedonic responses, such as those evoked by
images of appealing food. Differences are also beginning to
emerge in obese versus lean subjects. These platforms will
enable a new understanding of normal and disordered eating
behaviours in humans.
Addresses
1
University of Manchester, Manchester Academic Health Sciences
Centre (MAHSC), and Gastrointestinal Centre, Institute of Inflammation
and Repair, Faculty of Medical and Human Sciences, University of
Manchester, UK
2
University of Manchester, Manchester Academic Health Sciences
Centre (MAHSC), and Neuroscience and Psychiatry Unit, Institute of
Brain, Behaviour and Mental Health, University of Manchester, UK
Corresponding author: McLaughlin, John T
(John.Mclaughlin@manchester.ac.uk)

Current Opinion in Pharmacology 2016, 31:812

This review comes from a themed issue on Gastrointestinal
Edited by Graham J Dockray and Andrea Varro
For a complete overview see the Issue and the Editorial
Available online 28th August 2016
http://dx.doi.org/10.1016/j.coph.2016.08.006
1471-4892/# 2016 Published by Elsevier Ltd.

immunohistochemistry and electrophysiological recordings, often in response to direct application of nutrients or

gut hormones. Selective afferent vagotomy studies in
rodents demonstrated that signals neurally transduced
from the gut are essential for nutrient-induced responses
in the brainstem and hypothalamus [2]. Crucially then,
homeostatic responses to feeding are not solely metabolic
post-absorptive responses to circulating nutrients, which
access the cerebrospinal fluid and brain without hindrance [3]. Rather, the CNS responds directly to luminal
gut content, prior to absorption. This has major implications for understanding feeding behaviour, and how to
approach consequences such as obesity.
This review will initially focus on recent advances in
imaging technologies, review the limited literature on gut
hormones and nutrients effects on the human brain, and
explore some of the potential interactions between homeostatic signals and higher domains being uncovered in
functional experiments.

Studies in humans: limitations and

opportunities
There are clearly major experimental limitations, yet only
human studies can answer fundamental questions about
the complex mechanisms leading to over-consumption or
under-consumption of food. Recently, the use of functional MRI scanning has begun to map out human brain
responses and interactions.
Conventional fMRI study models used in other areas of
neuroscience are task based, so in early studies this
involved simple consumption of a test meal. The technique is based on the change in blood oxygenation level
dependent (BOLD) signal, a marker of regional activity
since blood flow is altered in active brain regions allowing
mapping of activity in regions of interest.

The presence of nutrients in the gut inhibits appetite and

reduces food intake. The secretion of gut hormones in
response to luminal nutrients is pivotal to these responses
by signalling to the brain to influence both homeostatic
functions and appetitive behaviours. There are also hedonic and emotional drives to eat, opposing homeostatic
physiological pathways: how these interact is a focus of
current attention [1].

A further development is physiological (phys)MRI in

which a nutrient is infused into the gut after a short
baseline period and the change in BOLD signal over
time compared to the baseline period. This is analogous
to pharmacological challenge MRI (phMRI) which uses
psychoactive drug infusions instead of nutrient ingestions
therefore physMRI has the same advantages and disadvantages as phMRI [4].

Most experimental data on brain responses to nutrients

and hormones necessarily come from animal models that
have employed a variety of techniques for example

There are advantages of physMRI over conventional food

based fMRI tasks [5]. physMRI has the ability to map the
direct effect on the BOLD signal of the nutrients being

Current Opinion in Pharmacology 2016, 31:812

www.sciencedirect.com

Human brain responses to nutrients McLaughlin and McKie 9

Time effect for

glucose vs saline
ingestion in
hypothalamus

glucose vs saline ingestion

(mean %BOLD signal change)

Figure 1

2
1
0
1
2
3
4
5
6
1

11

13

15

17

19

21

23

25

Time from glucose ingestion (mins)

Current Opinion in Pharmacology

A representative image of the selected human brain region (hypothalamus, left panel) and time course of the BOLD-signal response to glucose
directly infused into the stomach. The response is always subtracted from a control condition in the same subjects, in this case saline.

ingested. As the nutrient-induced BOLD signal change is

slow in comparison to the 30 second blocks used in
conventional fMRI paradigms, nutrient-induced changes
will not be detected using fMRI tasks. Also, the fMRI
tasks are designed to target the hedonic regions of the
brain, such as the ventral striatum and basal ganglia,
rather than the more physiological brain regions such
as the brainstem and hypothalamus. With physMRI both
the reward and deep brain regions can be probed [6]. A
representative image of the effect of glucose is shown in
Figure 1.
For all fMRI, there is a low frequency signal drift due to
mechanical vibrations causing increases in temperature in
the gradient coil system [7]. For conventional task-based
fMRI, the drift can be filtered out of the time series
however as the nutrient-induced BOLD signal change is
slow, and in the same frequency space of the drift, then no
filtering can be applied to physMRI data. In order to
eradicate signal drift, a saline control scan is needed per
person, so that any drift can be modelled and subtracted
with respect to time per voxel. This leads to multiple
scans per person and therefore increases costs and time.
Another disadvantage is that physMRI, like fMRI, is a
non-quantitative measure (BOLD signal change is compared to a pre-ingestion baseline) so there is no direct
comparison with circulating nutrient or hormone concentrations and only temporal correlations are possible [8].
Though not yet applied to this field, MR acquisition
techniques such as multi-echo EPI [9] or arterial spin
labelling (ASL) [10] can be used to separate slow changing BOLD effects from drifts. ASL can also provide
quantitative information on cerebral blood flow and in
some instances arterial arrival time which can be used to
www.sciencedirect.com

provide a direct comparison with nutrient and hormone

levels.

Imaging brain responses to nutrients present

in the gut
The literature is currently small, mostly addressing glucose. Studies focussed on oral taste are not reviewed here.

Carbohydrate
A study by Liu et al. was one of the first to use fMRI
following eating, that is the ingestion of glucose solution. Decreased hypothalamic BOLD signals occurred
from around 7 minutes [11]. Subsequent studies have
reinforced these findings showing a dose-dependent
and prolonged decrease in BOLD signal in the hypothalamus following glucose [12]. A larger response was seen
following oral than intravenous glucose [13]. In addition
no hypothalamic BOLD decrease occurred following
artificial sweetener (aspartame) or non-sweet maltodextrin. This suggests activity is not due to sweetness [14]. It
may depend on the ability to release gut hormones or
affect gut function, which may not occur in response to
sweetness per se in humans [15,16].
These studies are potentially confounded by sensory
responses via oral tasting. Movement of the head and
neck during swallowing results in imaging artefacts, preventing analysis of key early time points.
More recently, detailed imaging of brain responses to
glucose have been investigated by physMRI [8]. In
particular, detailed imaging of the brainstem and hypothalamus and other regions of interest was investigated
immediately following intragastrically administered glucose. In line with previous observations BOLD signal
Current Opinion in Pharmacology 2016, 31:812

10 Gastrointestinal

decreased in the hypothalamus, medulla and pons beginning within 5 minutes of infusion. Very widespread clusters of activity were noted throughout the brain,
particularly the cerebellum, right occipital cortex, putamen and thalamus. As lipid responses had already been
studied with the CCK1 receptor antagonist dexloxiglumide [17] this was used again to demonstrate that
responses were almost all CCK-independent, whilst temporally correlated to the rise in glucose and insulin. A clear
exception was the motor cortex: the response to glucose
was abolished by dexloxiglumide. The significance of this
finding is not clear.
Recent fMRI studies have addressed responses to fructose, since there is evidence that fructose is less satiating
than glucose and is a weaker stimulant of gut hormone
secretion. These suggest differences in responses to
glucose. Fructose differentially increased activity of the
left amygdala, left hippocampus, right parahippocampus,
orbitofrontal cortex and precentral gyrus more than glucose [18]. Glucose had a positive effect on several brain
areas (left caudate and putamen, precuneus and lingual
gyrus) which was more marked than in response to
fructose. The authors correlated these dissociable
responses with greater insulin,GLP-1 and GIP release
in response to glucose than fructose. Brainstem and
hypothalamus responses were not reported. However
the investigators used 75 g glucose (standard tolerance
test dose) versus 25 g fructose (maximum tolerated).
Ascribing differential effects is not possible until matched
doses are studied.
Another fMRI study addressed differential responses to
glucose and fructose in lean and obese adolescents,
including homeostatic brain areas [19]. They detected
no changes in BOLD in response to glucose or fructose in
the hypothalamus or other brain areas of lean subjects, but
did so in obese participants. The opposite was true in the
executive prefrontal cortex where glucose exerted lesser
effects. Fructose caused stronger striatal effects in all
subjects, arguably driving stronger reward responses.
It was suggested that heightened responses in obesity
may drive further food intake.

Lipid
Using physMRI to investigate whole brain responses, we
have shown that intragastrically administered lipid (dodecanoic acid) increases BOLD signal in key human brain
areas, including the brainstem and hypothalamus [17].
Unlike glucose, lipid-activated effects were wholly abolished by the CCK1 receptor antagonist, dexloxiglumide,
and subsequently shown to be strongly suppressed by
ghrelin [20].
A recent novel approach was taken to reduce food intake
via a diet designed to enhance luminal short chain fatty
acid content [21]. Using food image responses the
Current Opinion in Pharmacology 2016, 31:812

investigators showed reduced activity in striatal pathways with an impact on eating behaviour. No changes in
gut hormones were however detected.

Protein and amino acids

No published studies have addressed these nutrients.

Does the brain sense caloric load, or are there

nutrient-specific responses?
One key drawback is that nutrients in the gut are unavoidably absorbed, circulating and accessing the brain. If
it is true, as widely thought, that calories are the key
sensed factor, then the pattern of responses to different
caloric nutrients would be predicted to be broadly the
same. Interpreting the studies described above, this is
clearly not so. Almost all studies using glucose have
shown a decrease in BOLD; lipid induces an increase,
with a different spatial matrix. Glucose, maltodextrin and
fructose effects differ also, whilst CCK antagonism disrupts activity selectively. Only 2 g of fatty acid was
needed to achieve effects, versus >50 g of glucose, further indicating that systemic calorie sensing cannot be
key. Currently this remains observational, but already has
clear implications for the design and conduct of future
studies, and interpretation of any literature dependent on
measuring calories.

Effects of gut hormones on the human brain

Very few studies have been undertaken. These have
necessarily used infused or injected peptides (in contrast
to direct CNS application that has been employed in
animals), and so it is impossible to know whether effects
are direct or indirect. Peptides do not readily cross the
blood brain barrier in general, but the barrier is leaky in
regions associated with homeostatic mechanisms. This
includes the hypothalamus and the brainstem via the area
postrema, part of the dorsal vagal complex. Therefore
circulating hormones and metabolic cues may be sensed
in these salient areas, potentially integrating signals from
the periphery.
The most studied are the L-cell peptides PYY and GLP1,
driven by their relevance to diabetes, and possible involvement in the effector mechanisms of bariatric surgery
[22].
A seminal study involved the infusion of PYY3-36 to
humans [23], with fMRI whole brain analysis. Caloric
intake was reduced, associated with a switch in activity
from homeostatic areas to more hedonic regions. No other
studies using PYY have been reported. GLP-1 is an
incretin hormone released in response to carbohydrate.
One PET study that measured regional cerebral blood
flow responses to a meal showed correlations between
GLP-1 responses and changes in blood flow to the hypothalamus and prefrontal cortex [24]. However, correlations to plasma levels of a single hormone are of limited
www.sciencedirect.com

Human brain responses to nutrients McLaughlin and McKie 11

value. The GLP-1 receptor agonist exenatide reduced

food intake and diminished brain responses to food
images in obese and diabetic subjects (insula, amygdala,
putamen, and orbitofrontal cortex), with effects blocked
by the GLP-1 receptor antagonist exendin 9-39 [25]. A
recent study using fMRI explored both endogenous
GLP-1 and an exogenous analogue, liraglutide, on
CNS responses to a meal, again using exendin 9-39, in
healthy and diabetic subjects [26]. Endogenous GLP-1
was shown to activate the insular cortex, and this was
reduced in type 2 diabetes. However, homeostatic brain
stem and hypothalamic areas were not reported in either
study so GLP-1s effects on these homeostatic areas
remain unknown yet may be crucial if signalling is vagally
transmitted.
No fMRI studies involving CCK infusion have been
reported, and it is not known whether it can cross directly
to the brain, although leaky regions of the blood-brain
barrier may be permissive, and a short CCK-4 fragment
does exert central effects [27]. However, the studies
above using the CCK1 receptor antagonist dexloxiglumide suggest a key role in the response to lipid but not
glucose [8,17] This is probably related to its action at
vagal afferent terminals. This may also be the case for
ghrelins modulatory effect on the effect of lipid on
BOLD signal as noted above [20].
Intriguingly, these factors may also be implicated in links
between mood, brain activation and nutrient (lipid)-induced signals [6]. The highly differential effects of dexloxiglumide on the CNS responses to glucose versus fatty
acid therefore reflect very specific patterns of activity
indicative of differential responses to nutrient classes.
The situation is further complicated in that glucose is
metabolically widely sensed in the brain [28]: how this
relates to or impacts on BOLD signalling is unknown.

food intake. The opposite had previously been shown for

ghrelin, increasing the responses to food cues in broadly
the same brain areas, and gastric banding also exerts
similar effects [5].

Summary
Although very limited human data are currently available,
there is now considerable technical potential to study the
effects of nutrients and the role of gut hormones on
human brain responses, and to apply these to a better
understanding of feeding behaviour in health and disease.
It is particularly important to move away from a model of
caloric sensing.
The key, urgent question is how to move this field
beyond current observation and phenomenology, and into
mechanistic experiments that truly explore the underlying neurobiology. This is a necessary step change towards
applying these methodologies to translational clinical
research.

Conflict of interest statement

Nothing declared.

References and recommended reading

Papers of particular interest, published within the period of review,
have been highlighted as:
 of special interest
 of outstanding interest
1.

Murray S, Tulloch A, Gold MS, Avena NM: Hormonal and neural

mechanisms of food reward, eating behaviour and obesity. Nat
Rev Endocrinol 2014, 10:540-552.

2.

Schwartz GJ: The role of gastrointestinal vagal afferents in the

control of food intake: current prospects. Nutrition 2000,
16:866-873.

3.

Schaeffer M, Hodson DJ, Mollard P: The bloodbrain barrier

as a regulator of the gutbrain axis. Front Horm Res 2014, 42:
29-49.

4.

Anderson IM, McKie S, Elliott R, Williams SR, Deakin JFW:

Assessing human 5-HT function in vivo with pharmacoMRI.
Neuropharmacology 2008, 55:1029-1037.

5.

Scholtz S, Miras AD, Chhina N, Prechtl CG, Sleeth ML, Daud NM,
Ismail NA, Durighel G, Ahmed AR, Olbers T et al.: Obese patients
after gastric bypass surgery have lower brain-hedonic
responses to food than after gastric banding. Gut 2014, 63:891902.

6.

Van Oudenhove L, McKie S, Lassman D, Uddin B, Paine P, Coen S,

Gregory L, Tack J, Aziz Q: Fatty acid-induced gut-brain
signaling attenuates neural and behavioral effects of sad
emotion in humans. J Clin Invest 2011, 121:3094-3099.

7.

Foerster BU, Tomasi D, Caparelli EC: Magnetic field shift due to

mechanical vibration in functional magnetic resonance
imaging. Magn Reson Med 2005, 54:1261-1267.



Ghrelin has recently been further studied [29 ]. It

exerted no effects on key regions during a control task,
but increased the appeal of high calorie food images,
effects replicated by fasting alone. This was paralleled
by orbitofrontal cortex activation on fMRI. Low calorie
and high calorie food images both induced greater hippocampal activation during fasting or ghrelin infusion
than a control infusion.
A single study has used infusions of PYY3-36 and GLP-1
separately and together [30]. Six regions of interest were
studied, not including brainstem or hypothalamus. Differential effects were seen. PYY3-36 reduced BOLD
signal responses to food images (insula, left nucleus
accumbens, left orbitofrontal cortex), broadly supporting
prior data [23]. GLP-17-36 amide reduced BOLD signal
in all 6 regions of interest, adding the amygdala, caudate
and putamen to this matrix. Combined infusions were
shown to be summative: only the combination reduced
www.sciencedirect.com

8.


Little TJ, McKie S, Jones RB, DAmato M, Smith C, Kiss O,

Thompson DG, McLaughlin JT: Mapping glucose-mediated gutto-brain signalling pathways in humans. Neuroimage 2014,
96:1-11.
The first detailed physMRI mapping of the whole spatiotemporal matrix of
human brain responses to glucose infused into the gut, obviating the prior
problems with movement and other artefacts.
Current Opinion in Pharmacology 2016, 31:812

12 Gastrointestinal

9.

Evans JW, Kundu P, Horovitz SG, Bandettini PA: Separating slow

BOLD from non-BOLD baseline drifts using multi-echo fMRI.
Neuroimage 2015, 105:189-197.

10. Pollak TA, De Simoni S, Barimani B, Zelaya FO, Stone JM,

Mehta MA: Phenomenologically distinct psychotomimetic
effects of ketamine are associated with cerebral blood flow
changes in functionally relevant cerebral foci: a continuous
arterial spin labelling study. Psychopharmacology 2015,
232:4515-4524.
11. Liu Y, Gao J-H, Liu H-L, Fox PT: The temporal response of the
brain after eating revealed by functional MRI. Nature 2000,
405:1058-1062.
12. Smeets PA, de Graaf C, Stafleu A, van Osch MJ, van der Grond J:
Functional MRI of human hypothalamic responses following
glucose ingestion. NeuroImage 2005, 24:363-368.
13. Smeets PA, de Graaf C, Stafleu A, van Osch MJ, van der Grond J:
Functional magnetic resonance imaging of human
hypothalamic responses to sweet taste and calories. Am J Clin
Nutr 2005, 82:1011-1016.
14. Smeets PA, Vidarsdottir S, de Graaf C, Stafleu A, van Osch MJ,
Viergever MA, Pijl H, van der Grond J: Oral glucose intake
inhibits hypothalamic neuronal activity more effectively than
glucose infusion. Am J Physiol Endocrinol Metab 2007,
293:E754-E758.
15. Steinert RE, Frey F, Topfer A, Drewe J, Beglinger C: Effects of
carbohydrate sugars and artificial sweeteners on appetite and
the secretion of gastrointestinal satiety peptides. Br J Nutr
2011, 105:1320-1328.
16. Bryant CE, Wasse LK, Astbury N, Nandra G, McLaughlin JT: Nonnutritive sweeteners: no class effect on the glycaemic or
appetite responses to ingested glucose. Eur J Clin Nutr 2014,
68:629-631.
17. Lassman DJ, McKie S, Gregory LJ, Lal S, DAmato M, Steele I,
Varro A, Dockray GJ, Williams SCR, Thompson DG: Defining the
role of CCK in the lipid induced human brain activation matrix.
Gastroenterology 2010, 138:1514-1524.
18. Wolnerhanssen BK, Meyer-Gerspach AC, Schmidt A, Zimak N,
Peterli R, Beglinger C, Borgwardt S: Dissociable behavioral,

physiological and neural effects of acute glucose and fructose
ingestion: a pilot study. PLOS ONE 2015, 10:e0130280.
The first evidence that glucose and fructose produce different patterns of
brain response, albeit confounded by the use of different test sugar doses.

21. Byrne CS, Chambers ES, Alhabeeb H, Chhina N, Morrison DJ,

Preston T, Tedford C, Fitzpatrick J, Irani C, Busza A, Garcia
Perez I, Fountana S, Holmes E, Goldstone AP, Frost GS:
Increased colonic propionate reduces anticipatory reward
responses in the human striatum to high-energy foods. Am J
Clin Nutr 2016. [Epub ahead of print].
The only study to date demonstrating an effect of short chain fatty acids
on relevant aspects of brain function.
22. Abdeen G, le Roux CW: Mechanism underlying the weight loss
and complications of Roux-en-Y gastric bypass. Obes Surg
2016, 26:410-421.
23. Batterham R, Ffytche D, Rosenthal J, Zelaya FO, Barker GJ,
Withers DJ, Williams SC: PYY modulation of cortical and
hypothalamic brain areas predicts feeding behaviour in
humans. Nature 2007, 450:106-109.
24. Pannacciulli N, Le DS, Salbe AD, Chen K, Reiman EM,
Tataranni PA, Krakoff J: Postprandial glucagon-like peptide-1
(GLP-1) response is positively associated with changes in
neuronal activity of brain areas implicated in satiety and food
intake regulation in humans. Neuroimage 2007, 35:511-517.
25. van Bloemendaal L, Veltman DJ, Ten Kulve JS, Groot PF,
 Ruhe HG, Barkhof F, Sloan JH, Diamant M, Ijzerman RG: Brain
reward-system activation in response to anticipation and
consumption of palatable food is altered by glucagon-like
peptide-1 receptor activation in humans. Diabetes Obes Metab
2015, 17:878-886.
A mechanistic role for GLP-1 in brain responses anticipation of food
reward and intake is demonstrated using a GLP-1 receptor antagonist
peptide infusion.
26. Ten Kulve JS, Veltman DJ, van Bloemendaal L, Groot PF,
 Ruhe HG, Barkhof F, Diamant M, Ijzerman RG: Endogenous GLP1
and GLP1 analogue alter CNS responses to palatable food
consumption. J. Endocrinol 2016, 229:1-12.
The same group provided further evidence of GLP-1s role in the regulation of the brain reward systems to palatable foods.
27. Zwanzger P, Domschke K, Bradwejn J: Neuronal network of
panic disorder: the role of the neuropeptide cholecystokinin.
Depress Anxiety 2012, 29:762-774.
28. Levin BE, Magnan C, Dunn-Meynell A, Le Foll C: Metabolic
sensing and the brain: who, what, where, and how?
Endocrinology 2011, 152:2552-2557.

19. Jastreboff AM, Sinha R, Arora J, Giannini C, Kubat J, Malik S, Van

Name MA, Santoro N, Savoye M, Duran EJ, Pierpont B, Cline G,
Constable RT, Sherwin RS, Caprio S: Altered brain response to
drinking glucose and fructose in obese adolescents. Diabetes
2016. db151216.

29. Goldstone AP, Prechtl CG, Scholtz S, Miras AD, Chhina N,

 Durighel G, Deliran SS, Beckmann C, Ghatei MA, Ashby DR,
Waldman AD, Gaylinn BD, Thorner MO, Frost GS, Bloom SR,
Bell JD: Ghrelin mimics fasting to enhance human hedonic,
orbitofrontal cortex, and hippocampal responses to food. Am
J Clin Nutr 2014, 99:1319-1330.
fMRI studies demonstrate that ghrelin infusion induces the same
responses as fasting in appetitive and hedonic brain regions.

20. Jones RB, McKie S, Astbury N, Little TJ, Tivey S, Lassman DJ,
McLaughlin J, Luckman S, Williams SR, Dockray GJ,
Thompson DG: Functional neuroimaging demonstrates that
ghrelin inhibits the central nervous system response to
ingested lipid. Gut 2012, 61:1543-1551.

30. De Silva A, Salem V, Long CJ, Makwana A, Newbould RD,

Rabiner EA, Ghatei MA, Bloom SR, Matthews PM, Beaver JD,
Dhillo WS: The gut hormones PYY 3-36 and GLP-1 7-36 amide
reduce food intake and modulate brain activity in appetite
centers in humans. Cell Metab 2011, 14:700-706.

Current Opinion in Pharmacology 2016, 31:812

www.sciencedirect.com