- Email: [email protected]
Gender differences of exploratory eye movements
Life Sciences 68 (2000) 569–577
Gender differences of exploratory eye movements A life span study Ayako Miyahiraa,*, Kiichiro Moritaa, Hiroshi Yamaguchia, Kensaku Nonakab, Hisao Maedaa a
Department of Neuropsychiatry, Kurume University School of Medicine, 67, Asahi-machi, Kurume-shi, Fukuoka-ken, Japan b Horikawa Hospital, 510, Nishi-machi, Kurume-shi, Fukuoka-ken, Japan Received 12 October 1999; accepted 22 February 2000
Abstract Exploratory eye movements of normal subjects (39 male and 39 female) were recorded using an eye-mark recorder. Four parameters mean gazing time, total number of gazing points, mean scanning length, and total scanning length) were analyzed. Subjects were divided into three life spans as prepuberty (boys and girls), adult, and postpuberty. The mean gazing time of adult women was significantly longer than that of age-matched adult men, but not between men and women in prepuberty or postpuberty (postmenopausal older women and age-matched older men). The total number of gazing points of women was significantly smaller than that of men, but not significantly different between men and women in both prepuberty and postpuberty. Both the mean scanning length and total scanning length of adult women were shorter than those of age matched adult men, but no significant differences were found between men and women in both prepuberty and postpuberty. Furthermore, the mean gazing time of adult women was longer than that of men in prepuberty and postpuberty. The total scanning length of adult men was longer than that of women in both prepuberty and postpuberty. These findings suggest that gender differences of exploratory eye movements are observed only during the adult phase, which indicates that visual information processing may be regulated by gonadal hormones in humans. © 2000 Elsevier Science Inc. All rights reserved. Keywords: Exploratory eye movement; Gender difference; Life span; Visual cognition
Introduction Exploratory eye movements have been evaluated as useful biologic markers to investigate the mechanisms of visual information processing in healthy subjects and in patients who have * Corresponding author: Tel.: 0942-31-7564; fax: 0942-35-6041. E-mail address: [email protected] (A. Miyahira) 0024-3205/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 0 2 4 - 3 2 0 5 ( 0 0 )0 0 9 6 3 -2
570
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
cognitive disorders such as schizophrenia, depression, or dementia (1,2,3,4). Kojima et al (2) reported in detail on the exploratory eye movements in normal subjects and schizophrenic patients and suggested that these movements may be trait markers for schizophrenia. It is important to evaluate gender and age differences in normal subjects. Indeed, anatomical gender differences have been widely reported in the brain (2,5). A consistent finding is that men perform better than women in spatial , motor (6,7) and language tasks (8). Many reports indicate that in performance tasks men had consistently faster reaction times than did women (9), and women were reported to have a poorer performance in scanning the visual field (10). Thus, it is important that gender differences in exploratory eye movements be characterized in normal subjects. However, no such studies have been done. It has been suggested that the gender differences of ERP (Event related potentials) on visual information processing might involve the gonadal hormones (11). The present experiments examine the gender differences of exploratory eye movements during the life span of children (prepuberty), adult, and the elderly (postpuberty). Methods Subjects Subjects were 78 normal volunteers consisting of doctors, nursing staff and their children (39 male and 39 female).All subjects were right handed, had normal vision, and had no history of psychiatric or neurological diseases or drug addiction. All subjects were divided into three life spans as prepuberty (11 boys and 11 girls: mean age 6 SD; 7.46 6 2.70 and 8.40 6 1.89, respectively), adult (18 men and 18 women; 30.33 6 .85 and 28.67 6 9.15 respectively) and postpuberty (10 men and 10 women; 59.10 6 6.09 and 63.80 6 6.98 respectively). No significant age differences were observed in each life span. Women were not tested at any particular time during the menstrual period. All subjects provided informed consent for their participation. Eye mark recording Eye movements were recorded using an eye mark recorder (Nac, SK-2, Osaka, Japan) which consisted of two very small video cameras (left-and right-eye mark shooting units) fixed on the left and right sides of a head band and another camera (field shooting unit) fixed on the top of the helmet. The infrared light sources are positioned in front of each lower eyelid. The side cameras record the reflection of the infrared light from the cornea of the eye. The camera on the top records the pictures on the screen. These reflections are processed by a camera controller with a 1/100 seconds electronic timer which is then recorded on a video tape recording system. A movement of more than 18 and a duration of more than 0.2 seconds is scored as an eye movement. The gazing points were determined from a gazing time exceeding 0.2 seconds as real points in this study, because Poulton (12) concluded from his study on the relationship between eye movements and visual cognition that it required more than 0.2 seconds for the information processing system in the brain to deal with an image on the retina. Moreover, the gazing point was determined within 18 of visual angle(2,3). The instrument can detect movements as small as 0.38 of sight. This technique enables us to observe simultaneously the eye fixation points and eye movements on the figure. Recordings were
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
571
obtained between 9 AM and 12 noon. The recorded data were analyzed by a slow-motion replay of the video recorder. Slow-motion videotape replay enabled visualization of the sequencing of eye movements. Eye blink artifact has a characteristic vertical component and is easily eliminated. In the present study, the gazing points outside of the screen were omitted. The recorded data were analyzed by means of a computerized analyzing system (11). The exploratory eye movements were analyzed for four parameters: mean gazing time, total number of gazing points, mean scanning length of gazing points, and total scanning length of gazing points(2,3). The eye scanning length is the distance between two gazing points. Eye movement recording procedure In a dark room where non-visual sensory stimuli were attenuated, eye movements were recorded using an eye-mark recorder. Before the experiment, subjects were asked to draw each presented picture immediately after viewing in order to increase their visual attention. Exploratory eye movements and fixation points during perception of pictures were recorded. The pictures were projected onto a screen where they appeared 90 cm wide and 70 cm tall, and maximum angles of sight lines were 308 horizontally and 208 vertically. Each session consisted of a series of six views; subjects were required to view six pictures for 15 seconds each except for picture 3* (10 seconds) (2,3,13). Four kinds of pictures were used as shown in Fig. 1. Picture 1, an open circle which was simple and non-stressing but examined driving and/or motivation of subjects. Picture 2, a “happy face” which examined the effect of emotional influences. Picture 3, a “happy face” with lines added beside the mouth, which examined recognition of the differences. Picture 3*, identical to picture 3 but subjects were asked to search for any additional differences, which represented a test of confidence (2,4). Picture 4 (a scene) with 10 elements including four different animals, the sun, an airplane, five trees, a house, two mountains, and a river, which examined the subjects for visual ability and short term memory. Picture 5 (an open circle), identical to picture 1, which was examined to detect any influence of the stimulus of picture 4 and any habituation. For picture 4, all subjects were required to note all elements immediately after viewing. We evaluated the total number of
Fig. 1. Test pictures and comparison of two typical series of exploratory eye movements (adult men and women).
572
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
identified elements as element scores (the maximum possible score was 13 points) to check the subjects’ arousal levels. Data were analyzed from the left eye in this study. There are no significant differences in movements between the right and left eyes (unpublished observation). Statistical analysis For data considered reliable (epsilon , 1.0), a one-way ANOVA was used to compute the main gender effect (men and women). Post-hoc analyses were conducted using Fishers protected least square deviation (LSD) (13). Next, a two-way ANOVA was used to examine the interaction (sex 3 picture). A level of p, .05 was accepted as statistically significant. Results Mean gazing time The mean gazing time of girls (prepuberty) was 0.37960.089 (n511, mean6standard deviation, SD) and that of age-matched boys was 0.37960.079 (n511). For adult women it was 0.45360.113 and 0.39360.059 for age- matched adult men, and 0.40960.105 for postpuberty women and 0.39060.093 for age-matched men (see Table 1). In the adult, the mean gazing time of women was significantly longer than that of men by two-way ANOVA (sex 3 Table 1 Gender differences in the eye movement of three life spans
Statistical Significance was determined between men and women on four measurements (mean gazing time, total number of gazing points, mean scanning length of gazing points, and total scanning length). Statistical significance as calculated by Fisher F test is indicated by asterisks. The life span differences in each sex are indicated by a, aa, aaa. The significant differences between prepuberty, adult and/or postpuberty are indicated by b,bb bbb.
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
573
Fig. 2. Gender differences in mean gazing time of three life spans. A. The mean gazing time of women was significantly slower than that of men only in the adult group, as is indicated by the closed stars. Significant differences were observed in women between prepuberty and adult, and adult and postpuberty, as indicated by the open stars. B. Picture 1 (open circle), picture 4 (scene) and picture 5 (open circle identical picture 1) were significantly different between adult men and women.
picture) [F(1,204)523.88, p, .0001). Significant differences between adult women and agematched men were found with three pictures- the first open circle (picture 1; F[1,34]55.05], p,.05), the scene (picture 4; F[1,34]57.57, p, .001), and the last open circle (picture 5; F[11,34]50.93], p,.001). The mean gazing time of adult women was significantly longer than that of prepuberty (girls) and postpuberty (older women). There were no significant differences in mean gazing time among the three life spans in men as seen in Fig. 2. Total number of gazing points The total number of gazing points of adult women were significantly less than those of men as analyzed by two-way ANOVA (sex 3 picture; F[1,204]560.52], p, .0001). There were no significant differences between boys and girls or between postpuberty women and age-matched men. For each picture, significant differences were apparent between adult men and women. Picture 1(F[1,34]515.58, p,.001), picture 2(F[1,34]58.62, p,.01), picture 3(F[1,34]55.55, p,.05), picture 4 (F[1,34]519.72, p,.001) and picture 5 (F[1,34]514.01,
574
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
Fig. 3. Gender differences in total scanning length of three life spans. A. The total scanning length of women was significantly smaller than that of men only in the adult group, as indicated by the closed stars. Significant differences were observed in men between prepuberty and adult, and adult and postpuberty in men, as indicated by the open stars. B. The differences in scanning length between adult men and women were highly significant for picture 4 (scene) and picture 5 (open circle).
p,.001) were significantly different between adult women and age matched adult men. Pictures1, 4 and 5 showed larger differences than the others. The total number of gazing points of postpubescent women was significantly larger than those of girls and adult women. The total number of gazing points of boys was significantly larger than that of adult and senior men. Mean scanning length The mean eye scanning length of adult men was significantly longer than that of agematched women as analyzed by two-way ANOVA (gender 3 picture; F[1,204]55.33, p, .05). There were no significant differences between boys and girls, nor between postpubescent women and age-matched men. No significant differences were found for individual pictures. The mean scanning lengths of postpubescent women and age-matched men were shorter than those of boys and girls, respectively. In the male, the mean scanning length of senior men was significantly shorter than that of adult men.
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
575
Total scanning length The total eye scanning length of adult men was significantly longer than that of adult women as analyzed by two-way ANOVA (gender 3 pictures; F[1,204] 549.38. p, .001). There were no significant differences between boys and girls or between postpubescent women and age-matched men. The total scanning length shows significant differences on five pictures. The difference between picture 4 (F[1,34]515.28, p, .001) and picture 5 (F[1,34]5 19.47, p,.001), was larger than for the others (picture 1, F[1,34]57.48, p, .01; picture 2, F[1,34]55.37, p, .05; picture 3,F[1,34]55.56, p,0.05). The total scanning lengths of adult men and senior men were significantly longer than those of boys. However, there were no differences among the women in each life span.
Discussion The major finding of the present study was that a gender influence on eye movement parameters was observed in healthy subjects. Significant differences were found only in the adult phase, and they were attributable to gender. The mean gazing time of adult women was consistently longer than that of men, and the scanning length of adult women was shorter than that of men. Gender differences have been shown in a variety of electro-physiologic measures. Adult women were reported to have larger amplitude brainstem and cortical evoked responses (5) and shorter EP latencies than men (15,16). Anatomical differences between men and women have been suggested to underlie differences in EP source origins. However, anatomical differences are not likely to explain the present findings because gender differences varied according to the pictures . For example, on mean gazing time, there were significant gender differences only in pictures 4 and 5 but not in the other pictures. If anatomic differences were involved in exploratory eye movement, more consistent differences would be expected irrespective of the pictures. In a sensory vigilance task, young women had slower reaction times to target stimuli than men of similar ages, and women detected fewer targets (9). The authors suggested that these effects may have reflected different arousal levels in men and women. Thus, arousal may decline during the presentation of the six pictures. However, picture 1 showed certain gender differences, and the differences in mean gazing time and total scanning length were larger than in pictures 2 and 3. These findings indicate that a decline of arousal level may not result in gender differences during sustained viewing, such as with the 6 pictures. However, we cannot completely rule out the possibility that the arousal level of women may have been lower than that of men, especially in the last two pictures. Hormonal influences have been implicated in gender differences in ERP (event related potential). In that study, which spanned many subject ages, women reportedly had a higher mean amplitude response of P50 to a first stimulus than men (17,18). The gender differences in pattern reversal EP (evoked potential) amplitude over many subject ages suggested a hormonal influence on visual spatial frequency processing (11). Gender differences in the brain’s functions of language and spatial ability are also known. We speculate that in the present study hormonal changes are the likely cause of the gender differences because they were observed only between adult women and age-matched men. Furthermore, there were no gender
576
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
differences between girls (prepuberty) and age-matched boys, and also between postpuberty women and age-matched older men. It has been reported that spatial ability in androgen deficient men is similar to that of women (19). Indeed, gonadal hormones are reported to effect EEG changes (20). Another interesting finding was the association of certain characteristics to each life span in each sex. For example, postpubescent women have a long mean gazing time, large total number of gazing points and short mean scanning length. The age-matched older men have a short mean scanning length. The mean gazing time of adult women is the longest and the total scanning length of adult men is the longest. These life span differences in each sex may also indicate the influence of gonadal hormones in eye movements. The mechanism of gonadal hormone regulation of visual information processing may be further clarified by measuring hormonal levels at all ages. It is interesting that the eye movements for the scene (picture 4) and the second open circle (picture 5) were largely or significantly different between adult men and women in the present study, suggesting that the addition of details, such as animals and a house, stimulated visual information processing differently in men and women. Additionally, as shown in picture 5, the recovery cycle from the visual stimuli may be different in men and women as suggested previously (21). In conclusion, this study evaluated the effect of gonadal hormones on visual information processing in healthy humans (2,11). Evidence was presented that men perform better than women in spatial and motor tasks (6). Indeed, It has been reported that gender differences may exist in spatial ability and in functional organization of the brain for language (22,19). In this study, gender differences were not observed in either childhood (prepuberty) or in the elderly (postpuberty), but clear differences were observed in the adult. It is possible that tasks such as viewing pictures, which require a great amount of cognitive processing, involve different cognitive strategies in men and women. The present finding may reflect gender related behavioral differences. It has been reported that men’s spatial ability may be enhanced to function in a human hunter society. As shown in Fig. 1, open circle, men would scan across the line and tend to look at the outer space, but women would scan across the line and tend to look at the inner space. It may be speculated that men concentrate on the target and thus pay attention to other parts, especially outside of the target, whereas women would also concentrate on the target but pay attention to the inside.(3) These visual differences may reflect certain social behaviors in age development. Further study is needed to explore possible visual desensitization or fatigue phenomenon during repetitive viewing, which was different in men and women. References 1. Holzman PS., Kringle E., Levy DL, Proctor LR., Haberman SJ. Yasillo N.J., Abnormal pursuit eye movement in schizophrenia: evidence for a genetic indicator. Journal Of Archives of General Psychiatry 1977; 34 : 802– 805. 2. Kojima T., Matsushima E., Nakajima K., Shiraishi H., Ando K., Shimazono Y., Eye movement in acute, chronic, and remitted Schizophrenics. Journal of Biological Psychiatry 1990; 27: 975–989. 3. Miyahira A., Morita K., Yamaguchi H., Nonaka K., Maeda H., Gender differences and reproducibility in exploratory eye movements of normal subjects. Jounal of Psychiatry and clinical neurosciences 2000; 54: 31–36. 4. Matushima E., Eye movement in schizophrenic patients. Psychiatria et Neurologia Japonica 1988; 2: 89–110.
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
577
5. Baumann SB., Rogers RL., Guinto FC., Saydjart CL., Papanicolaou AC., Eisenberg HM., Gender differences in source location for the N100 auditory evoked magnetic field. Jounal of Electroencephalography of Clincal. Neurophysiology 1991: 80: 53–59. 6. Bakan P. and Putnam W., Right-left discrimination and brain materialization : sex differences. Journal of Archives of Neurology. 1974; 30: 334–335. 7. Delacoste UC. Holloway RL., Sexual dimorphism in the human corpus callosum. Journal of Science. 1982; 216: 1413–1432. 8. Shaywitz B., Shaywitz S., Pugh K., et al. Sex differences in the functional organization of the brain for language. Jounal of Nature. 1995; 373: 607–609. 9. Giambla LM., Quilter RE, Sex differences in sustained attention across the adult life span. Journal of Applied psychology. 1989; 74(1): 91–9. 10. Efron R., Yund EW. Nichols DR., Scanning the visual field without eye movements; a sex difference. Journal of Neuropsychologia 1987; 25: 637–644. 11. Sheare DE., Emmerson RY. Dustman RE., Gender differences in pattern reversal evoked potential amplitude; Influence of check size and single trial response variability. Journal of American Journal of EEG Technology. 1992; 32: 196–203. 12. Poulton EC., Peripheral vision, refractoriness and eye movements in fast oral reading. British Journal of Psychology. 1962; 53:409–419. 13. Miyahira A., Morita K., Nonaka K., Nakamura J. Maeda H., Effects of caffeine and ATP on eye movement. Journal of Kyushu Neuropsychiatry. 1997; 43: 114–120. 14. Maeda H., Morita K., Kawamura N. Nakazawa Y., Amplitude and area of the auditory P300 recorded with eyes open reflect remission of schizophrenia. Journal of Biological Psychiatry. 1996; 39: 743–746. 15. Taylor MJ., Smith ML., Iron KS., Event-related potential evidence of sex differences in verbal and nonverbal memory tasks. Journal of Neuropsychologia. 1990; 28(7): 691–705. 16. Trune D., Mitchell C., Phillps D., The relative importance of head size, gender and age on the auditory brainstem response. Hearing Research. 1988; 32: 165–174. 17. Freedman R., Adler LE., Waldo M., Gating of the auditory evoked Potential in children and adults. Journal of Psychophysiology. 1987; 24(2): 223–227. 18. Hetrick WP., Sandman CA., Bunney WE., Jin Jr.Y., Potkin SG. White MH., Gender differences in gating of the auditory evoked potential in normal subjects. Journal of Society of biological Psychiatry. 1996; 39: 51–58. 19. Hier DB. Growley WF., Spatial ability in androgen deficient men. Journal of New England Journal of Medicine. 1982; 306 :1202–1205. 20. Vogel W., Broyerman DM., Klaiber EL., Kobayasht Y., Clarkson FE., Itil TM., Laudahn G., Herrmann WM.(Eds), A model for the integration of hormonal behavioral, EEG, and pharmacological data in psychopathology. Journal of Psychotropic Action of Hormones. Spectrum (New York), 1976; 121–134. 21. Lolas F., Inter hemispheric and sex differences in the visual evoked response recovery cycle. Journal of Neuropsychobiology. 1979; 5(6): 301–308. 22. Shaywitz BA., Shaywitz SE., Pugh KR., Sex differences in the functional organization of the brain for language. Journal of Nature. 1995; 372: 607–609.
Gender differences of exploratory eye movements A life span study Ayako Miyahiraa,*, Kiichiro Moritaa, Hiroshi Yamaguchia, Kensaku Nonakab, Hisao Maedaa a
Department of Neuropsychiatry, Kurume University School of Medicine, 67, Asahi-machi, Kurume-shi, Fukuoka-ken, Japan b Horikawa Hospital, 510, Nishi-machi, Kurume-shi, Fukuoka-ken, Japan Received 12 October 1999; accepted 22 February 2000
Abstract Exploratory eye movements of normal subjects (39 male and 39 female) were recorded using an eye-mark recorder. Four parameters mean gazing time, total number of gazing points, mean scanning length, and total scanning length) were analyzed. Subjects were divided into three life spans as prepuberty (boys and girls), adult, and postpuberty. The mean gazing time of adult women was significantly longer than that of age-matched adult men, but not between men and women in prepuberty or postpuberty (postmenopausal older women and age-matched older men). The total number of gazing points of women was significantly smaller than that of men, but not significantly different between men and women in both prepuberty and postpuberty. Both the mean scanning length and total scanning length of adult women were shorter than those of age matched adult men, but no significant differences were found between men and women in both prepuberty and postpuberty. Furthermore, the mean gazing time of adult women was longer than that of men in prepuberty and postpuberty. The total scanning length of adult men was longer than that of women in both prepuberty and postpuberty. These findings suggest that gender differences of exploratory eye movements are observed only during the adult phase, which indicates that visual information processing may be regulated by gonadal hormones in humans. © 2000 Elsevier Science Inc. All rights reserved. Keywords: Exploratory eye movement; Gender difference; Life span; Visual cognition
Introduction Exploratory eye movements have been evaluated as useful biologic markers to investigate the mechanisms of visual information processing in healthy subjects and in patients who have * Corresponding author: Tel.: 0942-31-7564; fax: 0942-35-6041. E-mail address: [email protected] (A. Miyahira) 0024-3205/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 0 2 4 - 3 2 0 5 ( 0 0 )0 0 9 6 3 -2
570
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
cognitive disorders such as schizophrenia, depression, or dementia (1,2,3,4). Kojima et al (2) reported in detail on the exploratory eye movements in normal subjects and schizophrenic patients and suggested that these movements may be trait markers for schizophrenia. It is important to evaluate gender and age differences in normal subjects. Indeed, anatomical gender differences have been widely reported in the brain (2,5). A consistent finding is that men perform better than women in spatial , motor (6,7) and language tasks (8). Many reports indicate that in performance tasks men had consistently faster reaction times than did women (9), and women were reported to have a poorer performance in scanning the visual field (10). Thus, it is important that gender differences in exploratory eye movements be characterized in normal subjects. However, no such studies have been done. It has been suggested that the gender differences of ERP (Event related potentials) on visual information processing might involve the gonadal hormones (11). The present experiments examine the gender differences of exploratory eye movements during the life span of children (prepuberty), adult, and the elderly (postpuberty). Methods Subjects Subjects were 78 normal volunteers consisting of doctors, nursing staff and their children (39 male and 39 female).All subjects were right handed, had normal vision, and had no history of psychiatric or neurological diseases or drug addiction. All subjects were divided into three life spans as prepuberty (11 boys and 11 girls: mean age 6 SD; 7.46 6 2.70 and 8.40 6 1.89, respectively), adult (18 men and 18 women; 30.33 6 .85 and 28.67 6 9.15 respectively) and postpuberty (10 men and 10 women; 59.10 6 6.09 and 63.80 6 6.98 respectively). No significant age differences were observed in each life span. Women were not tested at any particular time during the menstrual period. All subjects provided informed consent for their participation. Eye mark recording Eye movements were recorded using an eye mark recorder (Nac, SK-2, Osaka, Japan) which consisted of two very small video cameras (left-and right-eye mark shooting units) fixed on the left and right sides of a head band and another camera (field shooting unit) fixed on the top of the helmet. The infrared light sources are positioned in front of each lower eyelid. The side cameras record the reflection of the infrared light from the cornea of the eye. The camera on the top records the pictures on the screen. These reflections are processed by a camera controller with a 1/100 seconds electronic timer which is then recorded on a video tape recording system. A movement of more than 18 and a duration of more than 0.2 seconds is scored as an eye movement. The gazing points were determined from a gazing time exceeding 0.2 seconds as real points in this study, because Poulton (12) concluded from his study on the relationship between eye movements and visual cognition that it required more than 0.2 seconds for the information processing system in the brain to deal with an image on the retina. Moreover, the gazing point was determined within 18 of visual angle(2,3). The instrument can detect movements as small as 0.38 of sight. This technique enables us to observe simultaneously the eye fixation points and eye movements on the figure. Recordings were
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
571
obtained between 9 AM and 12 noon. The recorded data were analyzed by a slow-motion replay of the video recorder. Slow-motion videotape replay enabled visualization of the sequencing of eye movements. Eye blink artifact has a characteristic vertical component and is easily eliminated. In the present study, the gazing points outside of the screen were omitted. The recorded data were analyzed by means of a computerized analyzing system (11). The exploratory eye movements were analyzed for four parameters: mean gazing time, total number of gazing points, mean scanning length of gazing points, and total scanning length of gazing points(2,3). The eye scanning length is the distance between two gazing points. Eye movement recording procedure In a dark room where non-visual sensory stimuli were attenuated, eye movements were recorded using an eye-mark recorder. Before the experiment, subjects were asked to draw each presented picture immediately after viewing in order to increase their visual attention. Exploratory eye movements and fixation points during perception of pictures were recorded. The pictures were projected onto a screen where they appeared 90 cm wide and 70 cm tall, and maximum angles of sight lines were 308 horizontally and 208 vertically. Each session consisted of a series of six views; subjects were required to view six pictures for 15 seconds each except for picture 3* (10 seconds) (2,3,13). Four kinds of pictures were used as shown in Fig. 1. Picture 1, an open circle which was simple and non-stressing but examined driving and/or motivation of subjects. Picture 2, a “happy face” which examined the effect of emotional influences. Picture 3, a “happy face” with lines added beside the mouth, which examined recognition of the differences. Picture 3*, identical to picture 3 but subjects were asked to search for any additional differences, which represented a test of confidence (2,4). Picture 4 (a scene) with 10 elements including four different animals, the sun, an airplane, five trees, a house, two mountains, and a river, which examined the subjects for visual ability and short term memory. Picture 5 (an open circle), identical to picture 1, which was examined to detect any influence of the stimulus of picture 4 and any habituation. For picture 4, all subjects were required to note all elements immediately after viewing. We evaluated the total number of
Fig. 1. Test pictures and comparison of two typical series of exploratory eye movements (adult men and women).
572
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
identified elements as element scores (the maximum possible score was 13 points) to check the subjects’ arousal levels. Data were analyzed from the left eye in this study. There are no significant differences in movements between the right and left eyes (unpublished observation). Statistical analysis For data considered reliable (epsilon , 1.0), a one-way ANOVA was used to compute the main gender effect (men and women). Post-hoc analyses were conducted using Fishers protected least square deviation (LSD) (13). Next, a two-way ANOVA was used to examine the interaction (sex 3 picture). A level of p, .05 was accepted as statistically significant. Results Mean gazing time The mean gazing time of girls (prepuberty) was 0.37960.089 (n511, mean6standard deviation, SD) and that of age-matched boys was 0.37960.079 (n511). For adult women it was 0.45360.113 and 0.39360.059 for age- matched adult men, and 0.40960.105 for postpuberty women and 0.39060.093 for age-matched men (see Table 1). In the adult, the mean gazing time of women was significantly longer than that of men by two-way ANOVA (sex 3 Table 1 Gender differences in the eye movement of three life spans
Statistical Significance was determined between men and women on four measurements (mean gazing time, total number of gazing points, mean scanning length of gazing points, and total scanning length). Statistical significance as calculated by Fisher F test is indicated by asterisks. The life span differences in each sex are indicated by a, aa, aaa. The significant differences between prepuberty, adult and/or postpuberty are indicated by b,bb bbb.
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
573
Fig. 2. Gender differences in mean gazing time of three life spans. A. The mean gazing time of women was significantly slower than that of men only in the adult group, as is indicated by the closed stars. Significant differences were observed in women between prepuberty and adult, and adult and postpuberty, as indicated by the open stars. B. Picture 1 (open circle), picture 4 (scene) and picture 5 (open circle identical picture 1) were significantly different between adult men and women.
picture) [F(1,204)523.88, p, .0001). Significant differences between adult women and agematched men were found with three pictures- the first open circle (picture 1; F[1,34]55.05], p,.05), the scene (picture 4; F[1,34]57.57, p, .001), and the last open circle (picture 5; F[11,34]50.93], p,.001). The mean gazing time of adult women was significantly longer than that of prepuberty (girls) and postpuberty (older women). There were no significant differences in mean gazing time among the three life spans in men as seen in Fig. 2. Total number of gazing points The total number of gazing points of adult women were significantly less than those of men as analyzed by two-way ANOVA (sex 3 picture; F[1,204]560.52], p, .0001). There were no significant differences between boys and girls or between postpuberty women and age-matched men. For each picture, significant differences were apparent between adult men and women. Picture 1(F[1,34]515.58, p,.001), picture 2(F[1,34]58.62, p,.01), picture 3(F[1,34]55.55, p,.05), picture 4 (F[1,34]519.72, p,.001) and picture 5 (F[1,34]514.01,
574
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
Fig. 3. Gender differences in total scanning length of three life spans. A. The total scanning length of women was significantly smaller than that of men only in the adult group, as indicated by the closed stars. Significant differences were observed in men between prepuberty and adult, and adult and postpuberty in men, as indicated by the open stars. B. The differences in scanning length between adult men and women were highly significant for picture 4 (scene) and picture 5 (open circle).
p,.001) were significantly different between adult women and age matched adult men. Pictures1, 4 and 5 showed larger differences than the others. The total number of gazing points of postpubescent women was significantly larger than those of girls and adult women. The total number of gazing points of boys was significantly larger than that of adult and senior men. Mean scanning length The mean eye scanning length of adult men was significantly longer than that of agematched women as analyzed by two-way ANOVA (gender 3 picture; F[1,204]55.33, p, .05). There were no significant differences between boys and girls, nor between postpubescent women and age-matched men. No significant differences were found for individual pictures. The mean scanning lengths of postpubescent women and age-matched men were shorter than those of boys and girls, respectively. In the male, the mean scanning length of senior men was significantly shorter than that of adult men.
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
575
Total scanning length The total eye scanning length of adult men was significantly longer than that of adult women as analyzed by two-way ANOVA (gender 3 pictures; F[1,204] 549.38. p, .001). There were no significant differences between boys and girls or between postpubescent women and age-matched men. The total scanning length shows significant differences on five pictures. The difference between picture 4 (F[1,34]515.28, p, .001) and picture 5 (F[1,34]5 19.47, p,.001), was larger than for the others (picture 1, F[1,34]57.48, p, .01; picture 2, F[1,34]55.37, p, .05; picture 3,F[1,34]55.56, p,0.05). The total scanning lengths of adult men and senior men were significantly longer than those of boys. However, there were no differences among the women in each life span.
Discussion The major finding of the present study was that a gender influence on eye movement parameters was observed in healthy subjects. Significant differences were found only in the adult phase, and they were attributable to gender. The mean gazing time of adult women was consistently longer than that of men, and the scanning length of adult women was shorter than that of men. Gender differences have been shown in a variety of electro-physiologic measures. Adult women were reported to have larger amplitude brainstem and cortical evoked responses (5) and shorter EP latencies than men (15,16). Anatomical differences between men and women have been suggested to underlie differences in EP source origins. However, anatomical differences are not likely to explain the present findings because gender differences varied according to the pictures . For example, on mean gazing time, there were significant gender differences only in pictures 4 and 5 but not in the other pictures. If anatomic differences were involved in exploratory eye movement, more consistent differences would be expected irrespective of the pictures. In a sensory vigilance task, young women had slower reaction times to target stimuli than men of similar ages, and women detected fewer targets (9). The authors suggested that these effects may have reflected different arousal levels in men and women. Thus, arousal may decline during the presentation of the six pictures. However, picture 1 showed certain gender differences, and the differences in mean gazing time and total scanning length were larger than in pictures 2 and 3. These findings indicate that a decline of arousal level may not result in gender differences during sustained viewing, such as with the 6 pictures. However, we cannot completely rule out the possibility that the arousal level of women may have been lower than that of men, especially in the last two pictures. Hormonal influences have been implicated in gender differences in ERP (event related potential). In that study, which spanned many subject ages, women reportedly had a higher mean amplitude response of P50 to a first stimulus than men (17,18). The gender differences in pattern reversal EP (evoked potential) amplitude over many subject ages suggested a hormonal influence on visual spatial frequency processing (11). Gender differences in the brain’s functions of language and spatial ability are also known. We speculate that in the present study hormonal changes are the likely cause of the gender differences because they were observed only between adult women and age-matched men. Furthermore, there were no gender
576
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
differences between girls (prepuberty) and age-matched boys, and also between postpuberty women and age-matched older men. It has been reported that spatial ability in androgen deficient men is similar to that of women (19). Indeed, gonadal hormones are reported to effect EEG changes (20). Another interesting finding was the association of certain characteristics to each life span in each sex. For example, postpubescent women have a long mean gazing time, large total number of gazing points and short mean scanning length. The age-matched older men have a short mean scanning length. The mean gazing time of adult women is the longest and the total scanning length of adult men is the longest. These life span differences in each sex may also indicate the influence of gonadal hormones in eye movements. The mechanism of gonadal hormone regulation of visual information processing may be further clarified by measuring hormonal levels at all ages. It is interesting that the eye movements for the scene (picture 4) and the second open circle (picture 5) were largely or significantly different between adult men and women in the present study, suggesting that the addition of details, such as animals and a house, stimulated visual information processing differently in men and women. Additionally, as shown in picture 5, the recovery cycle from the visual stimuli may be different in men and women as suggested previously (21). In conclusion, this study evaluated the effect of gonadal hormones on visual information processing in healthy humans (2,11). Evidence was presented that men perform better than women in spatial and motor tasks (6). Indeed, It has been reported that gender differences may exist in spatial ability and in functional organization of the brain for language (22,19). In this study, gender differences were not observed in either childhood (prepuberty) or in the elderly (postpuberty), but clear differences were observed in the adult. It is possible that tasks such as viewing pictures, which require a great amount of cognitive processing, involve different cognitive strategies in men and women. The present finding may reflect gender related behavioral differences. It has been reported that men’s spatial ability may be enhanced to function in a human hunter society. As shown in Fig. 1, open circle, men would scan across the line and tend to look at the outer space, but women would scan across the line and tend to look at the inner space. It may be speculated that men concentrate on the target and thus pay attention to other parts, especially outside of the target, whereas women would also concentrate on the target but pay attention to the inside.(3) These visual differences may reflect certain social behaviors in age development. Further study is needed to explore possible visual desensitization or fatigue phenomenon during repetitive viewing, which was different in men and women. References 1. Holzman PS., Kringle E., Levy DL, Proctor LR., Haberman SJ. Yasillo N.J., Abnormal pursuit eye movement in schizophrenia: evidence for a genetic indicator. Journal Of Archives of General Psychiatry 1977; 34 : 802– 805. 2. Kojima T., Matsushima E., Nakajima K., Shiraishi H., Ando K., Shimazono Y., Eye movement in acute, chronic, and remitted Schizophrenics. Journal of Biological Psychiatry 1990; 27: 975–989. 3. Miyahira A., Morita K., Yamaguchi H., Nonaka K., Maeda H., Gender differences and reproducibility in exploratory eye movements of normal subjects. Jounal of Psychiatry and clinical neurosciences 2000; 54: 31–36. 4. Matushima E., Eye movement in schizophrenic patients. Psychiatria et Neurologia Japonica 1988; 2: 89–110.
A. Miyahira et al. / Life Sciences 68 (2000) 569–577
577
5. Baumann SB., Rogers RL., Guinto FC., Saydjart CL., Papanicolaou AC., Eisenberg HM., Gender differences in source location for the N100 auditory evoked magnetic field. Jounal of Electroencephalography of Clincal. Neurophysiology 1991: 80: 53–59. 6. Bakan P. and Putnam W., Right-left discrimination and brain materialization : sex differences. Journal of Archives of Neurology. 1974; 30: 334–335. 7. Delacoste UC. Holloway RL., Sexual dimorphism in the human corpus callosum. Journal of Science. 1982; 216: 1413–1432. 8. Shaywitz B., Shaywitz S., Pugh K., et al. Sex differences in the functional organization of the brain for language. Jounal of Nature. 1995; 373: 607–609. 9. Giambla LM., Quilter RE, Sex differences in sustained attention across the adult life span. Journal of Applied psychology. 1989; 74(1): 91–9. 10. Efron R., Yund EW. Nichols DR., Scanning the visual field without eye movements; a sex difference. Journal of Neuropsychologia 1987; 25: 637–644. 11. Sheare DE., Emmerson RY. Dustman RE., Gender differences in pattern reversal evoked potential amplitude; Influence of check size and single trial response variability. Journal of American Journal of EEG Technology. 1992; 32: 196–203. 12. Poulton EC., Peripheral vision, refractoriness and eye movements in fast oral reading. British Journal of Psychology. 1962; 53:409–419. 13. Miyahira A., Morita K., Nonaka K., Nakamura J. Maeda H., Effects of caffeine and ATP on eye movement. Journal of Kyushu Neuropsychiatry. 1997; 43: 114–120. 14. Maeda H., Morita K., Kawamura N. Nakazawa Y., Amplitude and area of the auditory P300 recorded with eyes open reflect remission of schizophrenia. Journal of Biological Psychiatry. 1996; 39: 743–746. 15. Taylor MJ., Smith ML., Iron KS., Event-related potential evidence of sex differences in verbal and nonverbal memory tasks. Journal of Neuropsychologia. 1990; 28(7): 691–705. 16. Trune D., Mitchell C., Phillps D., The relative importance of head size, gender and age on the auditory brainstem response. Hearing Research. 1988; 32: 165–174. 17. Freedman R., Adler LE., Waldo M., Gating of the auditory evoked Potential in children and adults. Journal of Psychophysiology. 1987; 24(2): 223–227. 18. Hetrick WP., Sandman CA., Bunney WE., Jin Jr.Y., Potkin SG. White MH., Gender differences in gating of the auditory evoked potential in normal subjects. Journal of Society of biological Psychiatry. 1996; 39: 51–58. 19. Hier DB. Growley WF., Spatial ability in androgen deficient men. Journal of New England Journal of Medicine. 1982; 306 :1202–1205. 20. Vogel W., Broyerman DM., Klaiber EL., Kobayasht Y., Clarkson FE., Itil TM., Laudahn G., Herrmann WM.(Eds), A model for the integration of hormonal behavioral, EEG, and pharmacological data in psychopathology. Journal of Psychotropic Action of Hormones. Spectrum (New York), 1976; 121–134. 21. Lolas F., Inter hemispheric and sex differences in the visual evoked response recovery cycle. Journal of Neuropsychobiology. 1979; 5(6): 301–308. 22. Shaywitz BA., Shaywitz SE., Pugh KR., Sex differences in the functional organization of the brain for language. Journal of Nature. 1995; 372: 607–609.