Autism and sensory sensitivity

Colour preferences are seldom mentioned in relation to mental illness; when food colour is discussed in conjunction with psychiatry, it is usually in reference to the potential relationship between artificial food colouring and symptoms of attention-deficit/hyperactivity disorder17 or other negative health consequences.18 Although numerous anecdotal reports of rigid colour-based food preferences exist—for example, a mother of two boys with autism has described how one of her sons would use ketchup to mask unwelcome colours, since he would only eat red foods19—there is relatively little research evidence on this specific phenomenon. Selective or ‘picky’ eating in a broader sense, as well as food neophobia (that is, the unwillingness to try new and unknown foods), is of course tremendously common among children. Most children display a reduced appetite and a decrease in growth rate between 2 years and 6 years of age as a feature of normal development.20 In a prospective study of 120 children in San Francisco, 13%–22% were reported to be selective eaters at any given age between 2 years and 11 years and 39% of the cohort were selective eaters at some point.21 Similarly, in a study of 426 German junior school children, 19% of the girls and 18% of the boys displayed selective eating—these children ate a more limited number of foods, more often required special preparation of meals, and their parents reported more struggles around food in the family as well as more worries about their children’s health.22 However, even though selective eating during childhood can certainly give rise to family conflict and turn restaurant visits and dinner parties into veritable obstacle courses, in a large majority of cases the child’s weight is, in fact, not negatively affected.23 Likewise, selective eating during childhood does not generally lead to eating disorders in the long run,24 although subgroups may experience rigidity around food of a more enduring nature.25

One such subgroup appears to be individuals with autism. Although not a definitive criterion in the current diagnostic classification systems, problems related to eating were repeatedly described in Leo Kanner’s original 1943 autism case series. Here, Kanner notes that ‘[f]ood is the earliest intrusion that is brought to the child from the outside’, an intrusion that can become overwhelming for children with autism.26 Most of the research on the nature of selective eating has, in fact, been performed with samples of individuals with autism spectrum disorder. A number of retrospective studies have reported low breastfeeding rates and early weaning in children who are later diagnosed with autism; this has often been interpreted as evidence for viewing breastfeeding as a protective factor, but could just as well point to feeding difficulties as an integral part of the disorder itself.27 A meta-analysis of research on food intake among children with autism found a fivefold increase in the odds of having a feeding problem in this population, selective eating being the most common obstacle.28 In a study comparing 53 children with autism with 58 typically developing children, the autism group was significantly more likely to refuse food based on texture and consistency (77.4% vs 36.2%), taste and smell (49.1% vs 5.1%), ingredients being served in a mixture (45.3% vs 25.9%), brand (15.1% vs 1.7%) and shape (11.3% vs 1.7%).29 In contrast, no differences were seen for food refusal based on temperature (30.2% vs 24.1%), foods touching other foods on the plate (20.8% vs 17.2%) or colour (15.1% vs 12.1%). Similar studies have repeatedly identified colour-based food selectivity as a real but relatively rare phenomenon30—time and again, texture is found to be the most common factor in food refusal. In general, children with autism tend to prefer crispy or crunchy food31 as well as salty and savoury dishes,32 but they may also be more likely to only accept foods of low texture, such as puréed foods.33 To parents’ chagrin, these children often become experts in detecting when unwanted food item is ‘hidden’ in other foods.34 Findings on selectivity based on food colour are, however, mostly anecdotal: one child will reportedly only eat red apples, another child refuses all vegetables that are not white,35 and so on.

This increased prevalence of food-related problems seems to extend to non-clinical populations with autistic traits as well.36 The reasons behind the observed sensory selectivity in individuals on the autistic spectrum are not fully known. A paradoxical simultaneous hyposensitivity and hypersensitivity to sensory stimuli has been noted in autism. Respondents may, for example, report being extremely sensitive to high-pitched noises or the buzzing of a fridge, while also requiring unusually powerful taste stimuli in order to experience food as appetising.37 A tendency to become overwhelmed by sensory stimuli is also common in autism, as well as strong preferences for certain stimuli qualities or sameness in stimuli. When it comes to general colour preferences, it has been hypothesised that an increased sensory sensitivity in autism results in aversion to specific colours. A comparison between neurotypical children and children on the autistic spectrum showed that while both groups liked the colours red and blue, the typically developing children also liked yellow and disliked brown, whereas the children with autism tended to avoid yellow and instead favoured brown and green.38 A possible explanation behind these findings is that in terms of retinal photoreceptor dynamics, yellow is the most fatiguing colour and may therefore become unbearably stimulating for children with autism. A number of visual function impairments have also been demonstrated in autism; among these, a generally poor colour detection capacity, rather than any specific colour defect (such as blue-yellow or red-green colour blindness), has been observed.39 But again, as noted above, sensory selectivity based on food colour specifically appears to be relatively rare even in groups with an overall inclination towards selective eating, such as individuals with autism. As described in more detail below, general colour preferences may also be distinctly different from food colour preferences in autism as well as in the population at large.

It should be noted that texture, rather than colour or other food properties, is generally also the most important sensory aspect behind food rejection in non-autistic, non-clinical populations:40

When customers complain about products available in a grocery store or at the market or when diners in a restaurant express dissatisfaction with a dish, it can almost always be traced back to mouthfeel. We seldom complain that something tastes bad. Instead we might say that the soufflé has collapsed, the meat is too tough, the French fries have gone soggy, the bread is dry, the coffee is tepid, the mustard lacks bite, and so on. […] It is much easier to describe an unfulfilled expectation related to texture than one that concerns chemical sensations, such as taste and aroma.41

In particular, the most frequently used word to describe positive connotations regarding texture, at least in Europe and the USA, is ‘crisp’;42 as seen above, this holds true in autism as well.43 In contrast, autistic treats or not, Satie’s white diet is hardly characterised by crispness. Rice, eggs and ‘the fat of dead animals’ are certainly not the kind of foods that are usually associated with a crunchy texture and they are highly unlikely to appeal to the typical individual with rigid texture-based food preferences.

The boy with the blue-tinted glasses

In a fascinating case report, Amanda Ludlow and colleagues describe a young boy, referred to as J.G., with autism spectrum disorder and strong reactions to multiple types of sensory inputs, including colour obsession and phobias.44 In particular, J.G. has a marked preference for blue and purple, combined with an aversion to bright colours. His family has attempted to adjust his home environment according to these preferences in order to reduce stress: J.G.’s bedroom is painted purple, his clothes are all dark blue, and the family drives a blue car with purple interior. Reportedly, J.G. once even attempted to paint the family dog blue. This colour obsession also impacts on his eating. Since blue or purple foods are highly unusual, J.G. opts for white food items and cannot tolerate more than two colours at once on his plate. Thus, between ages 1 and 9 years, most of his meals were served between two slices of bread.45 J.G. reports painful sensations in his eyes as well as nausea when exposed to bright colours—consequently, he wears sunglasses at all times, even indoors. On a chromatic discrimination task, perhaps somewhat surprisingly, J.G. displayed a pattern of selective deficit in the very colour region that he prefers, indicating that his attraction to blue and purple may be due to a hyporeactivity to these colours, thus avoiding sensory overload. Interestingly, in a follow-up report the authors describe how J.G. had subsequently received blue-tinted glasses, which significantly reduced his adverse reactions to aversive colour stimuli and enabled him to partake in everyday activities.46 It can be noted that blue-tinted spectacles are sometimes prescribed as a remedy for the ‘visual stress’ associated with the Mearles-Irlen syndrome,47 a proposed and somewhat controversial diagnosis involving visual distortions as well as (perhaps) learning disabilities and dyslexia.48 Furthermore, blue-tinted glasses have actually been marketed as a dieting aid by a Japanese company49—supposedly, they act as an appetite suppressant precisely due to the scarcity of naturally blue foods.50

Another psychiatric condition that may involve selective, colour-based eating is the so-called avoidant/restrictive food intake disorder, usually abbreviated as ARFID, that was introduced as a new diagnosis in the eating and feeding disorders chapter of the fifth edition of the Diagnostic and Statistical Manual of Mental Illness in 2013.51 ARFID is explicitly an umbrella term, used to cover several types of restrictive eating with the common denominator that they do not occur in a context of body image concerns or a wish for thinness, as in anorexia nervosa or bulimia nervosa. Emerging evidence exists for the subdivision of ARFID into three clinical dimensions: low homeostatic appetite, fear of aversive consequences such as choking or vomiting, and sensory sensitivity of the kind described above.52 These presentations are not mutually exclusive. It has been noted that the symptoms of an individual with ARFID can often be plotted along all of these three dimensions; a typical case could, for example, involve a history of chronic low appetite and ‘picky eating’ during childhood which is later acutely exacerbated after a traumatic event, such as choking on a piece of food.53 Interestingly, there is often a potentially confusing comorbidity between ARFID and autism spectrum disorder; here, the ARFID diagnostic criteria specify that if the restrictive eating is part of broader mental health issues, the ARFID diagnosis should only be used when the severity of the eating problems exceeds that which is commonly seen and warrants additional clinical attention.54

What colours do people like, and why?

Certainly, colour-based food preferences are not a unique feature of individuals with autism spectrum disorder or ARFID—far from it. In fact, the colour aspect of food and drinks is vital for our experience of flavour and can, if slightly off, act as an effective appetite killer. In an oft-cited 1970s study,55 dinner guests were served steak, chips and peas in a dimly lit environment specially designed so as to show the food on the plates but not its colour. When the lights were turned on fully halfway through the meal, it was revealed that the peas had been coloured red, the chips lime green, and the steak a vibrant blue. Reportedly, at the sight of the artificially coloured food, several dinner guests immediately fell sick and headed straight for the bathroom. There is a well-documented general aversion to blue food, which is especially striking given that blue is often rated as the most preferred colour in non-food contexts.56 This could be a result of there not being very many naturally blue food items, so that the colour blue in food has come to be associated with unappealing artificiality; alternatively, human evolution may have made us implicitly associate blue food with mould and risk of poisoning. Culinary experiments with sushi dipped in blue food dye,57 neon purple udon noodles served with blue tofu,58 or glow-in-the-dark ramen59 are fascinating as expeditions into the overlapping territories of cooking, art and science—nevertheless, in the case of blue sushi, none of the participants even dared to try it.60

Scholarly interest in the psychological aspects of colours can be traced back to German writer Johan Wolfgang von Goethe and his work Zur Farbenlehre (or Theory of Colours), which was published in 1810. In a wide-ranging exposé, Goethe here attempted to unravel the perceptual essentials of colours and their ‘sensory-moral’ effects,61 symbolically associating red and crimson with the beautiful, yellow with the good, green with the useful, and so on. Goethe’s ideas were later integrated with clinical observations by German neurologist and psychiatrist Kurt Goldstein, who suggested that the colours green and blue have a relaxing, stabilising and introspective effect, whereas red and yellow are stimulating and prompt an outward focus.62 Interestingly, in line with Goldstein’s theory, studies on the psychophysiological effects of coloured medicine tablets and capsules show that red, yellow and orange drug formulations tend to be perceived as stimulant, in contrast to green, purple and blue which are seen as tranquillising and are, consequently, more often used for sedative and anxiolytic drugs.63

In experimental research on colour preferences, people systematically prefer cool colours such as blue, purple or green over warm colours—blue tends to be the most preferred colour whereas yellow and yellow-green are least favoured.64 Higher saturation and increasing lightness, at least to a certain point, is also usually preferred. However, different colours are preferred at different levels of lightness. In particular, dark yellow and orange—olive and brown, that is—are much disliked relative to lighter tones of the same hues. As noted above, there may be evolutionary components at play behind these tendencies. It has been suggested that cyan and green are generally attractive because they elicit positive affects through their association with clean water, clear skies and lush foliage, whereas various shades of brown are implicitly coupled to faeces and decaying food that is best left untouched.65 Even so, for some reason, people are obviously capable of harbouring a strong affection for such brown food items as chocolate or coffee. As we have seen, colour preferences are highly contextual: even individuals with a love for the colour blue usually like their bananas yellow.66

Deceiving our taste buds

Not least, food colour helps us in identifying what it is we are eating. There is an extensive experimental research literature on the impact of food colour on taste identification and intensity and a full review is beyond the scope of this article (see instead, for example, Spence 2015). Nevertheless, a few illustrative examples are useful to contextualise the findings on sensory preferences in autism and ARFID. In a classic 1974 study, flavour chemist Joseph Maga demonstrated how adding tasteless red, green or yellow food colouring to clear water solutions altered the threshold sensitivity for the basic tastes sweet, sour and bitter.67 For example, after adding green colouring to a sucrose-water solution, participants described it as sweet at a lower sucrose concentration than before. In contrast, adding yellow or red colouring made the sweet solution more difficult to detect. In the case of sourness, both green and yellow colouring decreased sensitivity, and red colouring made it more difficult to detect bitterness. Notably, participants’ detection threshold for saltiness was not altered by adding food colouring, perhaps because various foods of different colours (potato chips, popcorn, olives, pickles and so on) are associated with a salty taste. Later studies, however, have demonstrated that people do tend to associate saltiness with the colour white.68

In the same vein, a study by wine scientist Wendy Parr and colleagues investigated olfactory perceptual bias in casual wine drinkers as well as in wine connoisseurs and demonstrated that both groups tended to describe odours typical to red wines (cherry, plum, black pepper and so on) when tasting white wine that had been coloured red.69 This study has recently been replicated and extended, demonstrating how the pink hue of rosé wine similarly influences the perceived flavour and aroma in wine novices a well as in experts.70 Context is also important: a blue-coloured drink in a plastic cup may be expected to taste of mint—as in mouthwash—whereas the same drink presented in a cocktail glass may instead signify the orange flavour of blue curaçao.71 In general, people tend to expect more intensely coloured food to have a more intense taste; however, at a certain point, vivid colours may lead consumers to associate food with artificiality and therefore like it less.72 Furthermore, there are cross-cultural differences in common connotations between colour and taste. In a study comprising 452 participants from China, India, Malaysia and the USA it was found that individuals from all four countries associated black colour with bitterness, white with saltiness, green with sourness and pink with sweetness.73 However, Chinese participants tended to associate purple with sourness, whereas Malaysian participants thought of green as distinctly sour. Interestingly, the US participants more often thought of blue as sweet—perhaps due to a culture-specific association between the colour blue and raspberry flavoured soft drinks, which, in turn, can be traced back to a more or less haphazard use of novel artificial food colouring in the early 1970s.74

In contrast, foods that are served in a colourless clear or white form usually do not taste very much at all.75 This may very well be a sought-after characteristic in certain contexts. As noted above, for lack of blue and purple foods, autistic boy J.G. preferred to have his meals served between slices of white bread. Likewise, a case report describing a 42-year-old man with ARFID reports how the patient displays a clear preference for ‘plain foods’, such as white bread sandwiches.76 In their manual for cognitive-behavioural therapy in ARFID,77 therapists Jennifer Thomas and Kamryn Eddy note that individuals with ARFID characterised by sensory sensitivity often follow a ‘white diet, consisting, for example, of white-coloured dairy and grains (e.g., pasta with butter, vanilla ice cream, crackers, breakfast cereal) and few or no fruits, vegetables, or proteins’—substitute camphorised sausage with spaghetti and ice cream and we have a 21st century version of Satie’s diet. It may also be noted that, in contrast, contemporary promoters of ‘healthism’ have launched a ‘nothing white’ diet,78 focusing on a sharply reduced intake of refined carbohydrates, although the rationale behind this can be questioned.79

There have been anecdotal reports of celebrities with a special fondness for white foods—however, it is usually highly uncertain whether this is due to ‘real’ sensory sensitivity, an overall air of eccentricity or something else. Reportedly, fashion designer Yves Saint Laurent preferred his food white80 (although his favourite colours in other contexts were black and pink81). Likewise, movie producer Leland Hayward is said to never have wanted anything but white foods, such as scrambled eggs, chicken, custard and vanilla ice cream.82 It has been suggested that the symbolic association between the colour white and purity, just as in the case of Satie,83 might explain these artistic types’ fondness for white foods.84

Such an association between whiteness and purity need not only be symbolic. Historically, it can be noted that in the latter half of the 19th century, just around the time that Satie was born, big food manufacturers first started employing industrial chemists who were tasked with inventing ‘new aids to deception—deodorants, dyes, flavours, crispers for flabby foods, and softeners for hard foods—which left consumers thoroughly confused about what they were eating’.85 Population growth and increased urbanisation called for novel methods for the production of cheap and durable food, which included improving the looks of mass-produced food items with artificial food colouring. Public and government reactions to these developments were highly ambivalent. On the one hand, scientists of the time assured the public that the new inventions in food technology were beneficial and safe; this was indeed the era of burgeoning positivism and engineers were eager to provide rational solutions to the societal problems at hand. On the other hand, many people found the technological advancements intimidating. Concerns about the safety of the various artificial additives were raised and a new league of food ‘purists’ warned against insidious ‘cheating’ and ‘adulteration’.86 For example, German-born scientist Frederick Accum authored a much-debated book in which he uncovered how the rainbow colours of boiled sweets were produced by using poisonous copper salts and how the orange rind of Gloucester cheese was created with the help of red lead.87 Antiadulteration associations were founded and bills prohibiting the use of various artificial components were brought before the parliaments of Britain, the USA and other countries. One can imagine how a widespread suspicion against food chemists and deceitful grocers might have contributed in steering individuals with rigid food preferences towards an even more minimalistic diet—excluding, for example, colourful foods—around the turn of the 20th century.

Conclusion: Satie, Huysmans and the culture of boredom

So, does not Satie indeed qualify for an ARFID diagnosis—he who never spoke while eating, ‘for fear of strangling’? Or are his mealtime preferences rather the hallmark of an individual with autism? Was he perhaps afraid of food adulteration? As it turns out, in the risky business of retrospective diagnostics, Satie’s writing on chicken and fuchsia juice does not tell us much of anything about himself, simply because it is not autobiographical, as some have implied;88 parts of his white meal are virtually inedible anyway.89 Instead, ‘The Musician’s Day’ should be read as an ironic take on contemporary literature. Readers may already have noticed that the white meal described by Satie is clearly inspired by yet another monochromatic meal in literature: the all-black dinner party thrown by the eccentric aristocrat Des Esseintes in French novelist Joris-Karl Huysmans’ À Rebours (also known as Against Nature), published in 1884:

The dining room, draped in black, opened out onto a garden metamorphosed for the occasion, the paths being strewn with charcoal, the ornament pond edged with black basalt and filled with ink, and the shrubberies replanted with cypresses and pines. The dinner itself was served on a black cloth adorned with baskets of violets and scabious; candelabra shed an eerie green light over the table and tapers flickered in the chandeliers. […] Dining off black-bordered plates, the company had enjoyed turtle soup, Russian rye bread, ripe olives from Turkey, caviare, mullet botargo, black puddings from Frankfurt, game served in sauces the colour of liquorice and boot-polish, truffle jellies, chocolate creams, plum-puddings, nectarines, pears in grape-juice syrup, mulberries and black heart-cherries.90

A number of aspects—the demoralising defeat in the Franco-Prussian war, undercurrents of philosophical pessimism inspired by Schopenhauer’s writings and an increased interest in the occult—interacted in steering French as well as broader European culture away from rationalist positivism towards pessimist symbolism, spiritism and a medical obsession with neurological degeneration in the last few decades of the 19th century.91 Huysmans can be seen as a portal figure of these tendencies. In À Rebours, a novel that has ‘come to a near standstill’, the main protagonist Des Esseintes is a hedonist Parisian dandy plagued by the ever-present ‘sickness of the century’ that afflicted so many of his contemporary young adults: disillusionment, rootlessness, crippling feelings of banality and a deep-seated sense that they had been born into ‘a world which seemed not to need them’.92 In response, Des Esseintes withdraws to live alone in an isolated villa on the outskirts of the city where he spends his days indulging in exotic jewels and perfumes as well as in his own impotence and digestive problems. Here, Huysmans combines naturalist interest in the very concrete—reflected in Des Esseintes’ meticulous attention to the details of refinement, such as in the famous description of the gem-encrusted shell of a living tortoise—with symbolist and decadent subjects of solitude, perversion and nausea. Indeed, these ingredients have over the years turned À Rebours into somewhat of a cult classic; the novel has also, however, been laconically dismissed as ‘a book about interior decoration’.93

Huysmans’ meticulous description of the black dinner party could be compared with modern-day ‘food porn’: sensual images of appealing dishes that are consumed visually on blogs or in glossy magazines rather than viscerally at the dining table.94 Surely, in this sense Satie’s ironic diet is not simply a white version of Des Esseintes’ spleen-drenched extravagance; there is nothing pornographic about the veal, salt and coconuts of his rigidly anaemic meal. Whereas Des Esseintes’ dinner is pure decadence, that of Satie is austerity and mortification. It has been suggested that the liquorice and boot-polish dinner party conjured up by Huysmans is ‘a feast for the eyes rather than for the palate […] served at the kind of party to which you want to be invited, but not the place to go if you’re hungry’.95 Certainly, Satie’s turnips and rice do not induce much of a party mood at all, but they are probably also unlikely to satisfy the hungry gourmand. If Satie’s white meal—although hardly an autobiographical account—evokes the idea of an individual with selective eating and sensory sensitivity, perhaps against a backdrop of autism-like traits, Des Esseintes’ black meal is that of exuberance, decadent artificiality and fin-de-siècle existential boredom. And, as Peter Toohey notes in his book Boredom: A Lively History, boredom ‘is an emotion usually associated with a nourished body: like satiety, it is not normally for the starving’.96 The spare, white meal described in ‘The Musician’s Day’ is, in contrast, indicative of a need to indeed create something radically new in reaction to the prevailing self-indulgent pessimism of the time—after all, this is the author-composer who, as an ironic take on boredom, created ‘muzak’. In retrospect, we can be certain that if anything, Erik Satie was not bored.