Preference to Eat Alone: Autistic Adults' Desire for Freedom of Choice for a Peaceful Space

Abstract

Backgrounds:

Autistic adults tend to prefer eating alone, which may result in further social isolation. This study aims to understand contexts of challenges autistic adults face when eating with other people outside home.

Methods:

We explored autistic subjectivities actively situated in eating environments. We conducted follow-up individual email interviews with autistic adults diagnosed with an autism spectrum disorder in adulthood (n = 10) and autistic adults without a formal diagnosis (n = 6), both of whom had participated in our prior online survey with Brazilian autistic adults. We used a thematic analysis to gain an in-depth understanding of challenging characteristics of social situations in the eating space.

Results:

To eat out comfortably, many participants did not talk while eating, chose places without sensory overload, chose places with eatable foods, ate at off-peak hours, used sensory tools, and/or left places when stressed. Participants developed these coping strategies to negotiate incongruences between their autistic personal characteristics and their physical environments. However, there were incongruences between the coping strategies and social contexts with codiners. Participants often could not freely use the strategies—both in work and in casual social situations. This lack of freedom of choice often resulted in feeling uncomfortable, sick, or having meltdowns during or after social dining. Because of these experiences, many participants decided to avoid social dining.

Conclusions:

The preference to eat alone in autistic adults may reflect their desire for freedom of choice to use coping strategies to eat in a peaceful food space. Creating a peaceful eating environment may be a feasible and humanistic approach to reduce unwanted aloneness and social isolation in autistic adults.

Lay summary

Why was this study done?

Eating together with other people is a common activity to socialize and feel a sense of belonging. But autistic adults often prefer to eat alone, though they have a desire to socialize.

What was the purpose of this study?

To study challenging aspects of eating out with other people, to understand why some autistic adults prefer to eat alone.

What did the researchers do?

We interviewed adults who considered themselves to be on the autism spectrum. Of the 16 interview participants, 10 were formally diagnosed with an autism condition and 6 were not. They once participated in our online survey with Brazilian autistic adults. The interviews were done by email. We asked about eating places and experiences in eating out. We looked for the common themes over all the interview responses.

What were the results of the study?

When eating out alone, participants had strategies to eat comfortably. They could eat without talking, choose places without sensory overload, choose places with eatable foods, eat at off-peak hours, use tools such as earphones, or leave the place when stressed. However, they often could not freely do so in both work and casual social situations. They knew that their coping strategies would be considered inappropriate by others. For example, many participants felt pressured to talk while eating. Also, other people had different preferences for the times and places to eat from participants. It was difficult for participants to deal with the stressful environment without using their coping strategies. So, participants often felt uncomfortable, sick, or had meltdowns during or after eating out with others. For these reasons, many of them avoided eating out with others.

What do these findings add to what was already known?

Autistic adults' preference to eat alone and difficulty in social situations involving eating have been understood as eating problems in autism. Our findings suggest that some dining companions reduce autistic persons' freedom to eat in comfortable social and physical environments, and that the preference expresses a desire to have such freedom.

What are potential weaknesses in the study?

Because we used email interviews, the study did not include any participant without access to email, and participants were all college educated, unlike most Brazilian adults with an autism diagnosis.

How will these findings help autistic adults now or in the future?

The findings will help autistic adults by letting others know ways that dining companions potentially reduce autistic persons' freedom to eat comfortably and by inspiring them not to do so. The findings also suggest that society should focus on creating a peaceful eating environment instead of focusing on finding ways to treat autistic aloneness as if aloneness in autism is caused by problems inside of the autistic individual.

Introduction

Eating with other people is typically associated with socializing and happiness.^1,2 Socializing is also a desire of autistic adults, especially women.³ Still, autistic adults of average to high cognitive abilities report preferring eating alone, contrasting with their neurotypical peers.⁴ Autistic adults' preferences have been speculated to arise from a “difficulty to do two things simultaneously” in “social situations at mealtime.”⁴ Their aversion to certain environments and behaviors of others, their routines and preferences around food, and their self consciousness around their own eating behaviors related to difficulty in social situations involving eating.⁵ In an online survey in Brazil,⁶ we have found that approximately half of both 43 diagnosed and 38 self-identified autistic adults find difficulty arranging social dining, having a conversation during a meal, and hearing the sounds of others eating. Other challenges included strict place preferences, dietary restrictions, eating timing, and feeling judged. Further studies are needed for an in-depth understanding of these challenges and their implications on their preference to eat alone.

In some eating behaviors, such as eating alone, the influences across external contexts, and their interplays among them and with personal characteristics, may be important. From an ecological perspective, “eating behavior is highly complex and results from the interplay of multiple influences across different contexts” and the different contexts—“individual, social environment, physical environment, and macrolevel environments—all interact, both directly and indirectly, to impact eating behaviors”⁷ (p. 254–255). From this perspective, there is little research related to preference to eat alone or challenges in eating with others. Approaches that focus on the person as the location of problems try to fix their autism and their “problematic” behaviors, have been challenged both by critical autism scholarship and by autistic self advocates.^8,9 Environmental psychology has explored the roles of the environment on autistic behaviors.¹⁰ Also, autistic self advocates have requested society shift its focus from pathologizing autism to creating supportive environments for their quality of life.⁹ Yet, subjective voices of autistic adults have been marginalized in research about the environment that surrounds and interacts with autistic persons. To understand autism-friendly social and physical environments,¹¹ it is important to listen to autistic subjectivity¹² and to lived experiences.

In the past, the first author¹³ explored how autistic adults negotiate and navigate their food environments, through the voices of U.S. adults on the autism spectrum. The study suggested they experienced food environments as invisible food deserts, which did not accommodate autism-related needs, while hostile characteristics were not obvious to nonautistic people. A characteristic of invisible food deserts was “‘unsafe’ people who enforce unwanted social interactions or diets”; the other characteristics were “limited availability of edible foods,” “prevalent over-stimulating food places,” and “hours of operation experienced as restricted” (p. iv). The study had the potential to explain autistic adults' difficulties in eating out with others and preferences to eat alone. However, the study did not thoroughly explore this aspect. A further study of characteristics of “unsafe” people in relation to the other characteristics of invisible food deserts may advance our understanding of autistic adults' difficulties in eating out with others.

Here we investigate relationships between contexts of challenges in eating out with others. Our purpose is to understand autistic adults' preferences to eat alone and to conceptualize an autism-friendly eating space. We draw on critical autism studies¹⁴ and an ecological perspective in environmental psychology.¹⁵ We based data collection on interviews with Brazilian autistic adults.

Methods

The institutional review board at the University of São Paulo approved this study. The data analyzed for this article were from our research project on autistic subjectivity and the production of food space based on the voices of Brazilian adults on the autism spectrum. This project's purpose was to understand the situated character of autistic adults in the food space that produces them and is produced by them. The project had an online survey, with results analyzed and published elsewhere,⁶ and optional email interviews, which we analyzed for this article. We conducted all surveys and email interviews in Brazilian Portuguese.

Participant recruitment

We recruited interview participants among participants of the prior online survey.⁶ For the survey, we had contacted personal acquaintances (including a nutritionist, university professors, and psychologists), a dentist for autistic people, and autism advocacy organizations, asking for help in disseminating the survey through their social networks to Brazilian autistic adults. The invitation for the survey informed that adults who identify as being autistic, either with or without a formal diagnosis, are eligible to participate. The other inclusion criteria were being able to communicate in written Portuguese, live in Brazil, and be between 18 and 65 years old. We recruited autistic adults both with and without a formal diagnosis, for many autistic adults do not have formal diagnoses,¹¹ and autism is underdiagnosed in Brazil.¹⁶ To screen out nonautistic individuals, we used the Ritvo Autism and Asperger Diagnostic Scale-14 (RAADS-14),¹⁷ as a part of the survey. A cutoff score for Autism Spectrum Disorder suggested by the developers of the scale is 14. The RAADS-14 scores of the 43 diagnosed participants (median = 34, mean = 34, standard deviation [SD] = 6, minimum [Min.] = 18, maximum [Max.] = 42) and of the 38 self-identified participants (median = 33, mean = 33, SD = 6, Min. = 20, Max. = 42) showed that there were no significant differences between the two groups in terms of their likelihood of scoring above the cutoff. For the interviews, we invited all but two survey participants who provided email addresses for an invitation for optional follow-up email interviews; we excluded the two for being personal acquaintances of one of us. Thirty-seven survey participants consented to participate in the interview, through an Internet-based informed consent form. A total of 16 participants provided their responses to the interview questions (See Table 1 for interview participant demographic information).

Table 1.

Participant Demographic 1, Autistic Adults (N = 16)

	Diagnosed	Self identified	Total
	n	n	n (%)
Age	Mean = 32.7	Mean = 40.2	Mean = 35.5
	(range 27–53)	(range 31–50)	(range 27–53)
Age at autism diagnosis	Median = 28	N/A	N/A
	(range 23–52)
RAADS-14 score	Median = 33	Median = 38	Median = 34
	(range 20–40)	(range 29–42)	(range 20–42)
Autism diagnosis
Autism spectrum disorder	5	0	5 (31)
Asperger's disorder	5	0	5 (31)
Without a formal diagnosis	0	6	6 (38)
Gender
Male	3	2	5 (31)
Female	6	4	10 (63)
Other	1	0	1 (6)
Race
White	7	3	10 (63)
Pardoa	2	3	5 (31)
Asian	1	0	1 (6)
Educational achievement
Some college	4	2	6 (38)
Attending college	1	0	1 (6)
Bachelor's degree	3	1	4 (25)
Academic specialization	1	1	2 (13)
Graduate degree	1	2	3 (19)
Marital status
Single	6	5	11 (69)
Married	2	1	3 (19)
Civil union	2	0	2 (13)
Employment
Full-time work	5	3	8 (50)
Part-time work	1	0	1 (6)
Do not work	2	2	4 (25)
Full-time student	1	1	2 (13)
Retired	0	1	1 (6)

^a Pardo refers to Brazilians of mixed ethnic ancestries typically of European, African, and/or Native Brazilian.

N/A, not applicable; RAADS-14, Ritvo Autism and Asperger Diagnostic Scale-14.¹⁷

Study design

We used structured individual email interviews with tailored questions for each participant based on the participant's survey answers. We adapted the interview questions (Supplementary Appendix SA1) from the first author's doctoral dissertation.¹³ We translated the original questions into Portuguese. Reminding participants of their own relevant survey answers, the questions ask about eating out experiences (i.e., favorite and least favorite places, frequented places, the underlying circumstances compelling to eat out at the particular places, a recent annoying experience, and efforts to change the situation that makes it difficult to eat comfortably) and probe coping strategies for eating out comfortably (i.e., how one came up with ideas for coping strategies, what are challenges in using the strategies). We added a question that probes challenges in eating with others.

Data collection and analysis

We conducted email interviews with 10 diagnosed and 6 self-identified autistic adults. We limited data collection to a single email exchange for data analyzed for this article. Most participants provided their responses as an attached file. The individual responses regarding eating outside home counted on average ∼1000 words.

We used an inductive thematic analysis—a qualitative method that is used to search for keywords, themes, or ideas that emerge from the data—to identify themes at a semantic level.¹⁸ We also used three criteria of a theme: recurrence, repetition, and forcefulness.¹⁹

Results

We have identified three themes from participants' accounts: (1) incongruences between coping strategies and social contexts with codiners; (2) experiences of feeling uncomfortable, sick, or having meltdowns; and (3) avoidance of social dining. An overarching theme, the desire for freedom of choice for peaceful eating space, emerged.

Randomly assigned participant identifiers (IDs) below with a letter D refer to diagnosed participants and those with a letter S refer to self-identified participants.

Theme 1: Incongruences between coping strategies and social contexts with codiners

All participants have developed various coping strategies to eat out comfortably. Their strategies included not talking while eating, choosing places without sensory overload, choosing places that catered to dietary requirements and preferences, eating at off-peak hours, using sensory tools, and leaving places when stressed. The strategies were to negotiate incongruences between their autistic personal characteristics and the physical environment. Many of their strategies, however, were incongruent with their social contexts if they dine with certain people.

Contexts of development of coping strategies

Most coping strategies related to sensory sensitivity to foods and ambient environments. Ambient environments in eating places were often overwhelming to many participants. Some strategies were specific for other autistic personal characteristics, such as dietary restrictions and communication difficulty, which also were influenced by sensory sensitivity.

Many participants rarely talked during a meal: “I enter quietly and leave in silence because that's the only way I can eat” [S3].^* Having conversations during a meal was very difficult, not necessarily for lack of communication skills per se. For some participants, overstimulating eating places were too distracting for conversations. Difficulty to multitask eating and talking also discouraged mealtime conversations: “I have a little trouble talking, because I feel like I can choke. I either eat or talk…” [D8].

Participants were very selective about eating environments to avoid sensory overload. Eating always in the same places was common partly for this reason: “I like to always go to 3 places, which follow this profile of being spacious and not too noisy” [D1]. Characteristics of their (least) favorite places related not only the levels of spaciousness and noise but also lighting, temperature, ventilation, smells, and availability of certain types of seats in certain locations in a place.

Many participants ate in a few eating places also because places where they can find eatable foods were rare. Sensory sensitivity to certain textures, tastes, and smells of foods was the main reason for dietary restrictions. For example, common food seasoning could create uneatable food: “I DETEST the fact that Brazilians put salt, pepper, onions, and garlic in everything” [D10]. Places' ability to accommodate certain special instructions, such as not putting raw onion, seemed critical for several participants. Also, some participants had a sensitivity to some allergens.

Eating at off-peak times was one of the most common coping strategies for participants to avoid sensory overload and eat comfortably: “The most efficient is still eating at odd times, which provides me with quiet environments for the absence of the main sources of noise” [D6]; “When I am alone, I usually choose places I know will be empty and I go to different ‘standard’ times at lunch or dinner” [D7].

Participants used several sensory tools to deal with a sensory overload when they eat alone. Earphones were the most commonly used sensory tool: “The earphones are a way of ‘muffling’ the outside noises and ultimately listening to a song that calms me and prevents me from having meltdowns” [D4].

Also, leaving eating places when stressed was a very important coping strategy. Almost all participants responded that they would just leave the place rather than trying to change the uncomfortable situation or environment.

Social contexts with codiners

Codiners often inadvertently created a space where social norms, their preferences, and judgmental attitudes, or just their presence hindered participants from freely using coping strategies to eat comfortably in both work and casual situations.

Many participants felt pressured to talk because codiners preferred to talk and because having conversations while eating with others was a social norm: “The person would want to talk, and I would often be paying more attention to the noise of the environment than to the subject…” [D8]; “Then I can't help but talk” [S3].

Places chosen by dining companions were often sensory overloading: “When other people eat out, they have one universe of considerations much smaller than mine. Many times, chosen places are uncomfortable for me because of sensory issues, but the sensory details in question go unnoticed by others” [D6]. Although places become more sensory overloading during peak times because of crowding, social dining usually happened during those times because of codiners' preferences: “Eating with friends is always a hassle because they always want to eat at ‘peak’ times in hot spots” [S3].

Many participants could not use a sensory tool such as earphones while eating with others: “I try to use earphones whenever I eat alone to block background sounds, but I can't do the same when I'm with others for social reasons” [D7]. Some participants mentioned it is impolite to wear earphones when eating with others.

Places that other people like to go to rarely had eatable foods for participants with dietary restrictions, but not eating while others eat was not a social norm. For this reason, a participant on a gluten-free dairy-free diet had to refuse to join her friends. Some participants who prefer to eat the same foods and dislike trying new foods experienced judgments from others on eating nothing: “…people complain that I go but eat nothing or that I ask for something but remove almost everything that came in the food, it becomes a comment, it becomes a joke…it is annoying sometimes…” [D1].

When participants were with companions, leaving the sensory overloading place was difficult because it would upset others: “I have a lot of difficulty with environments I don't control. If I'm alone, I eat as fast as I can and I'm gone, but if I'm with friends, there's no way” [S3]. Besides gatherings of family and friends where one's “presence at the place is really needed,” a year-end company party (confraternização de empresa) was a social dining situation where a participant who used to work felt uncomfortable but had to stay “in order not to make other people annoyed” [D7].

Theme 2: Experiences of feeling uncomfortable, feeling sick, or having meltdowns

When participants could not freely use their coping strategies to eat comfortably because of codiners, they experienced negative consequences from the lack of freedom of choice. They felt very uncomfortable, sick, or even had meltdowns.

The levels of discomfort were intense, even for common social dining situations. A participant who avoids eating at peak times felt it is “torture” to eat with a stranger nearby while having to eat out with friends during peak time in places with the practice of sharing tables [S3]. A participant feels “horrible” if she is at lunch with someone with whom she does not have intimacy, including work colleagues: “I can't swallow. My throat locks shut…What I do is eating very little and very slowly not to choke” [S1]. Many participants had experiences of feeling trapped and suffering in sensory overloading social occasions. A participant detailed,

I went to a restaurant with my parents and siblings to celebrate someone's birthday…I didn't bring earplugs, and the restaurant was packed, cutlery noises banging on the plate, lots of people (the table organization was chaotic, and the people were “sewing” between the tables), a lot of smells. I couldn't eat and kept staring at the plate, concentrating on my breath waiting for the time to leave. [S4]

Not leaving social dining situations not to upset others often resulted in getting sick: “The problem is, as I would get sick because of sensory stimuli when I am accompanied it is more difficult to leave the place because I would think that I don't want to disturb the person, that is, I would try my best to stay, even if I am feeling sick” [D8]; “If I am forced to eat somewhere noisy, with lots of light and social exposure, I usually become very sick afterward or not able to eat” [D3].

Meltdowns happened during or after social dining situations too stressful to bear anymore. One participant described stressful aspects of social dining that result in meltdowns:

I hate talking while I eat…I feel that because of all the bullying I've been through, people keep judging the way I usually eat, which is sometimes a little flustered (when I am forced to eat out in a stressful environment I end up eating from anxiety) and as my mouth is small, I end up dropping the food and being around people stresses me even more, … The noise of people swallowing and chewing stresses me…, so I often have a meltdown when I get home from these social situations. [D3]

Theme 3: Avoidance of social dining

Many participants decided to avoid joining social dining after experiencing repeated negative consequences from the lack of freedom to use coping strategies and realizing social dining is not worthy if it causes such suffering. For them, being able to eat in a peaceful space was worth more than socializing during a mealtime. A participant's response well summarizes many participants: “Lately, I have avoided situations that make me have a meltdown as much as possible—practically all places that are not part of my routine. I just got tired of forcing myself to stay in places and feel sick…I take food, but most of the time I feel embarrassed to eat in front of others, I look for a more isolated place. My strategy now is not to go” [D8].

To avoid stressful situations, some gave up joining gatherings in eating places, although they wanted to socialize with their friends. For example, a participant decided to decline an invitation after checking the pub website: “It was small, dark, and with tables very close together. Bars like that usually have music too. It would be on a Saturday night. … I would freak out. I failed to reunite with the friends I saw last year because I felt it would be too much of a problem for me…” [S3].

Some participants ate alone at work to avoid talking with colleagues. A participant ate boxed lunch (marmita) alone in her consultation office, not in the kitchen/dining area at work. This required extra effort and inconvenience, but it was worth it: “I want to be alone in silence and calm. This is fundamental to me” [S1]. For a participant who noticed eating alone at work was considered abnormal, her negotiated nonstandard work schedule was a “perfect excuse to eat alone at [her] table” ahead of others [S3].

Discussion

We explored challenges experienced by autistic adults in eating out with others. The purpose was to understand why autistic adults prefer to eat alone despite their desire to socialize. Our thematic analysis showed that dining companions often made it difficult for participants to exercise their freedom of choice to eat comfortably, which was crucial for the participants' mental and physical well-being. Freedom of choice here is the freedom to pursue one's own interests in the environment.²⁰

Autistic participants' preferences to eat alone arose from incongruences between them and their environments (see Fig. 1 for a thematic diagram). There were two levels of incongruences. The first level of incongruences was between the participants' autistic personal characteristics, most notably sensory sensitivity, and the physical environment, which included both foods and ambient environments. The second level of incongruences was between their coping strategies, which they developed to deal with the first level of congruences, and social contexts with codiners. In the social space created by codiners, where social norms and the codiners' preferences and judgmental attitudes tend to be dominating, the participants were hindered from freely using coping strategies to eat comfortably. Proshansky²⁰ had aptly stated the influence of the presence of others: “[t]he mere presence of other people may reduce freedom of choice if the individual either cannot or will not carry on particular activities in the presence of others” (p. 175). Many participants felt uncomfortable, sick, or had meltdowns during or after social dining for a social space where they found it difficult to control the environment. To prevent these negative experiences, many participants avoided eating out with others.

FIG. 1.

A thematic diagram with the overarching theme and themes in bold typeface.

Our finding that dining companions often hindered comfortable eating is in line with findings from previous research on autistic adults' eating⁵ and food environment.¹³ The first author¹³ had suggested that “‘unsafe’ people who enforce unwanted social interactions or diets” along with the other three characteristics of invisible food deserts (p. iv) make it difficult for autistic adults to access food. Our study provides a more detailed picture of “unsafe” dining companions, and how they may influence autistic persons' freedom to negotiate food environments of limited places with eatable foods and prevalent overstimulating places. We further examined challenges in eating out with others and how they are related to challenges in using coping strategies to eat out comfortably, in finding “eatable” foods, and in avoiding overstimulating places. Unsafe dining companions reacted negatively to participants for deviating from social norms or for being impolite, although the deviation was participants' effort to accommodate themselves in stressful situations. Participants' struggles in adhering to social norms were similar to those found in other studies on autistic persons.²¹ Unsafe people were also unresponsive to autistic people's needs for freedom of choice, even when the needs should be noticed with obvious signs of distress in the autistic companions; autistic differences were often not respected or accommodated.

The fact that this study included only college-educated participants might limit generalizations of our findings. Nevertheless, the responses of these Brazilian participants, who are under-represented in autism research, are in line with those of participants in autism studies from developed countries.

This study suggests aloneness in autism may result partly from interplays between personal autistic characteristics, overstimulating environments, and social contexts created by other (neurotypical) codiners. To use the words of Bronfenbrenner's bioecological model,²² there were enduring patterns of interactions between autistic participants and their nonresponsive environments to the needs for freedom of choice to seek comfort. The physical environment had been challenging participants' sensory regulation processes. The social environment had been discouraging and suppressing their coping behaviors or strategies to deal with the challenges, judging and pathologizing those “symptoms” of autism. This sociophysical environment might be “causing” unwanted aloneness.

Our results have implications for the methodologies of further studies on autistic people's eating behavior. Most research on this topic has framed autistic people's eating behavior as a problem at the individual level. But their eating behaviors were often decontextualized in the studies. Human beings are actively situated in their sociophysical environment and, thus, the study of human perceptions, minds, and behaviors requires contextualization.^23–25 In our study, we understood human behavior as a function of the situation, where the situation includes both the person and the person's environment.^24,26 We have accessed autistic subjectivities situated in the eating environment through the voices of autistic persons. This methodology allowed us to better understand the contexts of the challenges in eating with others and the preference to eat alone in autistic adults. From the perspectives of the participants, eating out alone, rather than a problematic eating behavior to be discouraged, was an active choice with the goal to eat comfortably despite the potentially stressful environment.

Finally, this study has implications for autism-friendly eating spaces. The findings suggest that eating in a peaceful, not stressful, space is critical for autistic adults' health. Stressful mealtime may cause gastrointestinal conditions. Such conditions are common among autistic people.²⁷ An autism-friendly eating space would be a peaceful eating space, which comprises (1) the autistic person, (2) eatable foods for the person, (3) a physical environment without overstimulation, and, optionally, (4) “safe” dining companions, who would not limit freedom of choice of autistic people for comfortable eating. “Unsafe” dining companions are not part of peaceful eating space. The fourth optional component of a peaceful eating space reflects our finding that eating with others is not always a healthy option for participants.

Conclusions

Our study suggests that autistic adults' preference to eat alone reflects their desire for freedom of choice to eat in a peaceful space. That freedom included the freedom to use coping strategies to eat comfortably, unconstrained by social contexts with codiners. Creating a peaceful eating environment may be a feasible and humanistic approach to reduce unwanted aloneness and social isolation in autistic adults. This study showed the potential of understanding autistic subjectivity situated in the environment in understanding autistic differences and the concept of an autism-friendly environment. It would be important to further study areas of freedom critical for autistic individuals, from their perspectives, to produce autism-friendly space, and reduce the alienation of this population.

Authors' Contributions

J.P.-C. and A.P.S.d.S. conceived the study. J.P.-C. designed protocols for data collection in Portuguese and A.P.S.d.S. revised them. J.P.-C. collected and analyzed interview responses and A.P.S.d.S. supervised the procedures. J.P.-C. drafted this article and A.P.S.d.S. participated in revising it. Both authors have reviewed and approved the article before submission.

Disclaimer

This article has been submitted solely to this journal and is not published, in press, or submitted elsewhere.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This study was financed, in part, by the Programa Nacional de Cooperação Acadêmica (PROCAD, National Program of Academic Cooperation) of the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, Coordination for the Improvement of Higher Education Personnel)—Brazil (PROCAD/CAPES)—Finance Code 001 (Grant No. 88887.124148/2014-00).

Supplementary Material

Supplementary Appendix SA1