Role of Odor Novelty on Olfactory Issues in Autism Spectrum Disorder Authors: Zoe A Scheier, Kassandra L Sturm, John A Colavecchio, Apekchha Pradhan, Gonzalo H Otazu Published: 2024 ABSTRACT ABSTRACTSensory processing abnormalities are a hallmark of autism spectrum disorder (ASD) and are included in its diagnostic criteria. Among these challenges, food neophobia has garnered attention due to its prevalence and potential impact on nutritional intake and health outcomes. This review describes the correlation between novel odor perception and feeding difficulties within the context of ASD. Moreover, this review underscores the role of odor processing in shaping feeding behaviors within the ASD population. It examines the psychophysics of odor perception in individuals with ASD and evaluates the behavioral and neurophysiological assessments conducted using novel odor stimuli in mouse models relevant to autism and wild‐type mice. Additionally, we explore the mechanism on how odor novelty affects neuronal circuitry, shedding light on potential underlying mechanisms for the effect of odor novelty on ASD.Keywords: autism, Cntnap2, food neophobia, fragile X, MEAL plan, mitral cells, neural variability, odor habituation, olfactory bulb, Shank3 Introduction This review examines the impact of odors on feeding problems in autism spectrum disorder (ASD), exploring current strategies used to introduce new food items. We cover the psychophysics of odor processing in ASD, reviewing the behavioral and neurophysiological assessments done with odors in autism mouse models. Additionally, we will address the neuronal circuitry that could be causing olfactory issues in ASD. Recent reviews have focused on olfaction within the social aspects of ASD and potential circuit mechanisms [1] and psychophysical aspects of olfaction in ASD [2, 3]. Feeding and Eating Behavior in Autism Spectrum Disorder According to DSM‐5, ASD is a developmental disorder diagnosed based on deficits to age‐appropriate social communication [4]. Although reliable diagnosis by clinicians is typically not made until the age of 2 years, symptoms are known to present themselves from early childhood [5]. Sensory processing issues are a common symptom of ASD and are a criterion for diagnosis communication [4]. One of the various domains affected by abnormal sensory processing is feeding, which involves a dislike of food texture, color, and smell [6, 7, 8]. Feeding difficulties in ASD can manifest from an early age and persist into childhood and adulthood, with feeding problems developing into eating problems. A multicenter study analyzed the feeding behaviors through a caregiver questionnaire of 1389 children with autism and 1190 typically developing children and found that children with ASD have a later and poorer acceptance of foods other than breastmilk or formula [9]. Similarly, in the Avon Longitudinal Study of Parents and Children, it was found that after 6 months of age, infants with ASD had late introduction of solid foods and were described by their parents as “slow feeders” [10]. When these feeding problems extend into early childhood, they become what is known as picky eating, which is also a behavior seen in neurotypical children in early childhood. The definition of picky eating varies, but typically includes an unwillingness to eat new or even familiar foods leading to a restricted food intake [11]. Age of onset and incidence vary among studies. One study showed an increase in picky eating behaviors starting at age 2, with new incidence declining by age 6 [12]. Due to differences in the definition of picky eating, tools to measure its prevalence also differ. This yields a wide range of results, some indicating the prevalence may be as high as 50% of all children at 2 years of age [13]. As a result, picky eating is thought of as a normal part of early childhood and is reported to be equal in prevalence between neurotypical children and those with ASD [6]. While it is more common for typically developing children to outgrow their picky eating behaviors within 2 years of onset [12], children with ASD maintain their limited diets well into adolescence and beyond [14]. Severe cases of picky eating can meet diagnostic criteria for Avoidant Restrictive Food Intake Disorder (ARFID), in which restricted food intake results in a failure to meet nutritional or energy needs communication [4]. In one study, nearly 30% of children with ASD met the criteria for ARFID [8], and another found comorbid ASD in 58.3% of patients with ARFID [15]. This more severe extension of picky eating includes the behavior of food neophobia, which is the severe reluctance or avoidance of new foods. Food neophobia can drastically limit an individual's diet, and lead to very specific food selectivity. Most studies report greater food selectivity in children with ASD [6]. Atypical eating behaviors are 5× more common in youth with autism compared to youth with other neurodevelopmental disorders and 15× more common when compared to their typically developing peers. Limited food preferences was the most common atypical eating behavior reported in this study [16]. In the ASD population, picky eating and food neophobia may also be accompanied by rejection of changes to the mealtime environment. A study comparing mealtime behaviors among typically developing children and children with ASD found that those with ASD were more likely to insist upon sameness to food preparation, utensils, and dishes [17]. Food neophobia, ARFID, and other eating and feeding issues are apparent in people with ASD. Previous research has focused on common aspects of ASD, such as sensory processing issues and cognitive inflexibility, with respect to these feeding problems. However, less focus has been placed on the role of olfaction in the feeding behaviors of people with ASD. The Role of Olfaction in Food Neophobia in ASD Sensory processing abnormalities significantly contribute to food neophobia in ASD [18]. While texture and taste are hypothesized as factors that influence food neophobia, less focus has been placed on olfaction ([19]). Alterations to smell can change eating behaviors, as is seen in patients after chemotherapy treatment and Covid‐19 infection ([20]; Burges [21]). As a result, abnormal olfactory processing may be a key factor in the feeding problems seen in ASD. Children with ASD have poor hedonic discrimination of pleasant and unpleasant odors, with worsening ability correlated to increased food aversions ([19]). Children with ASD are also more likely to choose a food when it is presented with a familiar odor ([22]). With this in mind, it is clear that olfaction plays a role in feeding problems in ASD. The Role of Novel Food Odor Habituation on Methods Used to Introduce New Food Items in ASD There are many different tools to help expand the nutritional intake of children with ASD, some of which focus on social aspects of mealtime and others that address the sensory aspects [7]. One method [23] involves gradually exposing the child to new foods, initially by just having the food at the same table and then focusing on positive odor associations. Gradual exposure to the new food's odor leads to increased acceptance of the food item [23]. While effective, it may take weeks to months for a child to be ready to consume the exposed food. With the Autism Managing Eating Aversions and Limited variety (MEAL) plan [24], a clinician provides 1.5 h of intervention per week for 10 weeks with one family. Early sessions are dedicated to examining eating issues and behavior. Later sessions implement child‐specific feeding plans. Each session includes homework to encourage the application of new abilities. While the MEAL plan improved dietary intake during its clinical trial, methods like these can be invasive and time‐consuming for a family. Despite the existence of many feeding interventions, it is still neither clear why food neophobia arises in ASD nor why it takes such elaborate exposure strategies to reach new food acceptance in this population. Human Psychophysics of Olfaction in ASD Literature on ASD and olfaction reveals conflicting findings. Research on odor detection thresholds in ASD has indicated both hypersensitivity [25] and hyposensitivity [26, 27], with some research finding no altered odor detection [28, 29, 30]. Similar heterogeneous findings have been noted for odor identification abilities in ASD. Both impaired [27, 28, 29, 31, 32] and nonimpaired [26, 33] odor identification has been reported. Additionally, sniff modulation in response to the valence of an odor is reduced in ASD [34]. A meta‐analysis [35] investigating this phenomenon of conflicting findings revealed significant variations in effect size for odor threshold and identification. This study determined that participant age and IQ were potential variables contributing to the heterogeneity of results. Participant sex may also be a variable of note, as sensory issues between males and females with ASD are different. Females tend to report more lifetime sensory issues [36] and olfactory abnormalities [37] than males. With the exception of Tavassoli and Baron‐Cohen [30], all the previously mentioned studies included a majority, if not only, male participants. The methodology used in olfactory research varies. Many studies utilize Sniffin' Sticks (Burghart, Wedel, Germany), which has been standardized to investigate olfactory function [38] and includes 16 common scents. To try and minimize potential cultural bias, Muratori et al. [27] utilized a novel odor identification test that included 32 scents. However, their study found similar impairments in odor identification to those that used Sniffin' Sticks and those that used the University of Pennsylvania Smell Identification Test (UPSIT) [39]. Kumazaki et al. (2016) also took a novel approach by developing a pulse ejection system that improves testing precision and utilizes two fruity odors. This study corroborated previous results, showing hyposensitive odor detection thresholds in ASD. Ashwin et al. [25] took a unique approach to threshold testing and used the Alcohol Sniff Test, a method less commonly used in ASD olfactory research. This study is one of the only to report hypersensitivity to threshold testing in ASD. A key variable that has not been controlled for in previous research is novelty. Unfamiliar events like changes in routine and novel sensory experiences can be aversive to individuals with autism [40]. Therefore, the inconsistent results seen in olfactory research may potentially be due to the variable of novelty in the stimulus set. Additionally, children with ASD have reduced habituation to odors and might be affected by a novel smell for longer periods of time [41]. Both the UPSIT and Sniffin' Sticks utilize smells that are deemed common, but researchers do not determine if participants are familiar with the scents or expose participants to the scents ahead of time. Although this is done to maintain experimental validity, this method could be a source of variability. If a participant with ASD is presented with the smell of turpentine (used by the UPSIT and Sniffin' Sticks) for the first time during experimentation, they may have a negative reaction to the novel scent, thereby impacting their performance on future trials. Additional variability may be seen due to the cultural relevance of certain scents and differing cultural backgrounds of study participants. The issue of novelty not only extends beyond the scents used in an experiment but also to unfamiliar research staff and equipment. Little research has been done to address olfaction and novelty outside the context of food neophobia. However, a 2013 study found that exposure to a familiar scent (maternal body odor) encouraged imitation behaviors in participants with ASD [42]. This finding highlights the importance of addressing novelty in future ASD olfactory research. The Role of Odor Novelty in Olfactory Behavior in Mouse Models Relevant to ASD Multiple mouse models relevant to autism have been assessed for their social and repetitive behaviors, but less attention has been given to their olfactory behavior and to the effects of novelty. Nevertheless, a very widely used behavioral test, the three‐chamber test, measures the preference of a mouse between an inanimate object or an unfamiliar mouse [43]. As mouse social behavior relies on olfactory cues [44], the novelty of the unfamiliar mouse smell might play a role in the reduced preference for unfamiliar mice observed in mouse models relevant to ASD. These behavioral findings suggest that odor novelty affects both inbred lines and monogenic mouse models relevant to ASD. Burket et al. [45] assessed olfaction utilizing the Balb/c mouse model, a genetically inbred strain that is a proposed model relevant to autism [46]. This study assessed locomotive behavior in response to various odors. The Balb/c mice showed significant hesitancy in both moving and sniffing when exposed to novel floral and novel social odors. However, the behavior of the Balb/c mice was comparable to wild‐type when exposed to familiar floral odors, shown as a reduced initial latency of movement. Similar results were seen in the BTBR T+ tf/J mouse model [47]. BTBR T+ tf/J (BTBR) is an inbred strain exhibiting behavioral phenotypes with face validity to all three diagnostic symptom categories of autism. BTBR mice showed a decreased sniffing response to novel social and novel nonsocial odors as compared to wild‐type mice. A similar finding is also shown in the MALTT mouse model of ASD [48]. In a habituation/dishabituation task, MALTT mice showed decreased sniffing in response to a novel social odor but an intact capacity to detect and differentiate between odors. These results indicate that novel odors cause a significant behavioral change in inbred mouse models relevant to ASD. Similar hesitation surrounding novel scents has been seen in monogenic mouse models relevant to autism. The Nrxn1α+/mouse has construct validity as reduced NEUREXIN1 gene expression resulting from a 2p16.3 deletion dramatically increasing the risk of developing autism in humans [49]. This study utilized an odor discrimination task that required mice to dig in the sand at the location of a correct novel spice odor. Interestingly, both male and female Nrxn1α +/− mice learned the task in fewer trials than the wild‐type mice. However, the Nrxn1α +/− mice took longer to make correct choices during the task. That is, the Nrxn1α +/− mice were quicker to learn the novel odors, but showed hesitancy when approaching and digging for these odors. While male Nrxn1α +/− mice were found to be hyperactive, this finding cannot be explained by locomotive differences, as results from baseline locomotor activity testing did not correlate with the hesitancy found during odor discrimination. Interestingly, the Cntnap2 −/− mouse model related to autism [50] does not show olfactory deficits as measured by the time it takes to find familiar palatable food items buried in a familiar environment. However, the Cntnap2 −/− mice do show differences from wild‐type mice in relation to odor novelty as tested in the context of social interactions with unfamiliar mice. As mice social interactions are heavily dependent on olfactory cues [44], the novelty of the unfamiliar conspecific odor might play a role. Freely moving Cntnap2 −/− mice avoid entering and spend less time exploring and sniffing environments where a novel odor is present when compared to wild‐type counterparts [51, 52]. Mutations in the SHANK3 gene are one of the most common monogenic forms of autism [53]. Mice with deletion of all Shank3 isoforms [54, 55] also exhibit reduced exploration of odors in a habituation/dishabituation task. The results from these monogenic mouse models are similar to those of the inbred strain mouse models, both showing locomotive hesitancy around novel odors. Odor novelty also affects learning in the fragile X mouse model relevant to autism [56]. Learning of difficult odor mixtures using operant conditioning was affected, whereas simple monomolecular odor discrimination was not affected. These studies highlight the impact of odor novelty on detection and discrimination. Recent studies have directly addressed the effect of background odor novelty on two monogenic mouse models relevant to autism. WT mice are able to detect odors in the presence of other background odors [57] similar to other species [58]. WT mice could discriminate known targets in novel background odors, a task similar to the visual CAPTCHA used to distinguish humans from computers [59]. The Cntnap2 −/− mouse model of autism matched wild‐type mice performance in the presence of known backgrounds, but performance fell almost to chance levels in the presence of novel backgrounds. The inability of Cntnap2 −/− mice to recognize target odors in a novel background could be rescued by training the Cntnap2 −/− mice extensively with that background. During this training, the mice were initially presented with the background odor at low concentrations. The background odor concentration was increased gradually over multiple days. This result indicates that novel odors might not only be aversive by themselves, but they negatively interfere in the detection of simultaneously present odors. Similar results (see Figure 1) were reported for the Shank3‐deficient mice [60]. (A) Head‐fixed mice were trained in a go, no‐go task to detect target odors. (B) Mice were trained with a set of known background odors. (C) Mice were tested with novel background odors. (D) Trials with novel background odors were interleaved with trials with known background odors as catch trials. (D–F) Shank3B −/+ and Cntnap2 −/− mice performance for known odors matched that of the WT mice performance. (E–G) Mouse models of autism had deficits for odor recognition in novel background odors. Data are from Li et al. [59] and Ryndych et al. [60]. The enhanced masking effect produced by novel odors might also affect responses to social odors in mouse models relevant to autism. Urine is a complex and variable mixture, with only some of the components triggering a behavioral response [44]. Cntnap2 −/− mice show less exploration time of the urine of unknown mice [61], and male BTBR mice produce reduced urinary markings in response to female urine [62], consistent with enhanced masking of behavioral relevant volatiles in complex mixtures. The Effect of Odor Novelty in Neural Responses in Mouse Models Relevant to Autism Although novel odors have an aversive effect on multiple mouse models relevant to autism, there are no studies investigating how the novelty of an odor affects neural responses in the olfactory system of ASD mouse models or on the circuit level adaptation to new odors. Nevertheless, the evolution of the odor responses as they transition from novel to habituated odors stimuli has been recorded in WT mice. Odorants bind olfactory receptors expressed in the olfactory epithelium. These receptor neurons project to specific glomeruli in the olfactory bulb. These patterns of receptor activation produced by an odor are modified in the olfactory bulb by a network of interneurons. The output of this computation is transmitted by mitral and tufted cells to different brain areas [63, 64, 65]. In an experiment using chronic multiphoton imaging [66], it was shown that a novel odor produced large activation of the mitral cells. As the mouse was exposed to multiple presentations of the odor over the span of a few days, the mitral activation was reduced. This mitral cell habituation was an active process. The adaptation disappeared under anesthesia, and it was not present at the glomerular input level [67]. Therefore, novel odors trigger stronger activation in the signal that the olfactory bulb broadcasts to other brain areas. Neurophysiological responses to odors have been measured in the olfactory bulb in monogenic mouse models relevant to autism. Using intrinsic optical imaging, the glomerular responses to a panel of 20 odors were compared between WT, Cntnap2 −/− mice [59], and Shank3 −/+ mice [60]. There was no stark difference in the number of activated glomeruli or in the strength of the glomerular activation. However, awake Cntnap2‐ and Shank3‐deficient mice had higher trial‐by‐trial variability in glomerular responses, which was previously reported in neural responses in the olfactory bulb [68]. Increased variability was also reported in mitral cell responses to electrical stimulation of the glomerular input in slices in fragile X mice [56]. Increased trial‐to‐trial variability responses in the olfactory bulb might underlie some of the deficits seen in mouse models relevant to autism. The increased magnitude of responses produced by novel odors might act synergistically with the increased variability of the olfactory bulb responses in mouse models relevant to autism to affect odor discrimination. This would explain how novel background odors could produce large and variable responses that would mask the presence of other odors. At this point, it is not clear what circuit differences might underly the increased variability in mouse models relevant to autism. Mouse models relevant to autism have reduced inhibitory GABAergic responses [69]. Cntnap2 and Shank3‐deficient mice also show a reduced number of synapses from the olfactory receptor neurons into the periglomerular neurons, an inhibitory interneuron population located in the glomeruli [68]. Lack of excitatory drive in this interneuron population as well as reduced GABAergic activity might cause increased variability of responses in the olfactory bulb. Piriform cortical feedback targets mostly inhibitory interneurons in the olfactory bulb [70], which may contribute to adaptation of mitral cell responses after repeated exposure to an odor. Consistent with this hypothesis, piriform cortical feedback increases its strength as an odor is learned [71]. Increased feedback from the piriform cortex would increase the recruitment of inhibitory interneurons in the olfactory bulb, resulting in increased inhibition of mitral cell activity. It is an intriguing idea that cortical feedback from the piriform cortex might be reduced in ASD mouse models, resulting in reduced levels of inhibitory activity in the olfactory bulb. Neuronal variability produced by sensory stimuli in other sensory modalities can be dynamically reduced by attentional modulation in the visual system [72, 73] as well as pairing sound with cholinergic activation [74]. It is not clear whether similar mechanisms might be acting to quench the trial‐by‐trial variability in the olfactory system but might point to potential therapeutic strategies for olfactory deficits in ASD. Discussion Olfaction may play a role in food neophobia of individuals with ASD, but the neural mechanisms underlying this are unknown. Strategies have been created to expand the diets of people with ASD, with methods involving gradual exposure to food items and their smells, but it is not known why long periods of habituation are required for the acceptance of a new food. Psychophysical studies in ASD show contradictory results, including hyposensitivity, hypersensitivity, or no difference when compared to olfactory responses of neurotypical individuals. Besides differences in gender and IQ of the populations studied, it remains possible that these diverse effects might be caused by differences in the subject's familiarity with the odors used. Both inbred lines and monogenic mouse models relevant to autism show aversion to novel odors. Interestingly, novelty of an odor strongly modulates the responses of mitral cells, the output of the olfactory bulb in WT mice. This modulation might explain the diversity of the results in odor responses seen in individuals with ASD. However, there are no studies that have assessed this adaptation process in mouse models relevant to autism. Potential differences in the adaptation process in mouse models relevant to autism could offer clues to help understand food neophobia behaviors in ASD. We also review evidence of increased variability in the olfactory bulb in response to odors in mouse models of ASD. Understanding the mechanisms of increased variability may point to the development of potential treatments. Conflicts of Interest The authors declare no conflicts of interest. Funding: The authors received no specific funding for this work. Data Availability Statement Data sharing not applicable to this article as no datasets were generated or analysed during the current study. References Associated Data Data Availability Statement Data sharing not applicable to this article as no datasets were generated or analysed during the current study.