Autism in Adults: A Survey - Part 4: Reaction to Eye Contact

This is part 4 of a series on a survey I did in 2012. To get background information, go to part 1. To read an analysis of executive dysfunction in the sample, go to part 2. To read an analysis of alexithymia in autism, go to part 3. To read about autistic internal experiences, see part 5. To read about functional language, go to part 6, and lastly, to read about independent living skills, go to part 7.

In this section, I discuss one of the two new measures used in this study - an assessment of autistic people's Reaction to Eye Contact on a 49-item, five-point scale. I also provided an 'unsure' option, which people used 0-26 times (mean 4.5+/-5.433). Women tended to choose the 'unsure' option more than men (p = .040), and had more variability in number of unsure items (p = .017). The 12 men had an average of 2.67+/-1.875 unsure items, while women had an average of 6.15+/-6.738 unsure items.

First, I decided to see which items correlated with AQ total, to see if I could come up with an autism scale. I found that items 1, 2, 4, 9, 10, 12, 13, 17, 19, 24, 25, 28, 30, 34, 35, 37, 40, 41 and 42 significantly correlated with AQ total. For items 24, 30, 34 and 35, the correlation was negative, ranging from r = -.404 to -.633 (p = .024 to <.001). The rest of the items were positively correlated, ranging from r = .366-.591 (p = .044 to <.001).

The items linked to AQ described individuals who avoid eye contact in multiple situations, find eye contact distracting, unpleasant or anxiety provoking, have difficulty reading emotions from eyes or recognizing faces, don't understand why most people find eye contact important, look at people's mouths to boost understanding of speech (ie, lipreading), prefer looking at things in peripheral vision, notice small details in a person's face, try to fake eye contact by looking elsewhere on the face, and are not entirely sure how much eye contact they should be making.

By reverse-scoring the negatively correlated items and summing the scores of all 19 items linked to the AQ, I came up with a Reaction to Eye Contact autism scale. Possible scores on this scale range from 0-76 (each item scored 0-4), although in my sample, scores ranged from 6-69. The mean score was 54+/-15.759, suggesting that most of my sample identified with the majority of the AQ-correlated items on this scale. However, without a normal sample, I can't identify a cut-off point on this scale.

Next, I performed a factor analysis. The scree suggested 5 factors, which I extracted. Many items significantly loaded on multiple factors, but I assigned each to the factor it loaded most strongly on. There was only one tie - item 39. I then derived scores by reverse-scoring negatively loaded items and summing all items together.

Factor 1 included, in order of strength of correlation, items 24 (-.969), 8 (.933), 7 (.921), 34 (-.868), 28 (.846), 1 (.845), 16 (.822), 44 (.822), 13 (.794), 36 (.765), 32 (-.758), 15 (.725), 40 (.722), 9 (.718), 43 (.718), 33 (-.684), 37 (.679), 19 (-.659), 31 (.626), 38 (.626) and 39 (-.626). Item 39 cross-loaded on both factor 1 and factor 3, and was not scored for either.

In general, factor 1 described individuals who get less benefit from eye contact, find eye contact unpleasant and benefit from lipreading. Unfortunately, these individuals also tended to have increased their rate of eye contact over time (items 36 and 32), to report that someone had taught them to make eye contact (43), to think eye contact would make them seem more honest (19) and to get unjustly accused of lying (31). So factor 1 generally describes a tendency to find eye contact pointless and unpleasant but feel pressured to make eye contact anyway.

The average score on factor 1 (not including item 39) was 51.13+/-12.896, with individuals scoring between 2-75 (possible scores 0-84). Factor 1 was significantly positively correlated with AQ total (r = .522, p = .038), AQ social (r = .464, p = .034) and AQ communication (r = .646, p = .001), suggesting that this factor was strongly linked with the person's level of autistic-style social and communication difficulties.

This factor also correlated with FrSBe total (r = .692, p = .001), FrSBe apathy (r = .514, p = .017) and FrSBe executive dysfunction (r = .680, p = .001), suggesting that it was correlated with difficulties with motivation and organizational skills. In addition, there was a significant correlation between this factor and TAS total (r = .484, p = .026), TAS difficulty describing (r = .727, p < .001) and TAS external (r = .744, p < .001). This suggests that high scorers on factor 1 tend to have a lot more difficulty describing their emotions, and a strong tendency to discount the importance of emotions. Factor 1 was also negatively correlated with age (r = -.537, p = .008), which mirrors the individuals' self-assessment that their eye contact has increased over time.

Factor 2 included items 6 (.961), 49 (.926), 22 (.894), 5 (.886), 26 (.838), 29 (.787), 2 (.699) and 3 (.663). In general, this factor described individuals who found eye contact extremely overloading and distressing, to such a degree that even being looked at by someone else made them nervous (5). Individuals who score high on this factor also had difficulty telling eyes from the rest of the face (26), suggesting that they either have significant baseline visual processing problems, or else find eye contact severely disrupts visual processing. Tragically, these individuals were also likely to have had an adult grab their face when they were a child, in order to physically force them to make eye contact (29).

The average score on factor 2 was 16.85+/-7.841, with a range from 3-30 (possible scores 0-32). This factor was correlated with FrSBe total (r = .735, p = .006), disinhibition (r = .769, p = .003) and executive dysfunction (r = .664, p = .018), suggesting a link between this factor and disorganization and poor self-control. It was also positively correlated with TAS difficulty describing (r = .608, p = .036), suggesting that high scorers also struggled to describe emotions.

Interestingly, one item that loaded on this factor, turning away while greeting others (6), is a commonly described behavioural trait in Fragile X Syndrome. Individuals with Fragile X Syndrome tend to avoid eye contact, have increased electrodermal response to sensory stimulation, and have high levels of social anxiety. This condition also overlaps with autism. It may be that both Fragile X Syndrome individuals and autistics who score high on factor 2 share a common physiological underpinning to their avoidance of eye contact, with both groups being prone to overload and anxiety due to increased physiological reactivity to eye contact and other stimuli.

Factor 3 included items 27 (.967), 30 (.948), 18 (.840), 42 (-.738), 23 (.730), 17 (-.630), 39 (-.626), 10 (-.615) and 45 (-.596). In general, this factor described individuals who made a lot of eye contact, sometimes excessive eye contact (23), and could successfully gather information from eyes, such as remembering people's eye color (27) and using eye contact to boost facial recognition (18). However, they did not notice irrelevant facial details (42), or changes in pupils size (39), suggesting that they paid attention mainly to the most socially relevant aspects of eye contact. In general, factor 3 reflected individuals who found eye contact useful and not distressing, and tended to make a great deal of eye contact as a result.

The average score on factor 3 was 8.92+/-6.887, with a range of 0-30 (possible range 0-36). Factor 3 and factor 1 were significantly negatively correlated (r = -.502, p = .029), the only significant correlation between any of the factors. This factor also negatively correlated with AQ total (r = -.626, p = .003), AQ social (r = -.468, p = .018), AQ communication (r = -.514, p = .007) and AQ imagination (r = -.587, p = .002). This suggests that this factor may reflect more normative reactions to eye contact.

This factor also negatively correlated with FrSBe executive dysfunction (r = -.460, p = .024), suggesting that this factor was linked to better organizational skills. In addition, it was negatively correlated with TAS difficulty describing (r = -.514, p = .009) and TAS external (r = -.454, p = .022), suggesting that high scorers on this scale viewed emotions as more important and had an easier time describing emotions.

Factor 4 included items 20 (.792), 18 (-.788), 48 (.780), 46 (.763), 41 (-.725), 14 (.670) and 47 (.645). This factor is a bit hard to interpret. It includes items regarding beneficial effects of eye contact on attention (47) and a tendency to make eye contact while listening (20). These individuals were less likely to close their eyes when eye contact is expected (41), but avoided looking at eyes in pictures or their reflection (46). They also tended to get greeted by people they didn't recognize (48), which hints at facial recognition problems. This factor seems to reflect individuals who show attention with eye contact, have highly situational fluctuation in eye contact, and have difficulty recognizing faces.

The average score on factor 4 was 13.69+/-4.277, with a range of 2-24 (possible range 0-28). This factor was significantly negatively correlated with FrSBe apathy, suggesting that high scorers had an easier time motivating themselves. The two non-white subjects with scores on this factor scored significantly (p = .017) higher than the white subjects, with an average of 20.50+/-4.950 vs 13.19+/-3.863. However, it's difficult to draw conclusions from only two subjects.

Factor 5 had items 12 (.910), 11 (.872), 21 (-.872), 25 (.817) and 4 (-.802). This factor generally consisted of items suggesting that eye contact is not instinctive for the individual, such as items about not noticing whether or not they're making eye contact (11), deliberately faking eye contact (12), and preferring peripheral vision (25). However, having to consciously think about eye contact (21) and being unsure how much eye contact they should make (4) both loaded negatively on this factor.

The average score on factor 5 was 7.09+/-3.025, with a range of 0-12 (possible range 0-20). This factor significantly positively correlated with FrSBe total (r = .376, p = .049) and executive dysfunction (r = .453, p = .012), suggesting that high scorers on this factor had poorer organizational skills.

Probably the saddest part of these results was the results from items 29 and 43. A little over half (53.3%) of the sample responded strongly or slightly agree to "At some point in my life, someone deliberately taught me to make eye contact." Worse, about a third (37.7%) of the respondents reported that "When I was younger, adults used to grab my face to force me to make eye contact."

These two items were both linked to factors that reflected adverse reactions to eye contact (factors 1 and 3), with the more severely intrusive item being linked to the factor with more severe reaction to eye contact (factor 3). This probably reflects a reaction by others to the lower eye contact among high scorers on these two factors, but it's very unfortunate, because these individuals are the ones most likely to suffer harm as a result of being pressured or forced to make eye contact.

It's interesting and concerning to note that these two factors were also correlated with higher alexithymia. Research has found higher rates of alexithymia in abuse survivors, with links between alexithymia and parenting practices that discourage authentic emotional expression (which are common in abusive families). Alexithymia has also been found to moderate links between abuse and depression, suicide attempts, eating disorders and self-injury.

Autistic individuals are also at higher risk of depression, suicide attempts, anorexia and self-injury, and one study suggested that positive parenting can reduce self-injury in autism. I have long speculated that alexithymia might not be inherent to autistic people, but rather an effect of growing up with highly atypical emotional reactions that others do not understand or acknowledge the way they do for most children.

I hypothesize that the link between alexithymia and factors 1 and 3 of the Reaction to Eye Contact scale might be at least partially explained by the pressure to make eye contact. For someone who finds eye contact unpleasant and disruptive to functioning, pressure to make eye contact represents pressure to ignore your own needs and desires for the sake of someone else. For someone who finds eye contact severely overloading, to be grabbed and forced to make eye contact, until you learn not to resist anymore, must be extremely frightening and traumatizing. For both these groups, learning to conform requires that they learn to do something unpleasant and harmful to themselves for no other reason than to appease someone else - possibly even your own parents. How do you stay true to your emotions when faced with a demand like that?

Of course, many of my participants likely weren't diagnosed as children. Discussions of going through the diagnostic process in adulthood are common on WrongPlanet forums, and given the age range of my sample (17 to 60 years), many of them grew up in time periods when autism was believed to be extremely rare, and the diagnostic criteria was far more restrictive. But if they had been diagnosed in childhood, would their lot have been any better?

The best-supported treatment program for autistic children is ABA, and a search of "autism eye contact applied behavioural analysis" in Google Scholar readily brings up examples such as this study, this study, this study, this study, this study and this review, in which autistic children are systematically trained to show eye contact, without any apparent consideration to the adverse effects that eye contact may have on the child. The fourth study even used physical force as a potential punishment (the child was punished by being ordered to do a series of head movements, and was physically forced if he didn't comply).

Unfortunately, pressuring to make eye contact is something that many early-diagnosed children have experienced, and are still experiencing. Despite being 'experts' in autism, the people involved in treatment often don't seem to consider whether or not eye contact is adaptive for an autistic child. Instead, eye contact is presumed to be essential, and is encouraged through rewards, ignoring communicative behaviour without eye contact, forcing the child to do a series of pointless head movements on command if they don't make eye contact, and so forth.

Given the reports from autistic adults describing highly negative consequences of making eye contact, and the possibility that alexithymia and many serious psychiatric disorders can result from childhood experiences in which a child is encouraged to ignore negative emotions, I have serious concerns about the implications of forcing an autistic child to make eye contact.

Appendix: Reaction to Eye Contact items and responses:

Item	Strongly agree	Slightly agree	Neutral	Slightly disagree	Strongly disagree	Unsure
1. I make much less eye contact when interacting than most people do.	27 60.0%	7 15.6%	2 4.4%	1 2.2%	1 2.2%	7 15.6%	Total: 45
2. People's eyes often distract me from what I am saying or doing.	15 34.1%	18 40.9%	2 4.5%	4 9.1%	4 9.1%	1 2.3%	Total: 44
3. Most people's eyes are continually moving.	2 4.4%	6 13.3%	6 13.3%	2 4.4%	1 2.2%	28 62.2%	Total: 45
4. I'm often unsure how much eye contact I should be making.	31 68.9%	9 20.0%	3 6.7%	1 2.2%	0 0.0%	1 2.2%	Total: 45
5. Even if I don't look at them, someone looking directly at me makes me nervous.	20 44.4%	15 33.3%	2 4.4%	6 13.3%	1 2.2%	1 2.2%	Total: 45
6. I tend to turn away when I greet people.	8 17.8%	16 35.6%	6 13.3%	5 11.1%	8 17.8%	2 4.4%	Total: 45
7. When interacting with others, I often pay more attention to what I can see going on behind them than I do to their face.	20 44.4%	13 28.9%	5 11.1%	4 8.9%	2 4.4%	1 2.2%	Total: 45
8. I find it difficult to recognize faces.	14 31.1%	17 37.8%	5 11.1%	3 6.7%	6 13.3%	0 0.0%	Total: 45
9. Eye contact makes me anxious.	25 55.6%	13 28.9%	2 4.4%	1 2.2%	3 6.7%	1 2.2%	Total: 45
10. I often look at a person's mouth when they are talking.	14 32.6%	17 39.5%	4 9.3%	2 4.7%	5 11.6%	1 2.3%	Total: 43
11. I don't really notice if I am making eye contact or not.	4 9.1%	10 22.7%	4 9.1%	10 22.7%	14 31.8%	2 4.5%	Total: 44
12. I have various tricks to make people think I am making eye contact when I'm not.	12 26.7%	13 28.9%	4 8.9%	3 6.7%	10 22.2%	3 6.7%	Total: 45
13. I find the sight of people's eyes unpleasant.	5 11.1%	11 24.4%	10 22.2%	8 17.8%	9 20.0%	2 4.4%	Total: 45
14. I find it harder to make eye contact with someone who is angry at me.	27 60.0%	6 13.3%	3 6.7%	3 6.7%	4 8.9%	2 4.4%	Total: 45
15. I often don't know in what direction someone else is looking.	10 22.2%	10 22.2%	8 17.8%	6 13.3%	6 13.3%	5 11.1%	Total: 45
16. I force myself to make eye contact even when I don't want to.	14 31.8%	14 31.8%	3 6.8%	4 9.1%	8 18.2%	1 2.3%	Total: 44
17. I usually try to avoid looking at other people while I'm walking.	25 56.8%	10 22.7%	5 11.4%	3 6.8%	1 2.3%	0 0.0%	Total: 44
18. I find it easier to make eye contact with someone I know.	18 40.0%	16 35.6%	4 8.9%	1 2.2%	4 8.9%	2 4.4%	Total: 45
19. I try to make eye contact so I will seem more honest.	10 22.2%	11 24.4%	8 17.8%	4 8.9%	11 24.4%	1 2.2%	Total: 45
20. I make more eye contact when I'm listening to the other person talk than when I'm talking.	13 28.9%	8 17.8%	5 11.1%	2 4.4%	8 17.8%	9 20.0%	Total: 45
21. I have to deliberately think about whether or not I'm making eye contact.	22 48.9%	12 26.7%	5 11.1%	1 2.2%	1 2.2%	4 8.9%	Total: 45
22. I find it hard to think about other things when someone is looking into my eyes.	24 53.3%	10 22.2%	2 4.4%	4 8.9%	1 2.2%	4 8.9%	Total: 45
23. I tend to make eye contact too long or too intensely.	4 8.9%	4 8.9%	3 6.7%	6 13.3%	16 35.6%	12 26.7%	Total: 45
24. I can tell a lot about how someone is feeling from their eyes.	1 2.3%	2 4.5%	0 0.0%	13 29.5%	24 54.5%	4 9.1%	Total: 44
25. I prefer to look at things in my peripheral vision rather than directly.	9 20.5%	12 27.3%	8 18.2%	7 15.9%	3 6.8%	5 11.4%	Total: 44
26. I often can't tell someone's eyes from the rest of their face.	2 4.4%	3 6.7%	2 4.4%	8 17.8%	18 40.0%	12 26.7%	Total: 45
27. If you asked me about someone I knew, I could probably tell you what color their eyes were.	1 2.3%	8 18.2%	3 6.8%	13 29.5%	15 34.1%	4 9.1%	Total: 44
28. I don't know why people think eye contact is so important.	24 53.3%	10 22.2%	3 6.7%	6 13.3%	2 4.4%	0 0.0%	Total: 45
29. When I was younger, adults used to grab my face to force me to make eye contact.	11 24.4%	6 13.3%	5 11.1%	4 8.9%	17 37.8%	2 4.4%	Total: 45
30. I often make eye contact with people I know across a crowded room, when they are too far away to interact with directly.	4 8.9%	9 20.0%	2 4.4%	7 15.6%	20 44.4%	3 6.7%	Total: 45
31. I often get accused of lying when I'm telling the truth.	12 26.7%	9 20.0%	6 13.3%	9 20.0%	8 17.8%	1 2.2%	Total: 45
32. I used to make more eye contact than I do now.	1 2.3%	8 18.2%	6 13.6%	6 13.6%	15 34.1%	8 18.2%	Total: 44
33. If I'm not looking at the other person while they are talking, I'm probably not listening either.	1 2.3%	2 4.5%	2 4.5%	4 9.1%	33 75.0%	2 4.5%	Total: 44
34. I enjoy looking at people's eyes.	1 2.3%	2 4.5%	6 13.6%	6 13.6%	25 56.8%	4 9.1%	Total: 44
35. I find it easier to recognize who a person is if I look them in the eye.	2 4.5%	3 6.8%	3 6.8%	8 18.2%	24 54.5%	4 9.1%	Total: 44
36. I make more eye contact now than I used to.	8 17.8%	13 28.9%	2 4.4%	6 13.3%	9 20.0%	7 15.6%	Total: 45
37. I find it easier to understand what someone is saying if I watch the way their mouth moves.	10 22.2%	16 35.6%	6 13.3%	1 2.2%	8 17.8%	4 8.9%	Total: 45
38. I like looking at my reflection in people's eyes or glasses.	0 0.0%	3 6.7%	9 20.0%	8 17.8%	23 51.1%	2 4.4%	Total: 45
39. I have noticed people's pupils changing sizes while I'm looking at them.	4 9.1%	3 6.8%	0 0.0%	6 13.6%	23 52.3%	8 18.2%	Total: 44
40. I can look or listen, but not both at once.	12 27.3%	18 40.9%	2 4.5%	5 11.4%	4 9.1%	3 6.8%	Total: 44
41. I sometimes close my eyes when others are trying to make eye contact with me.	4 9.3%	7 16.3%	5 11.6%	11 25.6%	14 32.6%	2 4.7%	Total: 43
42. When I look at a person's face, I notice every little detail (eg freckles, blackheads, pores, etc).	13 28.9%	15 33.3%	1 2.2%	4 8.9%	9 20.0%	3 6.7%	Total: 45
43. At some point in my life, someone deliberately taught me to make eye contact.	14 31.1%	10 22.2%	4 8.9%	4 8.9%	11 24.4%	2 4.4%	Total: 45
44. I tend to look away from a person's face when I'm thinking.	35 77.8%	7 15.6%	0 0.0%	1 2.2%	1 2.2%	1 2.2%	Total: 45
45. I find it painful to make direct eye contact.	13 28.9%	15 33.3%	4 8.9%	4 8.9%	6 13.3%	3 6.7%	Total: 45
46. I avoid looking at the eyes of pictures or my own reflection.	9 20.5%	5 11.4%	4 9.1%	12 27.3%	9 20.5%	5 11.4%	Total: 44
47. I find it easier to pay attention if I make eye contact.	1 2.2%	1 2.2%	3 6.7%	9 20.0%	28 62.2%	3 6.7%	Total: 45
48. I often find people come up and act like they know me when I don't recognize them.	10 22.2%	22 48.9%	1 2.2%	4 8.9%	7 15.6%	1 2.2%	Total: 45
49. When someone looks into my eyes, I feel exposed, like they can see into my soul.	17 37.8%	12 26.7%	6 13.3%	4 8.9%	4 8.9%	2 4.4%	Total: 45