Key words Obesity - motivation - GLP-1 - Dopaminergic midbrain
Introduction
The growing obesity pandemic is one of the biggest health problems in the 21st
century that gives rise to multiple comorbidities such as cancer or
neurodegenerative diseases and thus dramatically increases mortality [1 ]
[2 ]
[3 ]. In the western world,
excessive food intake beyond physiological needs as well as reduced physical
activity are well established as main causes promoting obesity. However, the
underlying mechanisms instigating this maladaptive behavior, which disregards or
overrides homeostatic needs, are poorly understood. In everyday life, we constantly
make decisions and adapt our behavior to our physiological needs and the surrounding
environment – e. g., we decide to go to the bakery across the street
to get a delicious sandwich avoiding lunch in the inhouse cafeteria. To ensure our
physiological homeostasis and to adapt our behavioral responses, our brain
constantly integrates information about the metabolic state (e. g., hunger)
with external environmental cues (bakery vs. inhouse cafeteria). External cues can
comprise varying motivational signals, which include the incentive value of an
expected reward (delicious sandwich vs. cafeteria lunch) but also the effort
required to obtain the reward (distant bakery across the street vs. close inhouse
cafeteria). Thus, everyday decisions in favor of or against food intake are based on
cost-benefit analyses weighing the potential food reward against the cost of
spending effort to obtain it.
The incentive theory of motivation regards motivational behavior to mainly depend on
anticipated rewards and reinforcement; hence, incentive motivation refers to the
processes that translate expected reward into the effort spent to obtain the reward
[4 ]
[5 ]
[6 ]. Importantly, the subjective valuation of the magnitude of a reward
depends on our internal state; a sandwich is regarded as more valuable in a hungry
than a sated state [7 ]. Consequently, our
motivation depends on the capacity of our brain to integrate internal state signals
(hunger) with environmental cues (distance to bakery, value of sandwich) to guide
our behavior.
There is increasing evidence that high-fat diet consumption and obesity perturbate
the underlying neural processes leading to maladaptive behavior and motivational
deficits. This mini-review aims to give a short overview of incentive motivation,
its neural encoding in the dopaminergic mesolimbic system and its metabolic
regulation with a focus on derangements of incentive motivation in obesity as one
mechanism underlying excessive food intake and reduced physical activity. We further
give an outlook on the emerging field of metabolic interventions to counteract
motivational deficits and their potential clinical implications.
Encoding of incentive motivation in the dopaminergic midbrain
Encoding of incentive motivation in the dopaminergic midbrain
Incentive motivation is encoded by the mesolimbic dopaminergic system. Dopaminergic
neurons projecting from the ventral tegmental area (VTA) to the nucleus accumbens
(NAc) encode both reward-driven learning and motivation to work for reward [8 ]
[9 ]. Learning signals are encoded by phasic dopamine release in the NAc.
On the other hand, dopamine release ramps up when we approach a reward reflecting
reward expectancy [10 ]. The amount of
dopamine, that is released during this ramping-up phase, provides information about
the value of the anticipated reward and motivates the amount of effort required to
obtain it [11 ]
[12 ]
[13 ]. While our understanding of the differential functions of dopamine in
reward learning and motivation are mainly derived from animal studies, human brain
imaging studies support these results, as changes in the activity of the NAc were
observed to correlate with the subjective value of rewards and its anticipation
[12 ]
[14 ]
[15 ]. Likewise, human pharmacological intervention studies lowering
dopaminergic tone have provided evidence for lower effort spending and motivation
[16 ]
[17 ]
[18 ].
Motivation in obesity – inconclusive results
Motivation in obesity – inconclusive results
In obese humans, alterations in the fronto-mesolimbic dopamine system (in particular
within the dopaminergic projections from the VTA to the NAc) are related to an
impaired reward system [19 ]
[20 ]. A high-fat diet –
interestingly even in the absence of obesity – induces downregulation of
dopaminergic D2 receptors as described in animal studies [21 ]
[22 ]. On a functional level, high-fat diet consumption also causes a
devaluation of standard chow diet, which is encoded by reduced dopamine release from
the VTA upon receival of standard food (amongst encoding by hypothalamic
agouti-related peptide neurons), and thus diminishes the rewarding properties of
food discovery [20 ]. In humans, the
direct impact of a high-fat diet on the mesolimbic system in the absence of obesity
has not been studied yet. The data comparing obese and healthy weight participants
is consistent with the abovementioned animal literature. Human positron emission
tomography (PET) studies revealed a negative correlation between body mass index
(BMI) and striatal D2 receptor density or binding potential in obese and overweight
humans [23 ]
[24 ]. This reduced binding potential of
striatal dopamine receptors seems to be associated with an altered striatal
dopaminergic tone leading to an imbalance of anticipation and consumption of food
reward [25 ]
[26 ]. In comparison to lean individuals,
humans with obesity show increased neural activation of the NAc when anticipating a
reward but experience less activation of reward circuits from the actual food reward
consumption [23 ]
[27 ]. These changes in D2 receptor binding
potential seem to be partly reversible by bariatric surgery-induced long-term weight
loss [28 ]
[29 ]
[30 ]. In animals, bariatric surgery even seems to change the motivation
for drug rewards via post-surgical increases in bile acid signaling, which reduces
accumbal dopamine [31 ]. However, only
little is known about incentive motivation in obese humans and findings portray a
heterogeneous picture of effort spending in obesity.
Mathar et al. [32 ] assessed motivational
differences between lean and obese humans in a cost-benefit decision-making
paradigm, in which participants had to exert physical effort on a handgrip to win
food and non-food reward. Obese participants were less willing to engage in physical
effort in particular for high-caloric sweet snack food. In contrast, Epstein et al.
as well as Giesen et al. suggest that obese humans may be willing to invest more
effort to obtain high-caloric food than lean individuals [33 ]
[34 ].
Metabolic modulation of the dopaminergic midbrain – animal
results
Metabolic modulation of the dopaminergic midbrain – animal
results
These studies might rest upon incomplete assumptions about modulatory influences on
midbrain dopaminergic function, as VTA dopaminergic neurons are not only involved in
reward learning and motivation but are also sensitive to nutritional signals [35 ], post-ingestive effects of food [36 ]
[37 ], and metabolic state signaled by peptidergic mediators [38 ]
[39 ]
[40 ]. Many orexigenic
agents (such as Ghrelin) and postprandial anorexigenic peptides (such as
glucagon-like peptide 1 (GLP-1), insulin or leptin) or their agonists can bind to
receptors on dopaminergic neurons of the VTA/Nac [41 ] and hence modulate dopaminergic
performance upon their activation affecting motivational behavior. In line, animal
studies reveal that the hunger hormone ghrelin applied directly into the VTA
increases dopamine levels in the NAc and thus induces motivational behavior for food
rewards [42 ]
[43 ]
[44 ]
[45 ]. Correspondingly, in a
state of overnight fasting with high endogenous levels of ghrelin, ghrelin receptor
blockade in the VTA reduces the motivation to work for food reward [46 ]
[47 ]. However, the effect of ghrelin on motivation seems to depend on the
surrounding environmental cues. In a classical place preference task in rodents,
peripheral injection of ghrelin in the absence of food-induced aversive behavior
[48 ], while in the presence of food,
ghrelin instigated conditioned place preference [44 ]
[48 ]) indicating that the interplay of endogenous metabolic signals and
environmental perception shapes adaptive motivational behavior.
Contrary to ghrelin, postprandial hormones such as insulin or GLP-1 reduce motivated
behavior for food rewards in mice by downregulating dopaminergic transmission [49 ]. Specifically, insulin action on
dopaminergic VTA neurons depresses excitatory synaptic transmission [50 ], decreases dopamine concentrations by
enhancing its clearance [51 ]
[52 ], and reduces dopamine release into the
NAc [53 ]. GLP-1 (and its analogues)
reduces phasic dopamine release of VTA dopaminergic neurons in response to
reward-indicating cues and attenuates synaptic drive onto mesolimbic dopamine
neurons [54 ]
[55 ]
[56 ]. Amylin (and its receptor agonists), also seems to affect
dopaminergic neurons in the VTA, reducing phasic dopamine action in the NAc and
consequently food intake [57 ]
[58 ]. Its complete role as a modulator of
dopaminergic activity and hence motivation still requires further investigation with
first data revealing an attenuating effect of Amylin receptor agonists on the
rewarding properties of alcohol [59 ].
Similarly, the adipocytokine leptin, which circulates in proportion to body fat to
signal the repletion of long-term energy stores, expresses its receptor on VTA
dopaminergic neurons. However, ablation of these leptin receptors does not alter
motivational behavior but increases anxiety-like behavior (as these neurons mainly
project to the amygdala, which is highly implicated in anxiety) [60 ]
[61 ]. Nonetheless, leptin reduces motivational behavior for food. The mode
of action seems to be more indirect, however, with leptin receptor-bearing neurons
of the lateral hypothalamus, decreasing mesolimbic dopaminergic function as a
consequence of increased dopamine uptake in the NAc [62 ].
Collectively, the orexigenic peptide ghrelin seems to enhance motivational behavior
in rodents by upregulating dopaminergic transmission in the mesolimbic system,
whereas postprandial and anorexigenic peptides (such as insulin, GLP-1, leptin, and
probably amylin) have the opposite effect on dopaminergic function and
motivation[63 ]. However, the food
itself, which is used as a reinforcer in motivational paradigms, exerts a
time-dependent effect on dopamine release with an immediate orosensory and delayed
post-ingestive dopaminergic response [37 ]. Considering the multitude and complexity of modulatory influences on the
dopaminergic mesolimbic system, the above-portrayed roles of peptidergic hormones in
the regulation of motivation might be multifaceted with varying effects depending on
nutritional/metabolic state.
Metabolic modulation of motivational behavior in humans and its derangements in
obesity
Metabolic modulation of motivational behavior in humans and its derangements in
obesity
While the modulatory effect of metabolic peptides on the dopaminergic neurocircuitry
and motivational behavior is well documented in rodents, evidence for a modulatory
role of peripheral peptides or metabolic state affecting motivational behavior in
humans is scarce. We, therefore, set up a randomized, placebo-controlled study to
assess the effect of GLP-1 on motivation in lean and obese individuals while
capturing their metabolic state, i. e., their hunger level and insulin
sensitivity. Participants exerted force on a hand-grip to win food and monetary
rewards. We could show that hunger increases incentive motivation in lean humans but
not in obese humans indicating that motivational irregularities in obesity are
state-dependent. We further observed that the effect of hunger on incentive
motivation is modulated by the peripheral insulin sensitivity of the individual with
impaired peripheral insulin sensitivity reducing the motivational effect of hunger
[64 ]. These results are in line with
previous studies showing that altered insulin sensitivity impacts dopaminergic
projections of the midbrain and denote a dysfunctional integration of metabolic
signals and external cues within the mesolimbic system as the foundation of impaired
motivational drive in obesity [65 ]. The
aforementioned heterogeneous results about incentive motivation in obese humans
showing both increased and decreased effort spending for rewards might thus be a
consequence of neglecting metabolic state, in particular, insulin sensitivity and
fasting time/hunger. We further demonstrate that administration of the GLP-1
analogue liraglutide normalizes the motivational effect of hunger in
insulin-resistant humans. Most importantly, this holds true for both food and
monetary reward, indicating that the modulatory effect of GLP-1 on motivational
behavior exceeds a mere food scenario and might prove beneficial in other disorders
with motivational deficits [64 ].
Outlook: Metabolic treatment of motivational deficits in psychiatric
diseases
Outlook: Metabolic treatment of motivational deficits in psychiatric
diseases
As GLP-1 normalized motivation in insulin-resistant humans but did not affect
motivation in insulin sensitive participants, GLP-1(analogues) might comprise
therapeutic potential for motivational dysfunctions in dopaminergic disorders, which
are associated with metabolic impairments such as insulin resistance. Insulin
resistance is a shared abnormality among many patients with type 2 diabetes mellitus
and major depression [66 ] hence, GLP-1
could be cautiously hypothesized to improve dopaminergic functioning in depression
and hence depressive symptom burden in patients suffering from both depression and
insulin resistance. Furthermore, GLP-1 receptor polymorphism has been associated
with anhedonia – the lack of motivation, which is a core symptom of
depression. In a first meta-analysis, treatment of diabetic patients with GLP-1
analogues resulted in a significant reduction of depression scores [67 ]. However, this meta-analysis is based
on eight publications only with heterogeneous cohorts so that the result must be
interpreted with caution. While the GLP-1 analogue Liraglutide is already approved
as weight loss medication in Germany [68 ], randomized controlled clinical trials on the efficacy and safety of
GLP-1 analogues as a treatment for motivational deficits in diabetic or
insulin-resistant patients suffering from depression are lacking.
In animal studies, GLP-1 also reduces the reinforcing efficacy of drugs of abuse so
that a potential therapeutic benefit of GLP-1 analogues could also be suspected for
detoxification therapies [69 ]. For
instance, GLP-1 analogues were shown to reduce cocaine, amphetamine, alcohol, and
nicotine use in animals [70 ].
Interventional studies in humans are fervently awaited.
Summary
In summary, both external cues and internal state signals are integrated into the
dopaminergic mesolimbic system to guide our everyday motivational behavior (see
[Fig. 1 ]). Dopamine release in the
NAc ramps up as a reward approaches, encoding reward expectancy. Metabolic peptide
hormones – such as insulin, GLP-1, leptin, or ghrelin – modulate
dopaminergic transmission thus regulating motivational behavior, that is, hunger
augments motivation to exert effort for rewards. In obesity, insulin resistance
reduces the motivational effect of hunger, indicating that a dysfunctional
integration of metabolic signals with external cues seems to lead to derangements of
incentive motivation representing one possible mechanism underlying excessive food
intake and reduced physical activity in obesity. Concomitantly, interventions with
metabolic messengers offer new opportunities to reverse motivational deficits. In
obese humans, for example, intervention with GLP-1 normalizes motivational behavior.
Future clinical research directions should therefore include the safety and efficacy
of clinical interventions with metabolic mediators in pathologies associated with
motivational insufficiencies.
Fig. 1 Encoding of motivation in the dopaminergic midbrain. Notes. 1)
Dopamine (DA) encodes motivation in the midbrain; dopaminergic neurons
project from the ventral tegmental area (VTA) to the nucleus accumbens (NAc)
with higher dopaminergic tone encoding higher motivation. 2) Dopamine
signaling is impaired in obesity and under high-fat diet. The binding
potential of dopaminergic D2 receptors is reduced in obesity and high-fat
diet (in animal studies) causing a reduced dopamine release from the VTA
upon receival of standard food and thus diminishing the rewarding properties
of food discovery. 3) Dopamine signaling is modulated by metabolic signals
as shown in animal studies. Ghrelin applied into the VTA increases dopamine
levels in the NAc and hence motivation; glucagon-like peptide 1 (GLP-1),
insulin, leptin and (probably) amylin reduce motivated behavior for food
rewards in mice by downregulating dopaminergic transmission. 4) In humans,
motivation increases with increasing hunger levels in normal weight humans
with good insulin sensitivity, while in obese humans with reduced insulin
sensitivity hunger does not affect motivation. Intervention with GLP-1 does
not affect motivation in insulin sensitive humans but normalizes the effect
of hunger on motivation in insulin resistant humans.
One sentence summary In this Mini-Review, we provide a comprehensive overview
on incentive motivation, its neural encoding in the dopaminergic mesolimbic system
as well as its metabolic modulation with a focus on derangements of incentive
motivation in obesity.
