설탕을 끊으면 우울증, 비만이 해결된다!!

[문형철]

The impact of sugar consumption on stress driven, emotional and addictive behaviors

Author links open overlay panelAngela Jacques a b, Nicholas Chaaya a b, Kate Beecher a b, Syed Aoun Ali a b, Arnauld Belmer a b 1, Selena Bartlett a b 1

Show more

Add to Mendeley

Share

Cite

https://doi.org/10.1016/j.neubiorev.2019.05.021Get rights and content

Under a Creative Commons license

open access

Highlights

A diet high in sugars has been linked to cognitive impairments, negative neuroplasticity and emotional disorders such as anxiety and depression. Sugar consumption increases the impulsivity to feed. Sugar overconsumption leads to changes in neurobiological brain function which alter emotional states and subsequent behaviours. Addiction, stress, fear, anxiety and depression involve overlapping neural mechanisms.

당분이 많은 식단(설탕)은 인지 장애, 부정적인 신경 가소성, 불안 및 우울증과 같은 정서 장애와 관련이 있습니다. 설탕 섭취는 먹으려는 충동을 증가시킵니다. 설탕 과잉 섭취는 신경 생물학적 뇌 기능의 변화로 이어져 감정 상태와 그에 따른 행동을 변화시킵니다. 중독, 스트레스, 두려움, 불안, 우울증은 중첩된 신경 메커니즘을 수반합니다.

Abstract

In 2016 the World Health Organization reported 39% of the world’s adult population (over 18 y) was overweight, with western countries such as Australia and the United States of America at 64.5% and 67.9% respectively. Overconsumption of high fat/sugar containing food and beverages contribute to the development of obesity. Neural plasticity that occurs as a result of long term sugar consumption has been shown to reduce impulse control and therefore lower the ability to resist the high fat/sugar foods contributing to the obesity epidemic. There is significant overlap between the neural pathways involved in emotions that guide behavioural responses to survival situations with those regulating overconsumption of highly palatable food. This suggests that having a clearer understanding of the role of stress and emotions in the development of obesity will lead to the development of novel therapeutic strategies. Sucrose consumption activates the mesocorticolimbic system in a manner synonymous with substances of abuse. There is overwhelming evidence to support the hypothesis that sucrose consumption results in pathophysiological consequences such as morphological neuronal changes, altered emotional processing and modified behaviour in rodent and human models. In this comprehensive review, we examined >300 studies investigating the interaction between sugar consumption, stress and emotions. Preclinical and clinical trials investigating highly palatable foods and stress, anxiety, depression and fear are reviewed. Importantly, the synergy between sugar consumption and neurobiology is addressed. This review summarizes the neurochemical changes and neural adaptations ö including changes in the dopaminergic system ö that influence emotion and behaviour following sugar consumption.

2016년 세계보건기구는 전 세계 성인 인구(18세 이상)의 39%가 과체중이며, 호주와 미국과 같은 서구 국가는 각각 64.5%와 67.9%라고 보고했습니다. 고지방/당분이 함유된 음식과 음료의 과잉 섭취는 비만의 원인이 됩니다. 장기간의 설탕 섭취로 인해 발생하는 신경 가소성은 충동 조절 능력을 감소시켜 비만 유행에 기여하는 고지방/설탕 음식에 대한 저항력을 떨어뜨리는 것으로 나타났습니다. 생존 상황에 대한 행동 반응을 유도하는 감정에 관여하는 신경 경로와 기호성이 높은 음식의 과잉 섭취를 조절하는 신경 경로 사이에는 상당한 중복성이 있습니다. 이는 비만 발생에서 스트레스와 감정의 역할을 더 명확하게 이해하면 새로운 치료 전략 개발로 이어질 수 있음을 시사합니다. 자당 섭취는 남용 물질과 유사한 방식으로 중변연계를 활성화합니다. 설치류와 인간 모델에서 설탕 섭취가 형태학적 신경세포 변화, 감정 처리 변화, 행동 변화와 같은 병리 생리학적 결과를 초래한다는 가설을 뒷받침하는 압도적인 증거가 있습니다. 이 종합 리뷰에서는 설탕 섭취, 스트레스, 감정 사이의 상호작용을 조사한 300여 건의 연구를 검토했습니다. 기호성이 높은 음식과 스트레스, 불안, 우울, 공포를 조사하는 전임상 및 임상시험을 검토했습니다. 중요한 것은 설탕 섭취와 신경생물학 사이의 시너지 효과에 대해 다루고 있다는 점입니다. 이 리뷰에서는 설탕 섭취 후 감정과 행동에 영향을 미치는 도파민 시스템의 변화를 포함한 신경 화학적 변화와 신경 적응에 대해 요약합니다.

• Previous article in issue
• Next article in issue

1. Introduction

A sedentary lifestyle combined with a high caloric diet plays a significant role in obesity (Organization, W.H., 2017). In 2016 the World Health Organization reported more than 1.9 billion adults were overweight (Organization, W.H., 2016). World obesity has essentially tripled since 1975 (Organization, W.H., 2016). Excessive sugar consumption has been shown to be one of the leading contributors to weight gain (Malik et al., 2013). Furthermore, a diet high in sugars has been linked to cognitive impairments, negative neuroplasticity changes such as hippocampal dysfunction (Kanoski and Davidson, 2011; Noble et al., 2017) and emotional disorders such as anxiety and depression (Kim et al., 2018). High sugar intake increases the risk of cancer, oxidative stress, inflammation, and obesity (Makarem et al., 2018), as well as impacting cognitive function (Barnes and Joyner, 2012) and mental health (Peet, 2004). For example, a diet higher in refined sugar has been shown to predict a worsening of schizophrenic behaviour over a two year period (Peet, 2004). Despite the many psychological, physical and neurological burdens of sugar overconsumption and consequent obesity, there are no therapies directed at reducing sugar consumption, and few therapies capable of successfully treating obesity (for review see (Kanoski and Davidson, 2011)).

Obesity, arising from overconsumption of rewarding foods such as those with a high sugar content, may result in negative consequences, such as a loss of self-control and subsequent poor decision-making (Beilharz et al., 2014). This loss of control poses a significant challenge for overweight individuals attempting to lose weight. The desire to eat, is regulated by brain regions known as feeding centers, located in the arcuate nucleus of the hypothalamus. Importantly, these regions are interconnected with the limbic system and cerebral cortex (specifically the hippocampus and amygdala), which are responsible for the modulation of emotions (Sweeney and Yang, 2017; Liu et al., 2014; Stunkard et al., 2003; Simon et al., 2006; Ulrich-Lai et al., 2010; Ahima and Antwi, 2008).

고칼로리 식단과 함께 앉아서 생활하는 생활 방식은 비만에 중요한 역할을 합니다(세계보건기구, 2017). 2016년에 세계보건기구는 19억 명 이상의 성인이 과체중이라고 보고했습니다(WHO, 2016). 세계 비만은 1975년 이후 본질적으로 세 배로 증가했습니다(WHO, 2016). 과도한 설탕 섭취는 체중 증가의 주요 원인 중 하나로 밝혀졌습니다(Malik et al., 2013). 또한 설탕이 많이 함유된 식단은 인지 장애, 해마 기능 장애와 같은 부정적인 신경 가소성 변화(Kanoski and Davidson, 2011; Noble et al., 2017), 불안 및 우울증과 같은 정서 장애(Kim et al., 2018)와 관련이 있는 것으로 밝혀졌습니다. 설탕을 많이 섭취하면 암, 산화 스트레스, 염증, 비만의 위험이 높아질 뿐만 아니라(Makarem et al., 2018), 인지 기능(Barnes and Joyner, 2012)과 정신 건강(Peet, 2004)에도 영향을 미칩니다. 예를 들어, 정제 설탕 함량이 높은 식단은 2년 동안 조현병 행동의 악화시키는 것으로 나타났습니다(Peet, 2004). 설탕 과잉 섭취와 그로 인한 비만으로 인한 심리적, 신체적, 신경학적 부담이 많음에도 불구하고 설탕 섭취를 줄이기 위한 치료법은 없으며 비만을 성공적으로 치료할 수 있는 치료법도 거의 없습니다(검토는 (Kanoski and Davidson, 2011) 참조).

당 함량이 높은 식품과 같은 보상성 식품의 과다 섭취로 인해 발생하는 비만은 자제력 상실과 그에 따른 잘못된 의사 결정과 같은 부정적인 결과를 초래할 수 있습니다(Beilharz et al., 2014). 이러한 통제력 상실은 체중 감량을 시도하는 과체중 개인에게 중대한 도전이 됩니다. 식욕은 시상하부의 아치형 핵에 위치한 섭식 중추로 알려진 뇌 영역에 의해 조절됩니다. 중요한 것은 이 영역이 변연계 및 대뇌 피질(특히 해마 및 편도체)과 상호 연결되어 있으며, 이는 감정 조절을 담당한다는 것입니다(Sweeney and Yang, 2017; Liu 등, 2014; Stunkard 등, 2003; Simon 등, 2006; Ulrich-Lai 등, 2010; Ahima와 Antwi, 2008).

1.1. Sugar, artificial sweeteners and obesity

Sugar became embedded in the food chain in the late 1960s and replaced fats to mask bitterness and make food more palatable (Bakke et al., 2018; Moss, 2013). In the 1970’s a shift toward increased sugar-sweetened beverages became apparent (Wolf et al., 2008). Our early ancestors obtained sugar from either fruit, limited by seasons, or honey, protected by bees. In the last half century, sugar consumption has tripled worldwide, partially due to the hidden use of added sugars in processed food (Lustig et al., 2012). The first artificial sweetener (saccharin) was introduced in 1879 with low production costs during wartime, increasing its popularity (Weihrauch and Diehl, 2004). In the 1950s as sugar became readily available, the use of sweeteners shifted to so-called ‘diet products’ with low caloric content.

The development of obesity relies on both the hedonic, sweet taste of food in conjunction with the negative emotional properties of food consumption (Meye and Adan, 2014; Yu and Kim, 2012; Berthoud et al., 2017). Hedonic reactions to a 10% sucrose solution (and as low as 3.4%) were tested and found to be significantly higher in adolescent rats when compared to adult rats (Wilmouth and Spear, 2009).

1.2. Sugar and emotions

Emotional eating has been show to stem from the desire to mitigate the effects of stress (Van Strien et al., 1986) and stress is partially regulated by the hypothalamic-pituitary-adrenal (HPA) axis. Interestingly, activity of the HPA axis has been shown to be reduced through the consumption of sugar containing foods (Ulrich-Lai et al., 2011). Following consumption, hormones are released to reduce the feelings of stress, which also increase the desire for comfort foods, thus perpetuating emotional eating habits (Ulrich-Lai et al., 2011; Ursano et al., 2009).

The objective of this review is to summarize research that examines how the consumption of sugar leads to changes in neurobiological brain function that alters emotional states and subsequent behaviours. We will examine findings from studies at the intersection between the consumption of sucrose and changes in the performance of tasks with a stressful or emotional component and review neurobiological and neurochemical mechanisms involved in addiction, stress, fear, anxiety and depression to determine whether there are overlapping neural mechanisms. Lastly, we will determine whether there are novel pharmacotherapeutics and/or interventions that target these brain circuits or neurochemical pathways to improve the current approaches for the treatment of obesity.

2. Common neuronal pathways for sucrose consumption, addiction, emotions and obesity

Addiction is characterized by a difficulty to control habitual behaviour even in the face of negative consequences (Lindgren et al., 2018). Early addiction research focused on drugs of abuse such as alcohol, morphine and nicotine. This has since been extended to include gambling, eating and more recently, sugar consumption (Comings et al., 2001; Weeks, 1962; Jellinek, 1952; Benowitz, 1988; Gearhardt et al., 2009; Klenowski et al., 2016; Benton, 2010). Addiction to substances of abuse relies on the drug binding to specific protein targets which elicits certain physiological and behavioural responses unique to that drug (Lee and Messing, 2008). Psychoactive drugs commonly result in rewarding sensations that lead to repeated use and, depending on genetic susceptibility, environmental factors and subsequent addiction (Lee and Messing, 2008). Stress has long been associated with both the motivation to use rewarding substances and the result of not attaining those substances (for review see (Sinha, 2008)). The negative symptoms produced through withdrawal are common to all forms of addictive substances including highly palatable food. These include prolonged sensitization to the substance of choice and associative learning where environmental cues become associated with the pleasure derived from the substance (Hebebrand et al., 2014). These associative memories, combined with intense cravings, increase the incidence of relapse even after sustained abstinence. These commonalities may be due to substances of addiction utilizing the same circuitry within the brain’s mesocorticolimbic system (Fig. 1) (Volkow et al., 2004; Koob and Le Moal, 2001; Baron et al., 2018).

1. Download : Download high-res image (712KB)
2. Download : Download full-size image

Fig. 1. Reward pathway encompassing the mesocorticolimbic distribution of dopaminergic neurons. The neural regions of the reward pathway include the prefrontal cortex (PFC), amygdala (AMG), ventral tegmental area (VTA) and nucleus accumbens (NAc). Each anatomical region modulates individual behaviours and contributes to general behaviours through cross-connectivity. Regions shown and behaviours listed are consistent between human and rodent brains (Belin and Dalley, 2012).

Food consumption is necessary to regulate homeostatic energy balance. However, humans also consume food for pleasure or comfort. The hedonistic desire for palatable food is considered reward-related and may result in maladaptive or negative neuroplasticity that can override homeostatic regulation and result in overeating behaviours (Kenny, 2011). Reward is delivered through stimuli that produce pleasurable or enjoyable experiences, contrasting with addiction, which involves compulsive and sometimes painful behaviour, yet both share similar neuroadaptive responses and overlapping neuronal pathways (Adinoff, 2004). The reward or mesocorticolimbic pathway plays an influential role in what we choose to eat as demonstrated by studies where rats appear willing to endure noxious stimuli such as extreme cold, heat and foot-shock to procure highly palatable foods over standard rat chow (Cabanac and Johnson, 1983; Foo and Mason, 2005; Oswald et al., 2011).

The reward pathway consists of the prefrontal cortex (PFC), amygdala (AMG), ventral-tegmental area (VTA) and nucleus accumbens (NAc) and in accord with drugs of abuse (Sinha, 2018; Alsiö et al., 2012), is thought to be stimulated by the overconsumption of sugar, thus contributing to the development of obesity (Leigh and Morris, 2016; Murray et al., 2016; Shariff et al., 2016). The reward pathway is highly associated with the efflux of dopamine which regulates the motivational state of wanting or craving that substance or behaviour (Adinoff, 2004). Once overstimulated, the pathway becomes primed to require that particular stimuli when presented with contextual cues or emotional stress (Adinoff, 2004). The VTA induces this sensitization while the NAc modulates its expression‎ through dopaminergic control (Huang et al., 2018; Xiong et al., 2018; Zhang et al., 2018a).

2.1. Nucleus accumbens

The NAc consists of two sub regions (core and shell), each containing specific neuronal populations fundamental in processes of motivation (Yang et al., 2018; Reker et al., 2018), aversion (Rosas et al., 2018), fear related avoidance (Moscarello and Maren, 2018), reinforcement learning (Floresco et al., 2018; Anderson et al., 2018), pleasure seeking, addiction (Kim et al., 2018; Lam and Jadavji, 2018; Barrientos et al., 2018) and behavioural sensitization (Nona and Nobrega, 2018). To examine motivation and the role played by α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors in the NAc, intra-NAc infusions of AMPAR antagonists were given to obesity prone and obesity resistant male rats (Derman and Ferrario, 2017). Compared with controls, findings showed the obesity-prone animals exhibited cue-triggered food seeking and the behaviour was mediated by increased surface expression‎ of AMPA receptors in the core of the NAc. This data suggests cue-triggered food seeking may elicit greater motivational feedback in individuals prone to obesity, and thus play a role in driving the compulsive over-consumption of food (Derman and Ferrario, 2017; Berridge et al., 2010). Similarly, consumption of a 10% sucrose solution resulted in alterations in synaptic strength via AMPA receptor trafficking, which modulates the compulsive tendencies characteristic of any drug-seeking (Peng et al., 2011).

Drug induced neuroplasticity has been observed in the NAc through alterations in dendritic morphology and altered gene expression‎ (Huang et al., 2018; Legastelois et al., 2014; Li et al., 2004). Similar to psychoactive substances, sugar binging has been shown to cause repeated increases in dopamine release and altered expression‎ of NAc Delta FosB as demonstrated in rats with an increase in Delta Fos B expression‎ after consumption of a high sugar diet (Avena et al., 2006; Wallace et al., 2008). Dopamine release increases as a direct effect of chronic drug use and results in postsynaptic changes of Delta FosB and CREB accumulation (Nestler and Aghajanian, 1997). The intracellular build-up of Delta FosB can alter the gene production of receptors which may result in reinstatement during withdrawal (Nestler and Aghajanian, 1997).This data suggest that neuroadaptations in the brain reward pathway in obese subjects may contribute to the progression of compulsive eating. The NAc is considered to be the main region to undergo neuroadaptation after sugar consumption, but changes have been noted in many brain regions encompassing the mesocorticolimbic system (De Jong et al., 2016).

2.2. Amygdala

The AMG plays a key role in negative reinforcement, the progression towards addiction and extensively in the learned associations that lead to relapse (See et al., 2003). Neuroplastic changes occur in the AMG which facilitate the level of dependence to a substance of abuse to move from impulsivity to compulsion (Koob, 2009a). This was demonstrated by long term sucrose consumption in rats which produced maladaptive alterations in the apical dendritic morphology of AMG principle neurons (Shariff et al., 2017). Optogenetic stimulation of central AMG neurons was paired with a sucrose reward to test reward incentive in rats (Robinson et al., 2014). Stimulation increased incentive to choose the sucrose reward over a similar sweet alternative and amplified the degree of effort the rats were willing to exert to obtain the reward (Robinson et al., 2014). These findings suggest neuroplastic changes in the AMG may occur due to sucrose consumption and increased addictive-like compulsive behaviour. The AMG and PFC share a role in motivation, associative learning (including negative reinforcement) (Koob, 2009a), compulsive behaviour and deficits in executive functioning (Zehra et al., 2018). Relapsing into addictive behaviours is most likely due to dysregulation of motivational processes. Negative reinforcement, or substance abuse to alleviate an hedonic, anxious, irritable or dysphoric state, is characteristic of addiction withdrawal (Koob, 2018).

2.3. Prefrontal cortex

The orbitofrontal cortex and anterior cingulate are associated with compulsive cravings for drugs of abuse (Adinoff, 2004). Interestingly, neurons in the orbitofrontal cortex of rats were shown to encode this same compulsion to seek a sucrose reward solution (15%) (Moorman and Aston-Jones, 2014). The PFC is the brain region responsible for executive functions such as planning and decision making. Addiction and periods of withdrawal from substances of abuse are generally accompanied by a loss of executive function, which is due to dysfunctional neurocircuitry in the medial PFC (Molnar et al., 2018). Sucrose and fructose consumption have also been linked to metabolic and electrophysiological changes in the hippocampus (Hsu et al., 2015; Lemos et al., 2016), the thalamus (Jastreboff et al., 2016) and the hypothalamus (Berthoud et al., 2017; Ulrich-Lai et al., 2011); brain regions involved in the associative learning of cues that lead to receiving rewards (Atkins et al., 1998; Anderson and Petrovich, 2018).

3. Physiological and neural substrates of sugar consumption

Feeding centers linked to limbic regions relaying emotional information influence our behaviour towards food (Liu et al., 2014). The current view of what drives us to eat takes into account the drive to maintain an internal balance between energy expenditure and consumption. However, the influence of external cues that promise immediate reward may negate this balance (Stroebe et al., 2008; Lowe and Butryn, 2007). Sucrose added to food provides this immediate reward, with the human desire for sucrose or any sweet taste being comparable to the degree of yearning and reward produced by drugs of addiction (Ahmed et al., 2013). Sugar or sucrose (table sugar) is a disaccharide composed of the monosaccharides glucose and fructose (Smith, 1960; Southgate, 1995). Unlike most drugs of addiction (excluding alcohol), sugar does not cross the blood brain barrier to bind to molecular substrates/receptors on the cell surface and subsequently alter neural plasticity. Recent research has observed astrocytes, a type of glial cell that helps maintain the blood-brain barrier appear to sense and uptake sugar to regulate neuronal signaling related to appetite (García-Cáceres et al., 2016). However, as sugar is a food, ingestion begins in the mouth.

Sugar is initially sensed by heterodimeric G protein-coupled receptors on taste cells located in the mouth and gut (Brown and Rother, 2012). Once in the small intestine, sucrose is broken down into glucose and fructose; which are metabolized by separate and disparate mechanisms (Southgate, 1995). Glucose increases the absorption of fructose from the gut, whereas fructose acts as a catalyst for the uptake and storage of glucose by the liver (Laughlin, 2014). Fructose is absorbed into the bloodstream at a slower rate and persists for longer (Teff et al., 2004). As opposed to glucose, fructose is not well absorbed by pancreatic beta cells and therefore stimulates very little insulin secretion (Kyriazis et al., 2012). Insulin increases satiety and subdues the reward value derived from food, suggesting fructose may play a more complex role in the development of obesity than glucose consumption (Grant et al., 1980). In addition to these observed features fructose can increase the rate of carbohydrate oxidation after a meal, also decreasing satiety levels (Laughlin, 2014).

감정 정보를 전달하는 변연계와 연결된 섭식 중추는 음식에 대한 우리의 행동에 영향을 미칩니다(Liu et al., 2014). 우리가 음식을 먹게 하는 원동력에 대한 현재의 견해는 에너지 소비와 소비 사이의 내부 균형을 유지하려는 욕구를 고려합니다. 그러나 즉각적인 보상을 약속하는 외부 신호의 영향은 이러한 균형을 무효화할 수 있습니다(Stroebe et al., 2008; Lowe and Butryn, 2007). 음식에 첨가된 자당은 이러한 즉각적인 보상을 제공하며, 자당 또는 단맛에 대한 인간의 욕구는 중독성 약물이 만들어내는 갈망과 보상의 정도와 비슷합니다(Ahmed et al., 2013). 설탕 또는 자당(식용 설탕)은 단당류인 포도당과 과당으로 구성된 이당류입니다(Smith, 1960; Southgate, 1995). 대부분의 중독 약물(알코올 제외)과 달리 설탕은 혈액 뇌 장벽을 통과하여 세포 표면의 분자 기질/수용체에 결합하여 신경 가소성을 변화시키지 않습니다. 최근 연구에 따르면 혈뇌장벽을 유지하는 데 도움이 되는 신경교세포의 일종인 성상교세포가 설탕을 감지하고 흡수하여 식욕과 관련된 신경 신호를 조절하는 것으로 나타났습니다(García-Cáceres 외., 2016). 그러나 설탕은 음식이기 때문에 섭취는 입에서 시작됩니다.

설탕은 입과 장에 있는 미각 세포의 헤테로이합체 G 단백질 결합 수용체에 의해 처음에 감지됩니다(Brown and Rother, 2012). 소장에 도달한 자당은 포도당과 과당으로 분해되며, 이 두 가지 당은 서로 다른 메커니즘에 의해 대사됩니다(Southgate, 1995). 포도당은 장에서 과당의 흡수를 증가시키는 반면 과당은 간에서 포도당의 흡수와 저장을 촉진하는 촉매 역할을 합니다(Laughlin, 2014). 

Glucose and fructose have the ability to cross the blood-brain-barrier. However, as the gut and liver rapidly break down fructose, the blood concentration levels are generally low, resulting in only small amounts available to cross the barrier (Gonzalez and Betts, 2018). Once across the barrier glucose signaling mechanisms are activated. As demonstrated in animal studies (Asnaghi et al., 2003; Kikkawa et al., 1987), emerging arguments suggest fructose is produced from glucose in the brain via the polyol pathway (glucose → aldose reductase → sorbitol → sorbitol dehydrogenase → fructose), driven by hyperglycemia (Hwang et al., 2017). Glucose consumption increases functional connectivity between the hypothalamus, thalamus and striatum where fructose does not affect striatal connectivity (Cha et al., 2008). Regional cerebral blood flow patterns also differ post ingestion of these monosaccharides, with reduced flow in appetite and reward regions of the brain after glucose consumption inclusive of the hypothalamus, thalamus, insula, anterior cingulate, and striatum. Alternate to this, fructose reduces flow in the thalamus, hippocampus, posterior cingulate cortex, fusiform, and visual cortex (Page et al., 2013). These differences may assist in clarifying the role of sucrose consumption in obesity and diabetes mellitus. Both human and rat studies have shown that fructose and not glucose consumption results in conditions such as metabolic syndrome and contributes to obesity (Nakagawa et al., 2006).

On the neuroanatomical level, regulation of energy homeostasis is dependent on the hypothalamus (Benoit et al., 2008), however, the NAc, part of the reward-pathway, is pivotal in the pathophysiology of sugar consumption (Geiger et al., 2008). Table 1 depicts the findings of several studies examining the effect of sugar consumption on brain regions involved in the reward pathway, defining the molecular mechanisms resultant of the neural adaptations and the subsequent behavioural changes demonstrated by the animals.

과당은 혈류로 흡수되는 속도가 느리고 더 오래 지속됩니다(Teff et al., 2004). 포도당과 달리 과당은 췌장 베타 세포에 잘 흡수되지 않으므로 인슐린 분비를 거의 자극하지 않습니다(Kyriazis et al., 2012). 인슐린은 포만감을 증가시키고 음식에서 파생되는 보상 가치를 억제하여 과당이 포도당 섭취보다 비만 발생에 더 복잡한 역할을 할 수 있음을 시사합니다(Grant et al., 1980). 이러한 관찰된 특징 외에도 과당은 식사 후 탄수화물 산화 속도를 증가시켜 포만감 수준을 감소시킬 수 있습니다(Laughlin, 2014).

포도당과 과당은 혈액-뇌 장벽을 통과하는 능력이 있습니다. 그러나 장과 간에서 과당을 빠르게 분해하기 때문에 일반적으로 혈중 농도가 낮아 장벽을 통과할 수 있는 과당의 양이 적습니다(Gonzalez and Betts, 2018). 일단 장벽을 통과하면 포도당 신호 전달 메커니즘이 활성화됩니다. 동물 연구에서 입증된 바와 같이(Asnaghi 외, 2003; Kikkawa 외, 1987), 과당은 고혈당에 의해 폴리올 경로(포도당 → 알도스 환원효소 → 소르비톨 → 소르비톨 탈수소효소 → 과당)를 통해 뇌에서 포도당으로부터 생성된다는 새로운 주장이 제기되었습니다(Hwang et al., 2017). 포도당 섭취는 시상하부, 시상 및 선조체 사이의 기능적 연결성을 증가시키지만 과당은 선조체 연결성에 영향을 미치지 않습니다(Cha et al., 2008). 이러한 단당류 섭취 후 뇌의 국소적 혈류 패턴도 달라져 포도당 섭취 후 시상하부, 시상, 섬, 전대상피질, 선조체를 포함한 뇌의 식욕 및 보상 영역의 혈류가 감소합니다. 이와는 반대로 과당은 시상, 해마, 후대상피질, 방추상피질, 시각피질의 흐름을 감소시킵니다(Page et al., 2013). 이러한 차이는 비만과 당뇨병에서 자당 섭취의 역할을 명확히 하는 데 도움이 될 수 있습니다. 인간과 쥐를 대상으로 한 연구에 따르면 포도당이 아닌 과당 섭취는 대사 증후군과 같은 질환을 유발하고 비만에 기여하는 것으로 나타났습니다(Nakagawa et al., 2006).

신경해부학적 수준에서 에너지 항상성 조절은 시상하부에 의존하지만(Benoit et al., 2008), 보상 경로의 일부인 NAc는 설탕 소비의 병태 생리학에서 중추적인 역할을 합니다(Geiger et al., 2008). 표 1은 보상 경로에 관여하는 뇌 영역에 대한 설탕 섭취의 영향을 조사한 여러 연구 결과를 보여 주며, 신경 적응과 그에 따른 동물의 행동 변화로 인한 분자적 메커니즘을 정의합니다.

Table 1. The effects of sugar consumption on the reward pathway.

Sugar consumption causes neural adaptation to brain regions in the reward pathwaySugar consumption alters activation of molecular substratesSugar consumption alters behaviour↑ Dopamine (VTA-NAc)↓ Dopamine↑ Kir2.1↓CREB (NAc) (Leão et al., 2012)↑ Serotonin↓ α6β2 nAChRs↑ α4β2 nAChRs↑ AMPAR (NAc core)IL-6↑ Leptin↑ IL-6↑ TNF-α↓ Protein oxidation (PFC)↑ Delta FosB (NAc)↑ Delta FosB (PFC)

Ventral tegmental area Nucleus accumbens Amygdala Prefrontal cortex	↑ Sensitisation (Wyvell and Berridge, 2001; Sharpe et al., 2016) ↑ Reward seeking (Spangler et al., 2004; Gosnell, 1987) (Avena and Hoebel, 2003)
	↑ Anxiety (Kim et al., 2018) ↑ Depression (Avena et al., 2008)
	↓ Motivation (Mathes et al., 2013)
	↑ Anticipation of reward ↑ Motivation (Shariff et al., 2016)
	↑ Motivational feedback ↑ Cue-triggered food seeking (Derman and Ferrario, 2017)
	↑ Anxiety ↑Fear ↑ Depression (Santos et al., 2018)
	↓Executive function ↑ Anxiety (Chepulis et al., 2009)
	↑ Reward seeking (Wallace et al., 2008) ↓ Fear extinction memory retention (Baker and Reichelt, 2016)

To determine separation of neural networks engaged in choosing to eat based on palatability or nutritional status, striatal dopamine levels were measured in sugar consuming mice (Tellez et al., 2016). Only when the sweet solution contained energy was the dorsal striatum, basal ganglia descending pathway recruited. Energy content drives the release of dopamine in the NAc, however, sweetness modulated this efflux, inhibiting dopamine release as palatability decreased (Tellez et al., 2016). Alternatively, suppression of the sweeteners nutritional value inhibited dopamine release in the dorsal striatum, demonstrating recruitment of different circuitry when separating energy requirements from pleasurable taste (Tellez et al., 2016).

3.1. The effect of sugar on the neurobiology of food consumption

The hedonic value of food, the reward associated with its consumption and the learned external cues which trigger the desire to feed are modulated via bidirectional circuitry between the reward pathway, hippocampus and orexigenic cells (which induce appetite and stimulate food intake) in the lateral hypothalamus (Petrovich and Gallagher, 2003; Petrovich et al., 2012; Kastin, 2013) (see Fig. 2) and for a comprehensive review see (Murray et al., 2016). Studies activating the lateral hypothalamus show glutamatergic inhibition of feeding (Jennings et al., 2013) and GABAergic stimulation of feeding (Jennings et al., 2015). Furthermore, a subset of these GABAergic neurons project from the VTA to the hypothalamus and express galanin (a neuropeptide) which enhances motivation for sucrose consumption (Qualls-Creekmore et al., 2017; Bocarsly, 2018).

1. Download : Download high-res image (799KB)
2. Download : Download full-size image

Fig. 2. Regulation of feeding behaviour and food intake by central and peripheral appetite-regulating hormones and peptides. Hypothalamic orexigenic and anorexigenic pathways and their regulation by NPY/AgRP and POMC/CART peptides, respectively, are depicted in the right-hand side panel. Peripheral appetite-stimulating (green) and –inhibiting (red) hormones and peptides crossing the blood-brain barrier (BBB), and the organs they originate from are also represented. The effect of these hormones/peptides on the stimulation or inhibition of the orexigenic (dashed green arrows) and anorexigenic (dashed red arrows) is depicted as (+) or (-), respectively.

Lateral hypothalamic orexin neurons gain input regarding food intake from the arcuate nucleus of the hypothalamus via endogenous melanocortin receptor antagonists; neuropeptide Y (NPY) and agouti-related peptide-expressing neurons (AgRP). Further, metabolic homeostasis is sensed by the lateral hypothalamus through surrounding glucose, ghrelin and leptin levels, and this drives food seeking behaviours (Yamanaka et al., 2000).

Hypothalamic orexinergic and anorexinergic pathways are regulated by NPY/AgRP and POMC/CART peptides and are affected in different ways by the consumption of sugar (Murray et al., 2016). The orexigenic pathway is influenced by two specific neuronal populations within the arcuate nucleus, the first population expresses NPY and AgRP and stimulates food intake (Benoit et al., 2008; Yamanaka et al., 2000). Variations in both NPY and AgRP were associated with rats consuming more chow after being provided with a 30% sucrose solution (Gaysinskaya et al., 2011). Reduced expression‎ of these peptides was noticed following sucrose consumption, which then increased thirty to sixty minutes later, prior to feeding (Gaysinskaya et al., 2011).

The link between p53 (a gene encoding a protein involved in regulating the cell cycle) in AgRP neurons (which regulate ghrelin-induced appetite) and obesity, has been demonstrated in mice, with its overexpression‎ resulting in excessive weight loss (Quiñones et al., 2018). Under normal circumstances, a lack of available nutrients drives AgRP neurons to initiate feeding behaviour. In mice, it has been shown that the desire to feed does necessarily involve AgRP neuronal activation when the food has been enriched with sugar and fat (Denis et al., 2015). In this case dopamine signaling initiates the desire to eat and feeding behaviour becomes driven by reward, through the neural circuits involved in emotion, as opposed to the orexigenic pathway induced by metabolic need. This appears to be the case when eating for comfort (Denis et al., 2015). Further examinations directly link sucrose consumption with the orexigenic pathway. In two separate studies sucrose intake was found to be increased following NPY infusion into the lateral ventricle or AgRP administration into the NAc shell (Badia‐Elder et al., 2003; Pandit et al., 2015). The effects of AgRP administration into the NAc shell is, however, halted by pre-treatment of α-flupenthixol, a non-selective dopamine receptor antagonist Pandit et al., 2015). Furthermore, chemogenetic and optogenetic manipulation of AgRP neurons activity modulates the emotional valence of feeding. Specifically, arcuate nucleus AgRP neurons were shown to regulate the emotional aspects of feeding involved with anxiety, fear-like behaviour and aggression (Sweeney and Yang, 2017; Burnett et al., 2016; Joly-Amado et al., 2014; Padilla et al., 2016).

The second, or anorexigenic pathway contains proopiomelanocortin (POMC), a pre-cursor to melanocortin receptor agonist α-MS, and cocaine- and amphetamine-regulated transcript (CART). This pathway is used to inhibit the consumption of food (Benoit et al., 2008; Lau and Herzog, 2014). Disruption of the genes encoding POMC, CART or the melanocortin-4 receptor may lead to obesity (Graham et al., 1997). Mice fed a high fat diet and then provided with sucrose-sweetened water showed a down-regulation of POMC mRNA expression‎ in the hypothalamus (Soto et al., 2015). Prolonged limited access to sucrose lead to decreased activity of the anorexigenic oxytocin system (associated with satiety and termination of feeding), in the hypothalamus of rats (Mitra et al., 2010). Together, these findings suggest sugar plays a role in limiting the activation of this pathway, thus increasing the desire to feed.

Neural pathways connecting feeding centres of the brain to the limbic system have been identified through optogenetic activation of melanin-concentrating hormone neurons during the intake of an artificial sweetener (sucralose) (Domingos et al., 2013). Sucralose was found to increase striatal dopamine levels which transposed the preference to sugar normally shown by mice into a preference for the sweetener (Domingos et al., 2013). Further to this, Domingos and colleagues showed that melanin-concentrating hormone neurons projecting to reward areas are necessary for the rewarding effects of sucrose (Domingos et al., 2013).

Furthermore, the serotonergic system is indicated in the regulation of hedonic feeding, with increases of serotonin causing decreased food intake and decreased serotonin increasing motivation to feed (Halford et al., 2010). The suggested mechanism of action for serotonin in feeding behaviour is via AgRP and POMC neurons in the arcuate nucleus (Sohn et al., 2013). The serotonin transporter-linked polymorphic region (5-HTTLPR) regulates vulnerability to stress, which is increased in cases of pathological fear, and influences energy intake, suggesting its role on stress related overeating (Capello and Markus, 2014). Neural adaptations resulting from sugar dependence in rodents include alterations in dopamine and opioid receptor binding in the mesolimbic cortex, changed expression‎ of encephalin mRNA and modified NAc release of dopamine and acetylcholine (Avena et al., 2008).

3.1.1. Leptin and ghrelin

There are several alternate neuronal populations, pathways and brain nuclei which drive overeating behaviour and sucrose consumption. Along with hypothalamic orexin neurons, MCH producing neurons and leptin-receptor cells projecting from the lateral hypothalamus to the VTA also influence hedonic feeding and reward seeking behaviour (Bocarsly, 2018). High fructose intake results in lower insulin levels, decreased levels of leptin and increased concentrations of ghrelin when compared to meals high in glucose (Teff et al., 2004). Alteration to levels of these satiety hormones (which result in the feeling of ‘fullness’ following food consumption) may be a precursor to overeating. These results concur with those found after intracerebroventricular injection of concentrated fructose or glucose into the hypothalamus of rats exposed to 2-deoxy-Dglucose (2DG) (Miller et al., 2002). DG is an analogue of glucose that cannot be metabolized and is known to cause increased food intake in rats by interfering with the process of glycolysis (Fiorentini and Müller, 1975). Rats provided with fructose injections showed enhanced food intake both in the presence and absence of 2DG, where glucose suppressed 2DG induced food intake suggesting a role for brain glucoreceptors in the control of food consumption (Miller et al., 2002). Glucoreceptors monitor blood glucose levels and are located within the hypothalamus. They have been shown to be stimulated by low blood sugar levels to increase feeding behaviour (Ritter et al., 1981).

Dopaminergic neurons projecting from the VTA to the NAc are inhibited by leptin and insulin and stimulated by ghrelin (Palmiter, 2007). Overconsumption of sugar has been shown to increase dopamine D1 receptor binding in the NAc core and shell, decrease dopamine D2 binding in the dorsal striatum and increase binding to dopamine transporters in the midbrain (Colantuoni et al., 2001; Bello et al., 2002). Increases in dopamine delays the release of acetylcholine during feeding, which postpones satiety and paves the way for overconsumption (Hoebel et al., 2009). Alternatively, opioid receptor binding (mu-1) was increased in the NAc shell, hippocampus, locus coeruleus and cingulate cortex after rats were permitted intermittent binging on a 25% glucose solution (Colantuoni et al., 2001). Together these findings suggest the overconsumption of sugar may sensitize dopamine D1 and opioid mu-1 receptors in a similar manner to drugs of dependence. Table 2 depicts studies examining the effect of sugar consumption on brain regions, molecular mechanisms and behavioural changes involved in the orexigenic pathway.

Table 2. The effects of sugar consumption on the orexigenic pathway.

Sugar consumption causes neural adaptation to brain regions connected to the orexigenic pathwaySugar consumption alters activation of molecular substratesSugar consumption alters behaviour↑ Dopamine↓Acetylcholine↓ Corticosterone↓CRH mRNA↑ Orexin / hypocretin↓BDNF↑AgRP↑5-HT2CRs↑Mu-1 binding↑TNF-α↑NPY

Hypothalamus Hippocampus	↓Satiety (Hoebel et al., 2009)
	↓Stress (Kinzig et al., 2008)
	↑Cue-induced over-consumption (Wise et al., 1995)
	↓Spatial learning (Molteni et al., 2002)
	↑Stimulation to feed (Badia‐Elder et al., 2003; Pandit et al., 2015)
	↓Incentive motivation (Valencia-Torres et al., 2017)
	↑Overconsumption (Colantuoni et al., 2001)
	↓Neurogenesis (Van der Borght et al., 2011)
	↑Desire to feed ↑Learned cue associations (Badia‐Elder et al., 2003)

3.2. Compulsive sucrose seeking

Since the 1980’s it has been known that neurons from the lateral hypothalamus encode reward-associated cues (Nakamura et al., 1987) and both feeding (Schwartzbaum, 1988) and drinking (Tabuchi et al., 2002) behaviour. In 2015 an attempt was made to define the neural circuits specific to compulsive sucrose seeking (Nieh et al., 2015). Photoinhibition was employed to show a selective pathway from the lateral hypothalamus to the VTA reduced compulsive sucrose-seeking in mice without affecting normal feeding behaviour (Nieh et al., 2015). The study found a bidirectional circuit of both inhibitory and excitatory projections from the lateral hypothalamus to dopaminergic and GABAergic neurons in the VTA (Nieh et al., 2015). It is possible that this circuit increases the focus and intensity of sucrose motivation, highlighting it as a potential therapeutic target for the compulsive overeating of sugar.

As obesity occurs due to excess caloric intake, the incentive value of the food consumed must be considered in parallel to the satiation of hunger. 5-hydroxytryptamine 2C receptors (5-HT2CRs) play a role in incentive motivation via hypothalamic-VTA feeding circuits and have therefore been considered a target for obesity treatments (Valencia-Torres et al., 2017). Lorcaserin, a 5-HT2CR agonist was administered to mice and shown to reduce both standard food intake and the desire for chocolate pellets with a corresponding increase in c-fos expression‎ in VTA 5-HT2CR GABAergic neurons (not dopaminergic neurons) (Valencia-Torres et al., 2017), validating their role in the inhibition of motivational behaviour. Similar findings arose from observations of sucrose drinking where leptin administered to the VTA was found to reduce food intake and knockdown of the leptin receptors increased sucrose seeking (Clifton, 2017). Interestingly, mediation of the orexigenic pathway between the VTA and NAc may affect behavioural changes towards reward seeking for drugs of abuse, including withdrawal and relapse behaviour (Kastin, 2013; Carr and Kalivas, 2006).

Decreased dopamine levels and simultaneous increased acetylcholine levels are associated with behavioural signs of withdrawal from drugs of abuse (Rada et al., 2001; Rada and Hoebel, 2005). Rats provided intermittent sugar access show similar imbalances to initiate withdrawal after provision of the opioid receptor antagonist naloxone (Colantuoni et al., 2002). An animal will crave a substance of abuse it has been deprived, as shown by elevated operant responding even when unrewarded and increased responding to associated drug cues (Bienkowski et al., 2004; Lu et al., 2004). Rats deprived of sugar after glucose overconsumption (i.e. permitted 25% glucose for 30 min per day for 28 days and glucose access in their home cages for an additional 11.5 h per day) responded significantly more in operant chambers compared to controls suggesting that sensitization of the dopaminergic system and associative learning leads to increased motivation to seek sugar (Avena et al., 2005). The neural adaptations required to cause the behavioural change appear long lasting as sugar consumption ceased 2 weeks prior to testing. Similar results were shown in rat studies of alcohol consumption (Heyser et al., 1997).

3.3. Sucrose consumption and the hypothalamic-pituitary-adrenal axis

Long term stress, depending on its severity, appears to correlate with a preference for high sugar foods, suggesting its contribution to the progression of obesity. Chronic stress may develop through an accumulation of physical (traumatic), chemical (dietary), physiologic (painful), psychologic (fear) or social (lifestyle) stressors (Powers and Howley, 2007). Extended periods of stress result in hyper activation of the hypothalamic-pituitary-adrenal (HPA) axis, a mammalian stress response system involving the endocrine and central nervous systems (Smith and Vale, 2006). Hyper activation of the HPA axis leads to increased release of corticosteroids (steroid hormones produced in the adrenal cortex) see Fig. 3 (Smith and Vale, 2006; BjÖrntorp, 2000).

1. Download : Download high-res image (423KB)
2. Download : Download full-size image

Fig. 3. Hypothalamic-pituitary-adrenal axis. Stress causes the release of corticotrophin-releasing hormone and vasopressin from the hypothalamus. These hormones are transported to the anterior pituitary. Adrenocorticotropic hormone (ACTH) is then released from the pituitary gland which stimulates the adrenal glands to release cortisol, catecholamines and aldosterone into the bloodstream. Cortisol released from the adrenal glands increases a desire for palatable foods such as sugar and targets systemic organs, including the brain where it exerts a feedback inhibition on the release of CRH by the hypothalamus and ACTH by the pituitary gland, via corticoid receptors (mostly glucocorticoid receptors (GRs), and to a lesser extent mineralocorticoid receptors).

Activation of the HPA axis leads to increased adrenocorticotropic hormone (ACTH) levels and increased glucocorticoids, which affect the utilization of energy stores (Ulrich-Lai et al., 2010). This is of particular importance, as people regularly report choosing to consume sweet tasting food (of higher caloric content) due to its ability to enhance their mood and relieve negative emotional states (Smith and Vale, 2006). In an observation of rats undergoing acute stress, the provision of a sucrose solution significantly reduced levels of both ACTH and corticosterone secretion (Ulrich-Lai et al., 2010). Furthermore, rats fed with a sweetener (Saccharin), also showed lower HPA axis responses to acute stress suggesting the hedonic nature of sugar may be responsible for the reduced stress response (Ulrich-Lai et al., 2010).

Contrastingly, a study involving nineteen women found consumption of sucrose but not the sweetener (Aspartame) resulted in reduced reduced stress-induced cortisol (Tryon et al., 2015). Sucrose and not Aspartame also produced greater activity in the left hippocampus (Tryon et al., 2015) suggesting the HPA negative feedback loop may assist in generating a tendency towards sugar consumption in people dealing with stress. The following section reviews the complex relationship between the HPA axis, food intake, energy stores and chronic stress.

Any discerned acute stressor may cause engagement of the HPA axis and result in an emotional response eliciting fear, anxiety or defensive behaviours designed to maximize chances of survival (Ulrich-Lai et al., 2010). The relationship between stress, cortisol levels and sugar consumption is far from understood, however, it appears that high sugar diets play a significant role in HPA axis regulation and binge-like eating associated with stress. When ingested, foods high in sugar release neuropeptides, elevating mood and reinforcing desire or selective preference towards greater amounts of high caloric food (Ulrich-Lai et al., 2011). Sucrose and the sweetener saccharin have been shown to dampen the HPA axis response and this effect is possibly why highly palatable foods are consumed to assuage unpleasant emotions (Ulrich-Lai et al., 2011). Sugar consumption regulates stress-like behaviour via the HPA axis however, the precise mechanisms of action are yet to be elucidated.

Contrarily, examples of chronic stress reveal elevated levels of glucocorticoids with increased risk of developing stress related illnesses such as depression. Long term stress appears to alter brain function through changes in negative feedback loops from energy stores and glucocorticoid modulation of neural circuits (for review see (Dallman et al., 2003)). As glucocorticoid levels are tightly controlled by the HPA axis, one avenue of investigation has been to examine levels of mRNA expression‎ of 11βHSD-1(a mediator of glucocorticoid metabolism in the liver) in cases of early life stress (Maniam et al., 2015). Interested in the role of sugar intake on lipid homeostasis, researchers found a 53% increase in transcriptional levels of 11βHSD-1 in animals permitted chronic sugar (25% sucrose solution) consumption in contrast to controls revealing a correlation between sucrose consumption, stress and increased glucocorticoid metabolism (Maniam et al., 2015). Additionally, when humans are exposed to stress at a young age, they become more likely to develop anxiety and depression associated with dysregulation of the HPA axis (Heim and Nemeroff, 2001). Increased levels of cortisol (the resultant product of the HPA axis in humans) have also been positively correlated with higher visceral fat deposits and insulin resistance (Misra et al., 2008). Similarly high cortisol levels (22% higher) were reported in overweight/obese adolescents after 2 or more sugar-sweetened beverages per day, suggesting significant increases in stress hormone were not only due to visceral fat deposits but also increased sugar consumption (Shearrer et al., 2016). One suggested mechanism may include morphological alteration of the adrenal glands as shown through consumption of high sugar content beverages which increased the risk of metabolic syndromes such as diabetes and lead to dysfunction of the adrenal glands (Díaz-Aguila et al., 2016). Young adult rats fed a 30% sucrose solution for 12 weeks showed increased visceral fat deposits and insulin resistance (Díaz-Aguila et al., 2016). Interestingly, the adrenal glands showed histomorphological changes in the adrenal cortex and medulla due to sucrose consumption, suggesting hyperplasia and indicators of metabolic syndrome (Díaz-Aguila et al., 2016).

3.4. Developmental neuroadaptation

In rodents prenatal stress can decrease glucocorticoid and mineralocorticoid receptor levels in the hippocampus thus decreasing receptor availability for the feedback inhibition of corticosterone, which may explain why novel stressors cause an increased and longer lasting corticosterone response (Henry et al., 1994). Pregnant rats underwent restraint stress during the third week of gestation (Henry et al., 1994). The HPA axis and hippocampal corticosteroid receptors in the male offspring were investigated. Plasma corticosterone was found to be significantly higher in the prenatally-stressed rats compared to controls and the receptor subtypes (hippocampal type I and type II corticosteroid receptors) were decreased in the hippocampus suggesting long term changes in the HPA axis may occur following prenatal stress (Henry et al., 1994). Numerous studies have examined maternal stress and corticosteroids, for a comprehensive review see (Welberg and Seckl, 2001). Fewer studies involve these neuroadaptations and diet, therefore more direct investigation of high sugar consumption in mothers may be significant in the study of childhood obesity.

Furthermore, evidence exists for the pre-programming of HPA axis hyper activation during prenatal periods (Welberg and Seckl, 2001). In adolescents deemed to be at higher risk of becoming obese (by eval‎uating their parent’s body mass index), a higher functional magnetic resonance imaging (fMRI) signal was observed in brain regions associated with reward learning, processing and motivation after the consumption of high sugar milkshakes as compared to high fat milkshakes (Shearrer et al., 2018a). Although fat provides greater energy and therefore contributes to obesity by providing excess calories, sugar is more often associated with modulation of habitual overeating and therefore the addiction-like behaviour associated with obesity (Shearrer et al., 2018a).

Overconsumption during adolescence leads to long-lasting changes in the dopaminergic reward system and may cause stimuli-induced sensitization that is observed in adulthood (Naneix et al., 2018; Robinson and Berridge, 2008). Stimuli-induced sensitization refers to an increased effect of the stimuli, following repeated exposures eg. repeated exposure to a loud noise may create a sensitization to noise, generating an enhanced response, the same being true for repeated exposures to a drug of abuse. In drug addiction, sensitization causes changes in dopamine transmission and delta FosB expression‎ which contributes to increased craving and relapse (Robinson and Berridge, 2001).

Adolescent rats permitted sucrose consumption (5% sucrose solution) for sixteen days were later tested for motivation to seek either saccharin, maltodextrin or cocaine (Vendruscolo et al., 2010). The adult rats showed a reduction in motivation to procure saccharin and maltodextrin, however this was not the case for cocaine (Vendruscolo et al., 2010). Further to these findings, rats presented with a choice between a saccharin solution and intravenous cocaine demonstrated the reward provided by the sweetness surpassed the desire for cocaine (Lenoir et al., 2007); an intense debate remains as to whether sugar is in fact addictive (Benton, 2010; Corsica and Pelchat, 2010; Wilson, 2010).

It has been hypothesized the addictive-like behaviours present after long-term sucrose consumption arise from the highly palatable nature of sugar and not the neurochemical effects of sugar itself (for reviews see (Ahmed et al., 2013; Hoebel et al., 2009; Westwater et al., 2016; Avena et al., 2009; Criscitelli and Avena, 2017; Kendig, 2014)). The purpose of this review is not to debate the argument, but to examine the role of sugar in emotional dysfunction.

4. Common neurochemistry underlie consumptive behaviours and emotions

Neuroimaging studies and animal models investigating the mechanisms involved in the progression to obesity have revealed neurobiological correlates participating in the neuroadaptations of obesity and sugar consumption (Lindgren et al., 2018; Novelle and Diéguez, 2018). Preclinical rodent models of consumption are useful for investigating neural regions and pathways underlying consumptive behaviours (Holgate et al., 2017; Bonin et al., 2018) with intermittent access to sucrose shown to have an effect on opioid, cholinergic and, importantly, dopaminergic receptors (Shariff et al., 2016; Avena et al., 2008; Colantuoni et al., 2002; Pratt and Kelley, 2004).

4.1. Opioids

Opioid receptors are expressed throughout the limbic system and play key roles in the regulation of fear, happiness, anger, arousal, motivation and reward related feeding (Nummenmaa and Tuominen, 2017; Levine and Billington, 2004). Opioid-induced behaviour includes the modulation of pain, drug addiction and control of the autonomic nervous system, which includes emotions, cognitive processes and stress coping (Benarroch, 2012; Belzeaux et al., 2018). Opioids influence the way we process rewards through interaction with the dopamine system (Bray and Bouchard, 2014). The way an animal behaves towards food is also modulated by endogenous opioid neurotransmitters, which are neural substrates that create susceptibility to reward (Wise, 1996; Kavaliers and Hirst, 1987). Alterations in hedonic processing are demonstrated to be highly dependent on endogenous opioid receptors in the NAc suggesting these changes may result in the development of reward-related disorders such as obesity (Naneix et al., 2016).

Opioid peptides (β-endorphin, enkephalins, and dynorphins), and activation of their receptor types, μ (mu), δ (delta) and κ (kappa) in the NAc cause inhibition at both pre and post synapses (Benarroch, 2012). Opioid agonists modulate pleasure, reward and reinforcement through activation of mu and delta receptor ligands via the mesolimbic dopamine system whereas the dysphoria associated with withdrawal relies on kappa receptors (Nummenmaa and Tuominen, 2017; Benarroch, 2012; Herz, 1997). Mu receptor binding in the NAc is significantly increased after cocaine (Bailey et al., 2005), morphine (Vigano et al., 2003) and sucrose (Colantuoni et al., 2001).

In the endeavor to define specific peptide and receptor involvement in the motivation and reward seeking behaviour characteristic of hedonistic feeding, several animal studies have combined opioid antagonists and agonists with sucrose consumption (Hayward et al., 2006; Beczkowska et al., 1992; Ruegg et al., 1997). Opioid receptor antagonists (naltrexone and naloxone), used to treat drug and alcohol dependence, were shown to use kappa and mu2 binding sites to inhibit sucrose intake, thus demonstrating the role of endogenous opioids in regulation of hedonic rewards in rats (Beczkowska et al., 1992).

Substance P and the NK1 receptor system, interact with opioid receptor systems to regulate reward related behaviours, suggesting its role in addictive behaviour (Steensland et al., 2010). Substance P is also a neuropeptide that affects the orexigenic pathway, via the NK1 receptor (Karagiannides et al., 2008). Its presence in the brain assists the regulation of feeding, however, it is also located in the stomach and small intestine suggesting it may be a potential therapeutic target for obesity. Administration of a substance P antagonist (CJ 012,255) was shown to prevent weight gain in obese mice following a 2 w high fat diet while treatment with CJ 012,255 in obese mice resulted in a loss of weight loss and reduction improved insulin sensitivity, partially due to reduced food intake (Karagiannides et al., 2008).

Using a two-bottle choice drinking paradigm, knockout mouse models (lacking either one or two opioid peptides) were used to identify opioid receptor ligands modulating sucrose preference (Hayward et al., 2006). Enkephalin and dynorphin were found to modulate preference to sucrose but were not considered necessary for its consumption, supporting the role these peptides play in the motivation to consume unnecessary calories208. A decrease in enkephalin mRNA expression‎ was observed in the NAc after rats were provided with intermittent access to sugar (Spangler et al., 2004). The link between opioids, dopamine and motivation for hedonic rewards is further reinforced with data showing rats preference for sweet taste after the introduction of morphine to the NAc (Berridge, 1996; Pecina and Berridge, 1995).

4.2. Acetylcholine

Neuronal nicotinic acetylcholine receptors (nAChRs) act as biological targets for ethanol, nicotine (Bito‐Onon et al., 2011) and sucrose (Shariff et al., 2016; Löf et al., 2010) and modulate the neurotransmission of GABA, glutamate, dopamine, serotonin, acetylcholine and noradrenaline (Wu, 2009). Homeostatic dysregulation of acetylcholine in the limbic system has been shown to modulate the motivation and reward seeking behaviours characteristic of addiction and relapse (Hoebel et al., 2007; Aosaki et al., 2010; Rahman et al., 2015), neurodegeneration (Alzheimer’s disease, mild cognitive impairment) and mental illness (anxiety, depression and schizophrenia) (Wu, 2009; Posadas et al., 2013). Within the NAc, cholinergic interneurons regulate the expression‎ and release of enkephalin (Kelley et al., 2005) and many rat studies support the theory that they inhibit feeding behaviou (Avena et al., 2006; Hoebel et al., 2007; Mark et al., 2006). As increased serotonin reduces the motivation to feed, a study using rats to perform a progressive ratio task and paroxetine (a selective serotonin reuptake inhibitor) found that increased serotonergic activity decreased the appetitive-based responses to the hedonic taste of sucrose and the aversive taste of quinine (Mathes et al., 2013). Withdrawal from morphine increases acetylcholine levels in the NAc, while dopamine simultaneously remains low, leading to the assertion that this mechanism is involved in the unpleasant aspects associated with withdrawal (Avena et al., 2008; Rada et al., 2001). Arecholine (a muscarinic agonist) inhibits feeding but can be blocked by pirenzapine (a muscarinic acetylcholine 1 receptor antagonist), suggesting a role for acetylcholine in food intake (Avena et al., 2006).

In 2010, Lof and associates utilized rats to demonstrate the role of nAChRs in the conditioned reinforcement of lever pressing for a sucrose reward (Löf et al., 2010). Using a selective antagonist at α7 nAChRs (methyllycaconitine) they showed reduced lever pressing for the sucrose solution down to the level of the control, suggesting mediation of the amount of influence the cue had on the desirability of the sucrose (Löf et al., 2010). More recently our laboratory elucidated the effects of varenicline (an FDA-approved drug to reduce nicotine cravings) on sucrose consumption in rats (Shariff et al., 2016). Varenicline, a partial agonist at α4β2* and antagonist α6β2* nAChRs subtypes, significantly reduced sucrose consumption in both short and long term binge like (intermittent access) drinking. While α4β2* nAChR binding sites were increased, α6β2* nAChRs were significantly decreased in the NAc as a result of both short-term and long-term sucrose consumption (Shariff et al., 2016). By modulating dopamine release in the NAc (Feduccia et al., 2014) the control of sucrose consumption by nicotinic receptors and their subtypes may provide promising therapeutic strategies for obesity (Cocores and Gold, 2008).

4.3. Dopamine

Sucrose is considered a primary or natural reinforcer that does not require a learning process to be considered desirable (Corsini, 1999). Primary reinforcers are known to trigger dopaminergic activation in the VTA (Wise, 1996; Olds and Olds, 1963; Moore et al., 2018), which act on the NAc via the medial forebrain bundles (Yu and Kim, 2012). Activation of dopaminergic neurons in the VTA modulates processes of memory, learning and motivation (Yang et al., 2018; Adcock et al., 2006; Otmakhova et al., 2013; Wise, 2004) as well as the intense emotions associated with love, sexual desire, orgasm (Ulrich et al., 2018; Holstege et al., 2003; Acevedo et al., 2012), fear, stress, anxiety (Bouchet et al., 2018; Zhang et al., 2018b; Moghaddam, 2018) and psychosis (De Nijs et al., 2018; Knolle et al., 2018). Rat models of obesity display low levels of dopamine and impairment in the release of dopamine (Geiger et al., 2008).

The mesocorticolimbic dopaminergic system has been implicated in modulation of addictive reward seeking behaviour, (for review see (Liu, 2018)). Drugs of abuse such as alcohol, nicotine, cocaine and amphetamines that cause increased extracellular dopamine and subsequent heightened pleasure levels in humans (Volkow et al., 2002; Robinson and Berridge, 1993) also cause increases in dopamine release in the NAc in rats (Wise et al., 1995; Di Chiara and Imperato, 1988). Commonly abused (opiates, ethanol, cocaine, amphetamine and nicotine) given to rats resulted in increased extracellular dopamine in the NAc and the dorsal caudate nucleus Di Chiara and Imperato, 1988). Similar results were found in rats permitted intermittent access to sugar with subsequent binge drinking resulting in increased dopamine release in the NAc (Avena et al., 2006; Rada et al., 2005).

Supporting both the hedonic potency of sucrose and its reliance on dopaminergic regulation, orosensory stimulation in sucrose sham-fed rats resulted in increased dopamine release within the NAc, thus increasing reward seeking behaviour (Hajnal et al., 2004; Schneider, 1989). Compelling evidence shows drugs not routinely abused, such as imipramine, atropine and diphenhydramine do not display rewarding properties and do not alter synaptic dopamine levels (Di Chiara and Imperato, 1988). Studies show an attenuated dopamine system in obesity prone rats in the NAc, PFC and dorsal striatum and suggest the reduction in hedonic response available to these animals is related to hyperphagia (increased appetite) and resultant obesity (Geiger et al., 2008). These findings were similar to those of a human study showing reduced levels of dopamine D2 receptors correlated with increased body mass index in 10 obese individuals (Wang et al., 2001).

4.3.1. Dopamine sensitization

In 1993 Robinson and Berridge postulated the incentive-sensitization theory, which suggested that repeated exposure to rewarding substances sensitizes the dopaminergic system resulting in exaggerated “cue-triggered wanting” which can transform into a compulsion to seek out an associated reward (Wyvell and Berridge, 2001; Robinson and Berridge, 1993). Cue-reward learning depicts the learned association between the rewarding effects of substances of abuse and environmental cues (Bianchi et al., 2018; Tye et al., 2008). This learning contributes to obsessive overuse and the tendency to relapse after long periods of abstinence (Bianchi et al., 2018; Ziauddeen et al., 2012). Repeated or prolonged substance abuse can cause modifications to neurotransmitter release and alterations in synaptic strengths (Gerdeman et al., 2003). For example, rats provided with intermittent access to sucrose display signs of dopamine sensitization through altered dopamine receptor function (Sharpe et al., 2016).

Behavioural sensitization or hyper-locomotion is a motor response that increases incrementally with repeated exposures to a drug. It is reflective of a hyposensitized (or attenuated) reward pathway and is said to contribute to the craving or compulsive seeking which characterizes addictive behaviour (Nona and Nobrega, 2018; Robinson and Berridge, 2001). In both human and rodent studies, behavioural sensitization has been observed as dose-dependent (Jing et al., 2018; Huber et al., 2018) and shows considerable individual variation (Robinson, 1988). Brain-derived neurotrophic factor (BDNF) is believed to modulate this behaviour through overexpression‎ of the dopamine D3 receptor (Bordet et al., 1997) and has been implicated in drug addiction, schizophrenia (Guillin et al., 2001; Staley and Mash, 1996; Gurevich et al., 1997) and novelty seeking behaviour in obese rats (Savage et al., 2014).

Animals sensitized to one drug are often cross-sensitized to other substances of abuse as well as non-drug substances (Avena et al., 2008). This has been demonstrated with rats exposed to amphetamines becoming sensitized to cocaine (Pierce and Kalivas, 1995), cocaine cross-sensitizing to alcohol (Itzhak and Martin, 1999) and cocaine cross-sensitizing with stress (Covington and Miczek, 2001). Similar to these, short term sucrose consumption increases binding affinity to opiates (Avena et al., 2008), resulting in cross tolerance to other opioids such as morphine, while chronic sucrose consumption reduces the analgesic properties of morphine (Steensland et al., 2010; D’Anci et al., 1996; Fidler et al., 1993; Kanarek et al., 1991). A number of studies have also shown opioid agonists such as morphine to increase the impulsivity to feed when injected systemically or directly into the NAc, paraventricular nucleus, AMG, hypothalamus and tegmentum (Gosnell, 1987; Stanley et al., 1988; Woods and Leibowitz, 1985).

Further investigations have demonstrated how sugar access can lead to cross-sensitization of dopamine-altering drugs. Intermittent access to sugar has been found to cross-sensitize with amphetamines up to 8 days post sugar ingestion (Avena and Hoebel, 2003) Female rats provided with either a 10% sucrose solution or a rotation of sucrose solution and withdrawal, displayed behavioural cross-sensitization to a small dose of amphetamine (Avena and Hoebel, 2003; Avena et al., 2008). Similar results for sucrose and cocaine (Gosnell, 2005) and sucrose and quinpirole (a dopamine agonist) have also been found (Foley et al., 2006). The rats maintained on the cyclic sucrose solution responded hyperactively to amphetamine in comparison to controls, suggesting the binge-like sugar consumption leads to increased amphetamine sensitivity, which may be due to neuroadaptations of the dopamine system (Avena and Hoebel, 2003). Compelling evidence suggests chronic stress leads to similar cross-sensitization (Leão et al., 2012; Koob, 2009b). As globally the consumption of sugar is on the rise an interesting avenue of investigation may be to examine the effect of sucrose consumption on stress derived cross sensitization.

4.3.2. Dopamine and impulsivity

Dopamine is released in response to food cues, making it essential for the motivational prompt to consume food (Flagel et al., 2011; Salamone and Correa, 2012). As dopamine neurons project from the VTA to the NAc, caudate, putamen, PFC, hippocampus and AMG, they modulate a wide variety of behaviours and emotions (de Jong, 2015) including impulsivity (Wade et al., 2000). In relation to eating disorders such as obesity, there appears to be an impulsive tendency to overconsume sweet foods in response to cravings for carbohydrates and sugars. When a subject is withdrawn from an addictive substance, impulsivity increases and can be quantified using a behavioural test called differential-reinforcement-of-low-rate-schedule performance (DRL) (Mangabeira et al., 2015). The test requires the subject to withhold a response for a set period of time before they are permitted to respond and thereby earn a reward (Kirshenbaum et al., 2008). Deprivation after long term sugar consumption in rats resulted in impairment of DRL performance, confirm‎ing its resemblance with drug addiction and suggesting an increase in impulsive behaviour following sugar deprivation (Mangabeira et al., 2015). The suggested mechanism of action for this behaviour is the dopaminergic system as dopamine receptor antagonists (pimozide (Wise, 2006), raclopride (Maldonado et al., 2006), SCH 23390 (Verty et al., 2004)) can attenuate the desirability of highly palatable foods in rats (Mangabeira et al., 2015; Wise, 2006).

The progression towards obesity is linked with a decline in neural responses to reward similarly induced by cocaine and heroin, which moderate the transition from casual user to the impulsive actions of a compulsive drug-taker (Johnson and Kenny, 2010). Impulsive actions such as over eating are modulated by the excitation of D1 dopaminergic neurons and inhibition of D2 dopaminergic neurons in the dorsal striatum and the NAc (Johnson and Kenny, 2010). Preclinical and clinical studies have been conducted to investigate the role of dopamine in obesity with findings showing obese subjects have an inverse relationship between the abundance of dopamine D2Rs and body mass index, suggesting a dopaminergic role in compulsive eating (Wang et al., 2001; Johnson and Kenny, 2010; Volkow and Wise, 2005; Pijl, 2003). Male rats self-administering amphetamine showed a reduction in the ability of D2Rs to inhibit dopamine release in the NAc supporting the theory that desensitization due to short term substance abuse modulates addiction-related behaviours such as impulsivity (Calipari et al., 2014; Bello et al., 2011; Cools et al., 2003). In a human study, twelve patients with Parkinson’s disease, a disorder of the mesolimbic dopamine system, underwent several cognitive tests designed to provide a measure of impulsivity. Increased impulsive behaviour was found to correlate with greater dopamine levels Cools et al., 2003).

The neural mechanisms modulating motivation and drive are implicated in both the loss of self-control exhibited by over eating and addiction to drugs of abuse (Avena et al., 2008; Berridge, 1996). Neurofunctional imaging has documented neuroadaptations in the PFC (Volkow and Fowler, 2000), hippocampus (Wang et al., 2006), AMG and NAc (Volkow and Wise, 2005). A functional MRI study was conducted showing female adolescents an image of either highly palatable food (dessert) or a vegetable (Batterink et al., 2010). In comparison to leaner adolescents, overweight adolescents had increased behavioural impulsivity and decreased neural activation in frontal inhibitory regions of the brain (e.g. medial prefrontal cortex, and orbitofrontal cortex) (Batterink et al., 2010). In addition, activation of the reward region was positively correlated with body mass index when shown the picture of dessert (Batterink et al., 2010). These findings suggest the reduced functioning of inhibitory controls combined with increased responses in food reward regions are relevant to weight gain (Batterink et al., 2010). The Stroop test requires an individual to suppress their automatic response to certain stimuli. This was used to assess the inhibitory control of overweight / obese 10 y old children in comparison to normal weight children (Reyes et al., 2015). Results indicated changes to inhibitory control functions in overweight children suggesting less emotional self-regulation, which in conjunction with dysfunctional impulsivity control may contribute to over eating behaviour in the overweight/ obese children (Reyes et al., 2015).

5. Common anatomical structures and neural substrates of stress driven, emotional behaviour

The insular cortex and cingulate cortex (part of the limbic system) are responsible for the processing of emotions in conjunction with processes involving higher cognition (Rajmohan and Mohandas, 2007). The hippocampus provides negative feedback for the HPA axis with both neuroadaptations to its volume and capacity for neurogenesis implicated in emotional disorders (Jacobson and Sapolsky, 1991). The AMG processes, stores and retrieves fear memories, and initiates appropriate behavioural responses (Martin et al., 2009). It is responsible for the manner in which we express fear, aggression and defensive behaviour (Martin et al., 2009). The PFC regulates executive functions and reward processing while impulse control and mood are modulated by the orbitofrontal cortex (Martin et al., 2009). Emotional disorders involve a complex interconnected number of neuroendocrine, neuropeptide, neurotransmitter and neuroanatomical adaptations (Gulpinar and Yegen, 2004). These alterations may occur due to genetic predisposition or as a result of environmental influence (Martin et al., 2009). We assert that abundant evidence exists regarding the role overconsumption of sugar plays in altering these brain regions and contributing to emotional disorders such as depression, anxiety and fear.

5.1. Anxiety

Generalized anxiety disorder is defined by uncontrolled, exaggerated concern over numerous endeavours Association, 2019. Treatment often consists of anxiolytic drugs that act primarily on the monoamines serotonin, noradrenalin and dopamine (Martin et al., 2009). In the central nervous system these neurotransmitters are released in conjunction with neuropeptides that have strong links to anxiety, such as neuropeptide Y and cholecystokinin expressed in the limbic cortex where they influence emotions and stress levels (Martin et al., 2009). Other molecular substrates involved in regulating the stress response include corticotropin-releasing factor and adrenocorticotropic hormone in the HPA axis (Patriquin and Mathew, 2017). In cases of anxiety, those carrying the short-allele of the serotonin transporter gene (5-HTTLPR) demonstrate over activation of the AMGs in response to viewing fearful and angry faces as compared to those carrying the long-allele (Hariri et al., 2005).

Two specific brain circuits are notable in anxiety disorders, the first involves the dorsolateral PFC, anterior cingulate cortex, dorsal parietal cortex and precentral gyrus, all associated with cognitive control, executive function, flexible cognition, working memory and attention (Coplan et al., 2018). The second circuit is known as the negative affect circuit and as the name suggests is activated by negative stimuli (Coplan et al., 2018). Anatomically the hippocampus, AMG, medial PFC, and the dorsal, subgenual and pregenual regions of the anterior cingulate cortex may undergo neural adaptation to skew negative bias and threat by reducing attenuation through the cognitive control circuit (Coplan et al., 2018).

Human studies show links between high caloric food consumption and anxiety. Recent evidence from epidemiological studies found a suggestive link between greater consumption of processed foods and widespread presence of anxiety disorders (Jacka et al., 2011; Quirk et al., 2013; Westover and Marangell, 2002). There are also numerous studies linking adverse childhood experiences (early life stress) with an increased risk of body weight gain and obesity during adolescence and adulthood (Noll et al., 2007; Ramasubramanian et al., 2013; D’Argenio et al., 2009; Gunstad et al., 2006; Lumeng et al., 2013). More specific behavioural studies observing acute and chronic withdrawal from sugar in rodents show anxiety is induced when withdrawal follows extensive periods of sugar consumption (Avena et al., 2008). Long exposures (1 month of 12 h daily access) to palatable foods did result in rats showing increased anxiety-like behaviours on the elevated plus maze (EPM) when tested 24 h after withdrawal (Avena et al., 2008). Rats exposed to 12 h intermittent access to 10% sucrose solution for 28 days also displayed anxious behaviour in the EPM after 36 h withdrawal (Avena et al., 2008). These rats showed reduced conditioned suppression after 1 and 28 days of abstinence, measured by a failure to significantly reduce the number of lever-presses for sucrose during the presence of a tone stimulus paired with shock (Avena et al., 2008). Similar results have been reported using longer cycling periods whereby rats given intermittent access to a sucrose diet showed increased anxiety in the EPM and defensive withdrawal tests after 8 and 9 weeks on the diet cycle followed by 48 h withdrawal (Parylak et al., 2012). Alternate investigations have shown rats fed a carbohydrate rich diet to have increased protein oxidation in the frontal cortex which correlated with anxiogenic behaviour (Souza et al., 2007). These studies suggest a high probability of long term sugar consumption may contribute to symptoms of anxiety.

A long term study has shown an opposite effect in terms of anxiety and depressive like behaviours (Cao et al., 2007). Findings show that sugar consumption had no significant effects on motor activity in an open field (test for locomotor and anxious like behaviour), on exploration in a T-maze, or on anxiety in an EMP (Cao et al., 2007). Another study looking at the anxiety effects of sugar after a 1 year intervention of sucrose and honey found that anxiety decreased after 3 months of consumption, as tested with the EPM and open field test (Chepulis et al., 2009). Importantly, this reduction was maintained until the end of the experiment at 12 months (Chepulis et al., 2009). Despite this, honey-fed rats showed significantly less anxiety throughout the study as compared with those fed sucrose (Chepulis et al., 2009). Together this data suggests that chronic (i.e., greater than 1 month) intermittent access to high energy foods may increase anxiety-like behaviour followed by a plateau observed after 3months (Chepulis et al., 2009).

5.2. Depression

Depression is characterized by feelings of sadness, hopelessness and reduced pleasure in daily activities Association, 2019. Functional, structural and neurochemical factors associated with the pathophysiology of depression include dysregulation of the HPA axis resulting in reductions in hippocampal, PFC and striatal volumes (Koolschijn et al., 2009). In areas of emotional processing such as the AMG and PFC abnormal metabolism of glucose and changes to cerebral blood flow have also been demonstrated (Drevets, 1998). Although the association between obesity and depression is well documented, less is known about the physiological and corresponding psychological influence long-term sugar consumption has on depression (Luppino et al., 2010). Recent research suggests a possible contributor to the incidence rate of depression is sugar overconsumption (Westover and Marangell, 2002; Sánchez-Villegas et al., 2012). We suggest the neuroadaptations that occur to depression-related brain regions, namely the hippocampus, PFC and AMG, following sugar intake (reviewed above) contributes to the incidence of depression (Robinson et al., 2014; Moorman and Aston-Jones, 2014; Hsu et al., 2015).

The first study reporting a possible link between sugar and depression was conducted in 2002 by Westover and Marangell (Westover and Marangell (2002)). Data collected data from 6 countries showed a correlation between the consumption of sugar (calories/capita/day) with the yearly rate of major depression disorder (Westover and Marangell (2002)). Fast food further increases the potential of developing depression, and commercial baked goods are also positively correlated with depressive disorders (Sánchez-Villegas et al., 2012). Not only highly palatable food, but sweetened beverages (either artificially or otherwise) also contribute to depression (Westover and Marangell (2002); Sánchez-Villegas et al., 2012). Regular consumption of sweetened beverages indeed increases the occurrence of depression and suicidal tendencies (Shi et al., 2010; Pan et al., 2011; Lien et al., 2006; Guo et al., 2014).

In rats given the same high sugar/chow diet cycle for 7 weeks, highly palatable foods induce depressive-like behaviour as evidenced greater immobility in the Forced Swim Test (test for depression) and decreased preference for 0.8% sucrose (test for anhedonia) (Iemolo et al., 2012). Correlations with symptoms of depression have been observed following long-term exposure to sucrose, with long term overstimulation of the dopaminergic system during adolescence, which may occur through sucrose binging (Rada et al., 2005) resulting in deficits in later life that affect motivation, memory and happiness (Avena et al., 2008; Naneix et al., 2018; Kendig, 2014). When a model designed to generate a depressed-like state (known as social defeat-induced persistent stress) was used, rats showed increased motivation to acquire a sucrose reward and reinstated sucrose-seeking induced by a cue (Riga et al., 2015). Importantly, these studies showed the long-term effects of stress exposure induced deficits in the ability to eval‎uate natural rewards (Riga et al., 2015).

Congenitally helpless rats, a genetic model of predisposition to major depression, were used to respond less to the reward of sucrose solution (Vollmayr et al., 2004). This lack of motivation to partake in sweet rewards is conversely shown after long term sucrose consumption by adolescent rats, which results in depressive like behaviour in adulthood including symptoms of anhedonia (inability to feel pleasure) and increased anxiety-like behaviour (Gueye et al., 2018). Adult rats consuming sucrose showed a similar depressive-like behaviour, but to a lesser degree, suggesting the critical period of brain development that occurs during adolescence can be moderated by sucrose consumption and may increase the instance of disorders related to rewards, such as depression (Gueye et al., 2018).

As depression correlates with altered glucose metabolism, it is not surprising that comorbidity of diabetes mellitus (DM) and depression occurs quite frequently, greatly increasing the mortality risk (van Dooren et al., 2013). A study highlighting the link between glucose and depression in patients whose histories included DM and cardiac disease found there was a 34% increased chance of depression in subjects with elevated blood glucose levels (Azimova et al., 2015). A study including 70,000 postmenopausal women found a higher risk of developing depression in women consuming a high-sugar diet than those consuming high naturally occurring sugars (Gangwisch et al., 2015). However, research suggests sugar may contribute to depression more in the male population as was demonstrated by the Whitehall Study II which tracked dietary regimes and the corresponding medical health of 8000 participants over a 22 year period (Knüppel et al., 2017). Observing a five year duration they found a 23% increase in likelihood of men being diagnosed with depression if they consumed ≥ 67 g of sugar per day compared to ≤ 40 g (Knüppel et al., 2017). Cumulatively, these studies, while not able to confirm‎ that sugar causes depression, appear to produce enough evidence to show sugar overconsumption contributes to an increased risk of developing depression.

5.3. Neurogenesis

Neurogenesis is the term given to new neurons generated from neural stem cells (Ming and Song, 2011). The molecular substrates of stress, although not well understood are implicated in the regulation of adult neurogenesis, through molecular pathways modulated by glucocorticoids, inflammatory mediators and neurotrophic factors (Egeland et al., 2015). Studies on antipsychotic pharmacology support these findings by linking the blockade of dopamine D2 receptors with reductions in parkinsonian symptoms, anhedonia and increased neurogenesis (Aringhieri et al., 2018; Chikama et al., 2017). The effect of diet on hippocampal neurogenesis has been relatively unexplored.

Rats given 1 month access to fructose showed almost a 40% reduction in the number of BrdU / NeuN-immunopositive cells (mature neurons) in the dentate gyrus of the hippocampus (Van der Borght et al., 2011). Alternatively, rats consuming glucose had a similar number of BrdU / NeuN-immunopositive cells as the water controls (Van der Borght et al., 2011). Density of immature neurons labelled with PSA-NCAM was decreased following sucrose, fructose and glucose (Van der Borght et al., 2011). Sugar consumption increased the number of proliferating cells positive for Ki67 (a marker of cell proliferation (Scholzen and Gerdes, 2000)) in all cases and rats offered sucrose or fructose showed more cell death in the dentate gyrus of the hippocampus (Van der Borght et al., 2011).

Rats given a high fructose corn syrup solution throughout adolescence were tested for hippocampal-dependent contextual memory. Impairments in memory function were found in later in life, suggesting that sugar consumption early in life may have long-term negative effects on memory function (Noble et al., 2017). High fructose consumption is also linked to insulin resistance (Stranahan et al., 2008) with reduced hippocampal neurogenesis (Van der Borght et al., 2011). Hence it appears convincing that a high-sugar diet negatively affects adult neurogenesis which may contribute to the anxiety and depressive like behaviour demonstrated by animals after withdrawal. Decreased neurogenesis in the hippocampus has further been implicated in memory dysfunction and cognitive impairment disorders such as Alzheimer’s disease (Price et al., 2001) and the learning and memory processes involved in appetitive control (Davidson et al., 2009; Lathe, 2001). The hippocampus may hold further insight into the cognitive processes by which food becomes entwined with motivation and new rewards.

5.4. Fear

The functional neuroanatomy of fear encompasses the AMG, which stimulates the HPA axis and the hippocampus, which suppresses activity of the axis (Martin et al., 2009). Hyper activation of the AMG was recorded in response to viewing faces construed as fearful (Bryant et al., 2008a) and has been implicated in post-traumatic stress disorder (PTSD) (Cortese and Phan, 2005). The PFC is required for the extinction of fear memories (Giustino and Maren, 2015), and in line with these findings changes in volume of the anterior cingulate cortex have been implicated in a reduced ability to extinguish fearful memories (Bryant et al., 2008b). When fear memories become dysfunctional it is thought that the intrusive, recurring thoughts result from an inability of the cognitive control circuit to repress the negative effect circuit (Morey et al., 2008). In patients with PTSD their information processing may be overpowered by the hyper activation of the AMG when exposed to threat related stimuli (Falconer et al., 2008).

At the molecular level corticotrophin-releasing factor from the paraventricular nucleus of the hypothalamus initially activates the HPA axis in response to threatening stimuli (Martin et al., 2009). Low levels of corticosterone, which acts through binding to mineralocorticoid and glucocorticoid receptors, as well as enhanced negative feedback of the HPA axis are associated with pathological fear (Han et al., 2014). Rats that underwent a single prolonged stress paradigm (a reliable animal model of PTSD) have shown a down-regulation of both these receptors in the hippocampus, enhanced fear and altered neuronal morphology of AMG neurons (Han et al., 2014). Glucocorticoid receptors and dopamine receptors have been implicated in processes within the PFC that drive extinction memory learning. Infusions of corticosterone or a glucocorticoid receptor antagonist (RU38486) into the infralimbic cortex and pretreatment with sulpiride (a dopamine D2 receptor antagonist) attenuated fear expression‎ suggesting enhanced fear extinction (Dadkhah et al., 2018). As glucocorticoid receptors are also required for glucose homeostasis and play a role in the development of hyperinsulinemia and obesity (Majer-Lobodzinska and Adamiec-Mroczek, 2017) they are suspected as key players in neuroadaptations to the AMG, hippocampus and PFC that occur after long-term sucrose consumption.

Juxtaposing the combined psychological and physical effects high sugar diets have on pathological fear has, until recently, been an untouched topic of research. Although no studies have been conducted on sugar consumption alone with regards to pathological fear, some research has begun to explore the combined effect of high-fat/sugar or high sugar and high carbohydrate diets (Santos et al., 2018; Baker and Reichelt, 2016; Williams-Spooner et al., 2017; Reichelt et al., 2015). These studies are beginning to delineate the impact high caloric foods can have on the intensity and duration of pathological fear; however research into the effect of sugar alone is required to delineate the mechanisms involved. More recent research has examined contextual fear following a standard contextual fear conditioning protocol which was contradictory to this finding (Santos et al., 2018). Specifically, high-sugar/carbohydrate diets significantly enhance fear-related freezing to context (Santos et al., 2018). Whilst both examine contextual fear memory formation, different fear conditioning protocols were used (trace fear conditioning versus contextual fear conditioning). Indeed hippocampal function in contextual fear memory has been noted as disparate, depending on the fear conditioning protocol (see recent review by (Chaaya et al., 2018)). Nevertheless, these studies highlight how diet can influence hippocampal-dependent memory, as well as hippocampal function (Santos et al., 2018; Reichelt et al., 2015).

Fear extinction is a protocol whereby, following excessive exposure to the previously neutral fear conditioning CS (e.g. the context in contextual fear conditioning, or a tone in auditory fear conditioning), a reduction of fear-related freezing is seen (Maren, 2011). This reduction in fear is similar to basic theories surrounding exposure therapy, whereby excessive exposure to fear-inducing stimuli (via mental imagery, for example) results in a reduction in fear-related symptoms (Maren, 2011; Milad and Quirk, 2012).

Investigation into the effect of PTSD symptoms on the consumption of highly palatable ‘fast’ food and sweetened beverages was conducted to see if there was a correlation with emotional eating behaviours and body mass index (Hirth et al., 2011). The answers to questions regarding frequency of consuming fast food were collected from 3154 females and analyzed using regression analyses (Hirth et al., 2011). To determine unhealthy eating habits, participants were asked if they used diet pills, laxative, diuretics, skipped meals or vomited after eating. Findings suggested that PTSD symptoms increased the frequency of consumption of high caloric food and sodas, as well as contributing to unhealthy eating habits but did not increase overall body mass index (Hirth et al., 2011). Cumulatively, data from these various studies show a strong link between PTSD, fear memory and caloric foods suggesting a possible correlation with high sugar consumption that requires further investigation.

6. Sucrose consumption investigated

Negative states of emotion driven by substances of abuse are often shown by examining anxiety-like behaviours, decreased pain tolerance, or an increase in the point at which reward becomes sufficient to produce a stimulating effect and memory deficits (Koob, 2018). The complexities of the systems involved, the interconnectedness of the systems and the myriad of protocols used to study sucrose dependence could potentially lead to a wide variety of outcomes and yet overall investigations into high-sugar consumption in rats, mice and humans result in similar outcomes with regards to memory, stress and emotion (see Table 3.)

Table 3. Published reports on the effect of sucrose and sweetener consumption on cognition, emotion and stress.

Sucrose or sweetener consumption in ratsAuthors and yearSubjectsTasksBrain region involvedFindings2h/day 10% sucrose for 28 daysRat chow supplemented with meat pies, cakes and biscuits for 10 weeksPowdered diet: 7.9% sucrose for 52 weeks35% sucrose solution for 9 weeks32% sucrose solution for eight weeks10% fructose in drinking water for 7 months30% sucrose solution, 30 m, twice daily, 2-4wkOr 0.1% saccharin10-34% sucrose solutionFree access to 5% sucrose or water for 16d1% sugar solution (high fructose corn syrup) throughout the adolescent phase of development (post-natal day 26-56).

Xu, T.J. and Reichelt, A.C., 2017 (Xu and Reichelt, 2018)	3 week-old male Sprague-Dawley rats (n = 8)	EPM, open-field, NPR and short- and long-term NOR (Organization, W.H., 2017)	hippocampus, basolateral amygdala	increased anxiety-like behaviours
Reichelt, A., et al., 2015 (Reichelt et al., 2015)	6 week-old male Sprague–Dawley rats (n = 16)	trace fear conditioning	hippocampus	strengthened visual fears and attenuated contextual fears
Chepulis, L.M., et al., 2009 (Chepulis et al., 2009)	2-month old Sprague Dawley rats (n = 10-14)	1EPM, NOR Y maze, C	NA	reduced spatial memory and increased anxiety
Lemos, C., et al., 2016 (Lemos et al., 2016)	12-weeks old, male Wistar rats	open field, object displacement, NOR, forced swimming test	Hippocampus	decreased memory performance and increased helpless behaviour
Jurdak, N. and Kanarek, R.B., 2008 (Jurdak and Kanarek, 2009)	6 week old male Long-Evans rats (n = 10)	1NOR	NA	decreased cognitive performance on object recognition
Sangüesa, G., et al., 2017 (Sangüesa et al., 2018)	female Sprague-Dawley rats	Morris water maze, (Organization, W.H., 2017) NOR	frontal cortex and hippocampus	reduced performance in the NOR test
Ulrich-Lai, Y.M., et al., 2010 (Ulrich-Lai et al., 2010)	Adult male Long-Evans rats	Restraint stress, social interaction test, open field, EPM	hypothalamic paraventricular nucleus, BLA	- reduced ACTH by sucrose and saccharin - reduced corticosterone only after sucrose sucrose - reduced restraint-induced tachycardia and behavioural anxiety
Wilmouth, C.E., and Spear, L.P., 2009. (Wilmouth and Spear, 2009)	Adolescent male Sprague–Dawley rats	Taste reactivity and voluntary consumption	NA	greater positive taste reactivity and reduced negative responding
Vendruscolo, L.F., et al., (Vendruscolo et al., 2010)	Male Wistar rats	fixed- and progressive-ratio self administration of: - saccharin - maltodex- trin - cocaine	NA	Sugar overconsumption during adolescence, reduced motivation for saccharin and maltodextrin
Noble, E.E., et al., 2019 (Noble et al., 2017)	Male rats	novel object in context	Hippocampus	- impairments in hippocampal-dependent memory function later in life - NOIC performance impaired at PN 175

Sucrose and sweetener consumption in miceHigh-carbohydrate diet (45% condensed milk, 10% sugar and 45% chow) for 8 weeks10% sucrose, 3 I2BC for four weeks, withdrawal animals received only water for one week after the four weeksAcesulfame potassium and low carbohydrate diet for 4 weeks

Authors and year	Subjects	Tasks	Brain region involved	Findings
Santos, C.J., et al., 2018 (Santos et al., 2018)	5 – 7 week old, male BALB/c mice	restraint stress, 1EPM, contextual fear conditioning, tail suspension test	NA	increased anxiety-like and depressive-like behaviour and aversive memory
Kim, S., et al., 2017 (Kim et al., 2018)	C57BL/6 mice	tail suspension test, (Organization, W.H., 2017) EPM, sucrose preference test	nucleus accumbens	withdrawal after sucrose overeating induces depression and anxiety-like behaviour.
Ibi, D., 2018 (Ibi et al., 2018)	Male ddY strain mice (7–9 weeks old)	Y-maze and NOR, glucose levels	frontal cortex	decrease in short-term and object cognitive memories decreased glucose levels

Sucrose and sweetener consumption in humansAuthors and yearSubjectsTasksBrain region involvedFindingsSugar, sweetened beverages, and fruit intake was assessed on a servings per day basisHigh-sugar content milkshakesSucrose- or aspartame-sweetened beverage consumption three times per day for 2 weeks

Cohen, J.F.W., et al., 2018 (Cohen et al., 2018)	mother and child pairings, during pregnancy and childhood mean ages 3.3 yrs, 7.7 yrs (n = 1,234)	2 PPVT-III KBIT-II KBIT-II WRAVMA WRAML and HOME-SF	NA	adverse impact on child memory and learning
Shearrer, G.E., et al., 2018 (Shearrer et al., 2018b)	133 adolescents	food picture exposure during fMRI	temporal gyrus, operculu, juxtapositional lobule, thalamus, caudate	increased signal in the reward learning, processing and motivation regions of the brain
Tyron, M.S., et al., 2015 (Tryon et al., 2015)	Nineteen women (age 18–40 y)	Salivary cortisol, Montreal Imaging Stress Task	hippocampus	sucrose consumption resulted in: - higher activity in the left hippocampus - reduced stress-induced cortisol - lower reactivity to naltrexone, lower nausea, and a trend toward lower cortisol

1 EPM: elevated plus maze; NPR: novel place recognition; NOR: novel object recognition. 2 PPVT-III: Peabody Picture Vocabulary Test, third edition for maternal testing; WRAVMA: Wide Range Assessment of Visual Motor Abilities, KBIT-II: Kaufman Brief Intelligence Test, second edition, WRAML: Wide Range Assessment of Memory and Learning for childhood testing; HOME-SF: Environment short form test to eval‎uate home environment for cognitive stimulation and emotional support. 3 I2BC: intermittent 2 bottle choice. 4 LTP: long-term potentiation.

7. Therapeutics for obesity, derived from studies of sucrose consumption7.1. Pharmacological approaches

Sucrose is shown to be addictive in rodents (Avena et al., 2008; Colantuoni et al., 2002; Rada et al., 2005) and the studies listed in Table 3 show correlations between sucrose consumption and emotional disturbances. The NK1- (Neurokinin-1) receptor system is involved in the reinforcement mechanism that motivates the desire to have stimuli that no longer create pleasure, a characteristic of addiction (Sandweiss and Vanderah, 2015), and is implicated in both anxiety and depression (Mantyh, 2002). Using the intermittent access model with a 5% sucrose solution, our laboratory showed the NK1-receptor antagonist ezlopitant (which possesses anxiolytic effects) was able to attenuate and inhibit sucrose intake in Long Evans rats (Steensland et al., 2010). This finding suggests the NK1- receptor system to be a potential target for sugar-related obesity therapeutics (Steensland et al., 2010). Other potential options include antidiabetic drugs such as sodium-glucose cotransporter 2 (SGLT2) inhibitors, which impede renal glucose reabsorption and are generally considered effective, though lose efficacy when taken long term (Gillies et al., 2007). Indeed, a 4 week oral administration of ipragliflozin, (a SGLT2 inhibitor) with antidiabetic effects on type 2 diabetes, decreases the caloric balance and improves symptoms of diabetes, including obesity in mice fed on 20% glucose or sucrose solution (Tahara et al., 2018; Chao and Henry, 2010)

A recent review detailing the current management for type 2 diabetes suggested a step wise approach; including diet and exercise the patient would take metformin (which lowers high blood glucose levels by decreasing the production of hepatic glucose, increasing insulin sensitivity and lowering intestinal glucose absorption (Rena et al., 2017)), a glucagon-like peptide 1 receptor agonist (to increase insulin output from the pancreas (Drucker and Nauck, 2006)) or a SCLT2 inhibitor in addition to one of the approved weight-loss drugs (to control appetite or the absorption of calories consumed (Padwal and Majumdar, 2007)) (Burguera et al., 2017). Table 4 lists doses and mechanisms of action of therapeutics used in trials to alter sucrose consumption in an effort to discover alternate treatments for diabetes and obesity.

Table 4. Therapeutics used in sugar consumption trials.

Author, yearDrug / DoseMechanism of actionSubjectsFindingsRichard, D., et al., 2000 (Richard et al., 2000)Beczkowska, I.W., 1992 (Beczkowska et al., 1992)Steensland, P., et al., 2010 (Steensland et al., 2010)Shariff, M., 2016 (Shariff et al., 2016)Tahara, A., et al., 2018 (Tahara et al., 2018)Muscat, R. and Willner, P., 1989 (Muscat and Willner, 1989)Lof, E., et al., 2010 (Löf et al., 2010)Patkar, O., et al., 2017 (Patkar et al., 2017)Lin, Z., et al., 2013 (Lin et al., 2013)Kurhe, Y., et al., 2014 (Kurhe et al., 2014)Badia‐Elder, N.E., et al., 2003 (Badia‐Elder et al., 2003)Pandit, R., et al., 2015 (Pandit et al., 2015)

Topiramate (30 mg / kg)	Blocks voltage gated sodium and calcium channels, ↓ glutamate, ↑ GABA (Anticonvulsant)	Rats	↓ Weight ↓ sucrose intake
Naltrexone Naloxone	Kappa and Mu2 (opioid) receptor antagonists (Treat drug and alcohol dependence	Rats	↓Sucrose intake
Ezlopitant (2,5,10 mg / kg)	NK1 receptor antagonist (Anxiolytic and antiemetic)	Rats	↓ obesity
Varenicline (0.3,1,2 mg / kg)	nAChR partial agonist (Smoking cessation)	Rats	↓ sucrose intake
Ipragliflozin (0.1 – 3 mg / kg)	SGLT2 inhibitor (Antidiabetic)	Mice	↓ sucrose intake
Sulpiride (20,40 mg / kg)	D2 and D3 receptor elective antagonist (Antipsychotic)	Rats	↓ desire for sucrose
Methyllycaconitine (0.3,1 mg/kg)	α7 nAChRs selective antagonist	Rats	↓ sucrose preference
Buspirone (1,2.5,5 mg / kg)	5-HT1A/1B partial agonist (Anxiolytic)	Mice	↓ sucrose consumption
Curcumin (40 mg / kg)	Upregulation of PPAR-γ activation (Herbal supplement)	Rats	↓ sucrose intake
Ondansetron (1 mg / kg)	Serotonin receptor (5-HT3) antagonist (Antiemetic)	Mice	↑ Sucrose consumption
NPY (5 μg / 5 μl) (10 μg / 10 μl)	Inhibits GAD67 expression‎ (Vasoconstrictor)	Rats	↑ sucrose intake
AgRP (0.66 nmol) (1 nmol)	MC 3/4 receptor inverse agonist (Decreases metabolism)	Rats	↑ motivation for sucrose

7.2. Lifestyle interventions

Diet and exercise are highly recommended approaches to treatment for obesity (Bray and Bouchard, 2014). A 2007 systematic review and meta-analysis that eval‎uated lifestyle interventions for patients with impaired glucose tolerance looked at 21 randomized controlled trials designed to delay or prevent the onset of type 2 diabetes (Gillies et al., 2007). Data revealed evidence in support of interventions such as diet and exercise being at least as effective as oral diabetes drugs, a Chinese herbal remedy (jiangtang bushen) and the anti-obesity drug orlistat (Gillies et al., 2007). The dilemma with lifestyle interventions is that regain of the weight lost is common after a period of time and often results in weight gain greater than that originally lost (Barte et al., 2010). Strategies that proved effective in maintaining weight loss included a consistent regular meal pattern that included breakfast, a high expectation to succeed, a good support system and behavioural self-monitoring (Barte et al., 2010). Mindfulness meditation is another intervention studied as a way to regulate emotional eating. A self-reported survey was conducted to examine whether mindfulness practice could alter emotional eating (Levoy et al., 2017). Surveys conducted before and after the intervention revealed lower emotional eating scores after the meditation, suggesting a possible role for mindfulness as a treatment for emotional eating (Levoy et al., 2017). Mindfulness studies do not show a direct effect on weight loss itself, but appear to be quite successful in lessening the addictive-like behaviours pursuant to relapse of weight gain (Moor et al., 2013; Goldbacher et al., 2016; Keesman et al., 2017).

7.3. Digital technology

Decreasing emotional eating has been proposed as a potential mechanism for the long-term maintenance of weight loss (Barte et al., 2010). Digital technology may be our best hope of achieving this goal. Interaction with online environments provide the social support often lacking during maintenance of weight loss (Hwang et al., 2010). Apps for exercise regimes, food tracking, meditation and positive thinking all provide a support system to the user and promote networking with others in similar situations. Support networks are useful tools as increases in body weight decrease an individual’s ability to make informed decisions regarding highly palatable food, resist temptation and regulate their emotions (Yeomans, 2017). These challenges were supported through a study of 17 obese women and their physiological reactions to verbal food cues which found that food and beverage preferences affected physiologic responses as well as cognition and attention (Fioravanti et al., 2004).

Wearable technology such as Fitbits (activity trackers), that interact directly with the individual, provide information about the individual and promote long term maintenance and support models of healthcare where the patient takes an active role in their own wellness (Handel, 2011; Sama et al., 2014). Wearable technologies appear to be extremely effective in assisting weight loss in patients with serious mental illness (Aschbrenner et al., 2016). Thirteen obese individuals diagnosed with a serious mental illness (e.g. schizophrenia and major depressive disorder) participated in 24 weeks of behavioural weight loss intervention encompassing group sessions for weight management and exercise, mobile health technology and social media for motivation, peer support and a self-monitoring tool (Aschbrenner et al., 2016). At the end of 6 months of interventions, 45% of participants had reduced their weight (below their baseline) and increased fitness (as measured through walking distance) (Aschbrenner et al., 2016).

8. Discussion

The data summarized in this article suggests overconsumption of sugar can lead to brain adaptations involving many different neural systems, molecular substrates and subsequent changes in behaviour. The ease of availability and cost effectiveness of high caloric, sweetened food and beverages appear to be a contributing factor in the world wide increase in obesity (Sinha, 2018; He et al., 2018). Paradigms designed to investigate emotional eating show increases in weight gain due to higher caloric intake, nevertheless examination of the correlation between environmental and social inputs, individual thought processes and behaviour that maintains emotional eating is yet to be defined. It seems feasible to suggest the pleasurable sensations brought about through the consumption of sugar may provide a self-medicating method to deal with daily stresses (Fortuna, 2010; Brewerton, 2011). Common hedonistic mechanisms play a role in both obesity and addiction to drugs of abuse (Johnson and Kenny, 2010). While the root cause of obesity remains elusive, the estimated global annual healthcare cost of treating illnesses related to obesity may reach US $1.2 trillion per year by 2025.

Every day new digital applications are being developed to assist our pursuit of health and happiness. In the future it may be possible to collect personal information about the consumption of high-sugar foods and beverages in conjunction with emotions felt on a daily basis. The possibilities of personal wellness monitoring, implantable in vivo monitoring and drug delivery devices will also require further robust study before they prove to be effective treatments for obesity and food addiction. It would be interesting to investigate a possible correlation between sugar consumption, emotions and body mass index (BMI > 30 denotes obesity). It would also be significant to know if such applications combining game rewards might assist with the childhood obesity epidemic as current predictions state that globally, 2.7 billion adults will be overweight and/or obese by 2025 (Haby et al., 2012).

Functional brain imaging studies substantiate the higher preference and increased emotional activation that occurs in response to images and verbal cues related to high sugar content foods and beverages, making it more difficult for overweight individuals to resist eating unhealthy food (Batterink et al., 2010; Pursey et al., 2014). Optogenetic and chemogenetic studies may assist in defining the combination of neural pathways and substrates involved. Other tools designed to unravel the complexities between neuronal organisation and behaviour include methods to quantify cellular populations that are functional for particular behaviours (Bergstrom et al., 2011). Using analytic methods (including micro-binning and density mapping) to accurately compare functional neural network activity it may be possible to produce microanatomical topography of molecular activity resultant from the influence of long term sugar consumption on a variety of stress and anxiety related behaviours (Jacques et al., 2018). This would assist in demonstrating pathophysiological neuroadaptive changes and perhaps lead to the development of enhanced pharmacotherapeutic and technological strategies to assist the reduction of excess sugar consumption.

In conclusion, the ease of access to sugar rich diets today is an environmental contributor to obesity, but it may be sugars ability to generate a superior neurological reward signal which overrides self-control mechanisms and leads to obesity (Lenoir et al., 2007; Yeomans, 2017). Obesity and long term sugar consumption both result in low basal levels of dopamine, particularly in the NAc, which may be the mechanism which induces the desire to overeat in the hope to restore homeostatic dopamine levels and avoid mild depression. Opioids, which induce feeding through an abundance of brain regions, may be responsible for cue-induced relapse into overeating behaviours, and binging on sugar postpones the release of acetylcholine required to signal satiety. Each of these neuroadaptations implicates sucrose consumptions ability to alter the way we perceive and process our emotions and consequential behaviour. Perhaps a greater understanding of the neural mechanisms of impulsivity and overeating are required to assist in the development of improved obesity treatments.

It has been estimated that by 2020, 1.5 million people will die each year by suicide, with 15 to 30 million attempting it (Weissman et al., 1999). Children suffering anxiety disorders are twice as likely to attempt suicide, while those suffering major depressive illnesses show a 3 fold chance at attempt (Weissman et al., 1999).In support of these statistics, over 300 million people were reported as suffering depression in 2016/17, with 264 million people reported to be suffering from anxiety disorders (Organization, 2017). This review has examined how negative emotion can exacerbate sugar overconsumption, and vice versa. If negative emotions are so preval‎ent in our children, and sugar intake so common, its consumption may be considered a threat to the emotional stability of our race (see reviews on mental health in children and adolescents (Wehry et al., 2015; Beesdo et al., 2009; Pine et al., 2002)). More importantly, reduction of sugar overconsumption may be capable of significantly reducing the preval‎ence of negative emotion in a vast number of individuals around the world.

Funding

This research was funded by NHMRC project grant (GNT1146417) and Lady Cilento Children’s Hospital Research Foundation.

Conflicts of interest

The founding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results. S.E.B is currently examining the role of emotions and high-fat/sugar containing foods using digital technology platform.

Acknowledgments

N/A.

References

1. Acevedo et al., 2012
2. Adcock et al., 2006
3. Adinoff, 2004
4. Ahima and Antwi, 2008
5. Ahmed et al., 2013
6. Alsiö et al., 2012
7. Anderson and Petrovich, 2018
8. Anderson et al., 2018
9. Aosaki et al., 2010
10. Aringhieri et al., 2018
11. Aschbrenner et al., 2016
12. Asnaghi et al., 2003
13. Association, 2019
14. Atkins et al., 1998
15. Avena and Hoebel, 2003
16. Avena et al., 2005
17. Avena et al., 2006
18. Avena et al., 2008
19. Avena et al., 2009
20. Azimova et al., 2015
21. Badia‐Elder et al., 2003
22. Bailey et al., 2005
23. Baker and Reichelt, 2016
24. Bakke et al., 2018
25. Barnes and Joyner, 2012
26. Baron et al., 2018
27. Barrientos et al., 2018
28. Barte et al., 2010
29. Batterink et al., 2010
30. Beczkowska et al., 1992
31. Beesdo et al., 2009
32. Beilharz et al., 2014
33. Belin and Dalley, 2012
34. Bello et al., 2002
35. Bello et al., 2011
36. Belzeaux et al., 2018
37. Benarroch, 2012
38. Benoit et al., 2008
39. Benowitz, 1988
40. Benton, 2010
41. Bergstrom et al., 2011
42. Berridge, 1996
43. Berridge et al., 2010
44. Berthoud et al., 2017
45. Bianchi et al., 2018
46. Bienkowski et al., 2004
47. Bito‐Onon et al., 2011
48. BjÖrntorp, 2000
49. Bocarsly, 2018
50. Bordet et al., 1997
51. Bouchet et al., 2018
52. Bray and Bouchard, 2014
53. Brewerton, 2011
54. Brown and Rother, 2012
55. Bryant et al., 2008a
56. Bryant et al., 2008b
57. Burguera et al., 2017
58. Burnett et al., 2016
59. Cabanac and Johnson, 1983
60. Calipari et al., 2014
61. Cao et al., 2007
62. Capello and Markus, 2014
63. Carr and Kalivas, 2006
64. Cha et al., 2008
65. Chaaya et al., 2018
66. Chao and Henry, 2010
67. Chepulis et al., 2009
68. Chikama et al., 2017
69. Clifton, 2017
70. Cocores and Gold, 2008
71. Cohen et al., 2018
72. Colantuoni et al., 2001
73. Colantuoni et al., 2002
74. Comings et al., 2001
75. Cools et al., 2003
76. Coplan et al., 2018
77. Corsica and Pelchat, 2010
78. Corsini, 1999
79. Cortese and Phan, 2005
80. Covington and Miczek, 2001
81. Criscitelli and Avena, 2017
82. D’Anci et al., 1996
83. D’Argenio et al., 2009
84. Dadkhah et al., 2018
85. Dallman et al., 2003
86. Davidson et al., 2009
87. de Jong, 2015
88. De Jong et al., 2016
89. De Nijs et al., 2018
90. Denis et al., 2015
91. Derman and Ferrario, 2017
92. Di Chiara and Imperato, 1988
93. Díaz-Aguila et al., 2016
94. Domingos et al., 2013
95. Drevets, 1998
96. Drucker and Nauck, 2006
97. Egeland et al., 2015
98. Bonin et al., 2018
99. Falconer et al., 2008
100. Feduccia et al., 2014
101. Fidler et al., 1993
102. Fioravanti et al., 2004
103. Fiorentini and Müller, 1975
104. Flagel et al., 2011
105. Floresco et al., 2018
106. Foley et al., 2006
107. Foo and Mason, 2005
108. Fortuna, 2010
109. Gangwisch et al., 2015
110. García-Cáceres et al., 2016
111. Gaysinskaya et al., 2011
112. Gearhardt et al., 2009
113. Geiger et al., 2008
114. Gerdeman et al., 2003
115. Gillies et al., 2007
116. Giustino and Maren, 2015
117. Goldbacher et al., 2016
118. Gonzalez and Betts, 2018
119. Gosnell, 1987
120. Gosnell, 2005
121. Graham et al., 1997
122. Grant et al., 1980
123. Gueye et al., 2018
124. Guillin et al., 2001
125. Gulpinar and Yegen, 2004
126. Gunstad et al., 2006
127. Guo et al., 2014
128. Gurevich et al., 1997
129. Haby et al., 2012
130. Hajnal et al., 2004
131. Halford et al., 2010
132. Han et al., 2014
133. Handel, 2011
134. Hariri et al., 2005
135. Hayward et al., 2006
136. He et al., 2018
137. Hebebrand et al., 2014
138. Heim and Nemeroff, 2001
139. Henry et al., 1994
140. Herz, 1997
141. Heyser et al., 1997
142. Hirth et al., 2011
143. Hoebel et al., 2007
144. Hoebel et al., 2009
145. Holgate et al., 2017
146. Holstege et al., 2003
147. Hsu et al., 2015
148. Huang et al., 2018
149. Huber et al., 2018
150. Hwang et al., 2010
151. Hwang et al., 2017
152. Ibi et al., 2018
153. Iemolo et al., 2012
154. Itzhak and Martin, 1999
155. Jacka et al., 2011
156. Jacobson and Sapolsky, 1991
157. Jacques et al., 2018
158. Jastreboff et al., 2016
159. Jellinek, 1952
160. Jennings et al., 2013
161. Jennings et al., 2015
162. Jing et al., 2018
163. Johnson and Kenny, 2010
164. Joly-Amado et al., 2014
165. Jurdak and Kanarek, 2009
166. Kanarek et al., 1991
167. Kanoski and Davidson, 2011
168. Karagiannides et al., 2008
169. Kastin, 2013
170. Kavaliers and Hirst, 1987
171. Keesman et al., 2017
172. Kelley et al., 2005
173. Kendig, 2014
174. Kenny, 2011
175. Kikkawa et al., 1987
176. Kim et al., 2018
177. Kinzig et al., 2008
178. Kirshenbaum et al., 2008
179. Klenowski et al., 2016
180. Knolle et al., 2018
181. Knüppel et al., 2017
182. Koob, 2009a
183. Koob, 2009b
184. Koob, 2018
185. Koob and Le Moal, 2001
186. Koolschijn et al., 2009
187. Kurhe et al., 2014
188. Kyriazis et al., 2012
189. Lam and Jadavji, 2018
190. Lathe, 2001
191. Lau and Herzog, 2014
192. Laughlin, 2014
193. Leão et al., 2012
194. Lee and Messing, 2008
195. Legastelois et al., 2014
196. Leigh and Morris, 2016
197. Lemos et al., 2016
198. Lenoir et al., 2007
199. Levine and Billington, 2004
200. Levoy et al., 2017
201. Li et al., 2004
202. Lien et al., 2006
203. Lin et al., 2013
204. Lindgren et al., 2018
205. Liu, 2018
206. Liu et al., 2014
207. Löf et al., 2010
208. Lowe and Butryn, 2007
209. Lu et al., 2004
210. Lumeng et al., 2013
211. Luppino et al., 2010
212. Lustig et al., 2012
213. Majer-Lobodzinska and Adamiec-Mroczek, 2017
214. Makarem et al., 2018
215. Maldonado et al., 2006
216. Malik et al., 2013
217. Mangabeira et al., 2015
218. Maniam et al., 2015
219. Mantyh, 2002
220. Maren, 2011
221. Mark et al., 2006
222. Martin et al., 2009
223. Mathes et al., 2013
224. Meye and Adan, 2014
225. Milad and Quirk, 2012
226. Miller et al., 2002
227. Ming and Song, 2011
228. Misra et al., 2008
229. Mitra et al., 2010
230. Moghaddam, 2018
231. Molnar et al., 2018
232. Molteni et al., 2002
233. Moor et al., 2013
234. Moore et al., 2018
235. Moorman and Aston-Jones, 2014
236. Morey et al., 2008
237. Moscarello and Maren, 2018
238. Moss, 2013
239. Murray et al., 2016
240. Muscat and Willner, 1989
241. Nakagawa et al., 2006
242. Nakamura et al., 1987
243. Naneix et al., 2016
244. Naneix et al., 2018
245. Nestler and Aghajanian, 1997
246. Nieh et al., 2015
247. Noble et al., 2017
248. Noll et al., 2007
249. Nona and Nobrega, 2018
250. Novelle and Diéguez, 2018
251. Nummenmaa and Tuominen, 2017
252. Olds and Olds, 1963
253. Organization, 2017
254. Organization, W.H., 2016
255. Organization, W.H., 2017
256. Oswald et al., 2011
257. Otmakhova et al., 2013
258. Padilla et al., 2016
259. Padwal and Majumdar, 2007
260. Page et al., 2013
261. Palmiter, 2007
262. Pan et al., 2011
263. Pandit et al., 2015
264. Parylak et al., 2012
265. Patkar et al., 2017
266. Patriquin and Mathew, 2017
267. Pecina and Berridge, 1995
268. Peet, 2004
269. Peng et al., 2011
270. Petrovich and Gallagher, 2003
271. Petrovich et al., 2012
272. Pierce and Kalivas, 1995
273. Pijl, 2003
274. Pine et al., 2002
275. Posadas et al., 2013
276. Powers and Howley, 2007
277. Pratt and Kelley, 2004
278. Price et al., 2001
279. Pursey et al., 2014
280. Qualls-Creekmore et al., 2017
281. Quiñones et al., 2018
282. Quirk et al., 2013
283. Rada and Hoebel, 2005
284. Rada et al., 2001
285. Rada et al., 2005
286. Rahman et al., 2015
287. Rajmohan and Mohandas, 2007
288. Ramasubramanian et al., 2013
289. Reichelt et al., 2015
290. Reker et al., 2018
291. Rena et al., 2017
292. Reyes et al., 2015
293. Richard et al., 2000
294. Riga et al., 2015
295. Ritter et al., 1981
296. Robinson, 1988
297. Robinson and Berridge, 1993
298. Robinson and Berridge, 2001
299. Robinson and Berridge, 2008
300. Robinson et al., 2014
301. Rosas et al., 2018
302. Ruegg et al., 1997
303. Salamone and Correa, 2012
304. Sama et al., 2014
305. Sánchez-Villegas et al., 2012
306. Sandweiss and Vanderah, 2015
307. Sangüesa et al., 2018
308. Santos et al., 2018
309. Savage et al., 2014
310. Schneider, 1989
311. Scholzen and Gerdes, 2000
312. Schwartzbaum, 1988
313. See et al., 2003
314. Shariff et al., 2016
315. Shariff et al., 2017
316. Sharpe et al., 2016
317. Shearrer et al., 2016
318. Shearrer et al., 2018a
319. Shearrer et al., 2018b
320. Shi et al., 2010
321. Simon et al., 2006
322. Sinha, 2008
323. Sinha, 2018
324. Smith, 1960
325. Smith and Vale, 2006
326. Sohn et al., 2013
327. Soto et al., 2015
328. Southgate, 1995
329. Souza et al., 2007
330. Spangler et al., 2004
331. Staley and Mash, 1996
332. Stanley et al., 1988
333. Steensland et al., 2010
334. Stranahan et al., 2008
335. Stroebe et al., 2008
336. Stunkard et al., 2003
337. Sweeney and Yang, 2017
338. Tabuchi et al., 2002
339. Tahara et al., 2018
340. Teff et al., 2004
341. Tellez et al., 2016
342. Tryon et al., 2015
343. Tye et al., 2008
344. Ulrich et al., 2018
345. Ulrich-Lai et al., 2010
346. Ulrich-Lai et al., 2011
347. Ursano et al., 2009
348. Valencia-Torres et al., 2017
349. Van der Borght et al., 2011
350. van Dooren et al., 2013
351. Van Strien et al., 1986
352. Vendruscolo et al., 2010
353. Verty et al., 2004
354. Vigano et al., 2003
355. Volkow and Fowler, 2000
356. Volkow and Wise, 2005
357. Volkow et al., 2002
358. Volkow et al., 2004
359. Vollmayr et al., 2004
360. Wade et al., 2000
361. Wallace et al., 2008
362. Wang et al., 2001
363. Wang et al., 2006
364. Weeks, 1962
365. Wehry et al., 2015
366. Weihrauch and Diehl, 2004
367. Weissman et al., 1999
368. Welberg and Seckl, 2001
369. Westover and Marangell, 2002
370. Westwater et al., 2016
371. Williams-Spooner et al., 2017
372. Wilmouth and Spear, 2009
373. Wilson, 2010
374. Wise, 1996
375. Wise, 2004
376. Wise, 2006
377. Wise et al., 1995
378. Wolf et al., 2008
379. Woods and Leibowitz, 1985
380. Wu, 2009
381. Wyvell and Berridge, 2001
382. Xiong et al., 2018
383. Xu and Reichelt, 2018
384. Yamanaka et al., 2000
385. Yang et al., 2018
386. Yeomans, 2017
387. Yu and Kim, 2012
388. Zehra et al., 2018
389. Zhang et al., 2018a
390. Zhang et al., 2018b
391. Ziauddeen et al., 2012

Cited by (80)

•  
•  
•  
•  
•  
•