An expected and normal activity of an individual is that they periodically blink, i.e., transient or very brief closures of the eyelids occur without any obvious stimulus. 1 This activity, often referred to as the spontaneous eye blink activity, 2 has been studied for may years by individuals with a wide range of interests. 1 Assessment of the spontaneous eyeblink rate (SEBR) in humans has been carried out in relation to research on visual task performance assessments, psychophysical state, and in relation to ocular surface disease and contact lens wear.
A recent meta-analysis of studies reported during a 75-year period indicated that whereas the basic descriptor of “spontaneous” is common to many studies, the SEBR values can be shown to be predictably different depending on the physical disposition and general state of the individual while their so-called spontaneous eyeblink activity was being assessed in some way. 1 The SEBR can be expected to be lowest when individuals are directing their gaze downward and are engaged in some form of reading activity; this should be described as a reading-SEBR. While subjects are sitting in silence and assuming their habitual straight-ahead eye position (primary eye gaze), such a primary gaze-SEBR can be expected to be slightly higher the reading-SEBR. Finally, the analysis revealed that whenever an individual is engaged in some type of conversation (e.g., casual conversation or a structured interview), the conversational-SEBR would be expected to be higher than the primary gaze- or reading-SEBR; the differences between the various SEBRs were statistically significant and were essentially consistent with independent studies in which SEBR was assessed during reading, silence, and conversation. 1,3 Notwithstanding, even with such robust subcategorization of spontaneous eyeblink activity, the analyses also revealed rather substantial interstudy differences in the methods by which spontaneous eyeblink activity was assessed, including SEBR studies. There appears to be little agreement as to when or for how long measurements should be taken and how the resultant data should be analyzed. The present studies were designed to further address these analyses for measures of spontaneous eyeblink activity in primary eye gaze.
In previous studies, 4 analyses were presented to indicate that an assessment of spontaneous eyeblink activity in primary eye gaze should be carried out during a period of several minutes, and a 5-min standard was proposed, 4 partly because such a period is also consistent with numerous older studies on eyeblink activity during reading. 1 Beyond assessments of SEBR per se, relatively few further data are actually available on the characteristics of spontaneous eyeblink activity in normal humans during primary eye gaze. Some key aspects relate to what the actual inter-eyeblink interval (IEBI) values are and how these relate to both the pattern of spontaneous eyeblink activity and the overall rate.
In early studies by Ponder and Kennedy, 5 IEBI values were reported for a group of 55 individuals, and data were presented that there were essentially four types of eyeblink patterns, as based on analysis of the distribution of IEBI values. These patterns were irregular (I-type), J-type, bimodal, and symmetrical and were obtained with an electromyograph, i.e., an apparatus, linked to a series of electrical contacts, attached to the upper eyelid and producing a drum recording of eyeblink activity. In I-type patterns, there were a range of IEBI values with no obvious modal value. In the J-type pattern, most IEBI values were very short and interspersed with a few longer intervals, leading to a very positively skewed distribution of IEBI values. The bimodal pattern was one in which the distribution of IEBI values included a set of very short values and a set of somewhat longer values clearly grouped around a secondary modal value. In the symmetrical pattern, the distribution of IEBI values approximated a Gaussian distribution with a clear modal value. That different types of eyeblink patterns could be observed in humans has been confirmed in later studies on small groups of individuals, but with the assessment of spontaneous eyeblink activity being by videography rather than with an electromyograph. 4,6
The present studies were designed to extend these previous studies 4,6 and to assess the frequency of occurrence different patterns of eyeblink activity and the detailed characteristics of these patterns.
The study protocol was approved by review from a departmental and university ethics committee. A total of 61 subjects were recruited, by personal contact, from the undergraduate student population, mostly from Vision Sciences. The subjects were aged 18 to 28 years (mean, 20.9; median, 20). There were 30 men and 31 women. None of the subjects had any history of major eye disease or abnormality other than mild to moderate refractive errors, none reported any neurological diseases, and none of the subjects were contact lens wearers (because this was an exclusion criterion for this study). Based on responses to questions, all subjects had normal spectacle-corrected visual acuities (for minor-to-modest refractive errors) and reported no vision problems.
The subjects were asked to respond to a general questionnaire on age, gender, contact lens wear, medication use, and whether they experienced any ocular symptoms. 7 and provide written consent. Based on responses to the questionnaire, none of the subjects experienced significant ocular symptoms in that whenever symptoms (e.g., dryness or itchiness) were reported, these were experienced only sometimes (as opposed to often or always); 27 of the subjects reported symptoms just sometimes (14 men, 13 women). Of these, 12 reported symptoms when exposed to provoking stimuli (e.g., radiators, cigarette smoke). Seven of the subjects reported occasional use of oral antihistamines for seasonal allergies, three subjects reported use of anti-asthma medications, and five of the women reported use of oral contraceptives.
The subjects were seated in a 12- × 8-ft clinic investigation room with an air exchange of about 5 changes/h, but without perceivable airflow (drafts). The room is adjacent to an internal corridor but, within reason, was generally free of significant background noise. The temperature (mean ± SD, 20.7 ± 0.4°C; range, 18 to 21°C) and humidity levels (mean, 36 ± 4%; range, 31 to 40%) were routinely monitored. Ambient lighting around the subject was about 600 lux, and the illumination at the target (see below) was about 350 lux. For a subject seated in the examination chair (see below), the luminance off the target was about 75 to 85 cd/m2.
Assessment of Spontaneous Eyeblink Activity
The subjects were first acclimated to the examination room. This was generally done during a period of about 5 min so the subjects could adapt to the temperature, humidity, and lighting in the room. During this period, the subjects were asked a number of questions as they completed a questionnaire (see above). The assessments were all carried out between 10 a.m. and 4 p.m.
Analysis of eye-blinking was done by analysis of videotape recordings made of each subject essentially following previous methods. 4 The subjects were seated in a comfortable, high-backed chair and asked to direct their gaze toward a 35-mm-high, vertically oriented black cross on the white background of the wall that was 2 m distant (Fig. 1). The chair height was adjusted so that the viewing was comfortable and the subjects adopted primary eye position. All subjects reported being able to see the target satisfactorily, and this included spectacle wearers who were asked to remove their spectacles during video recording.
A Panasonic VHS-compact movie camera (model NV-G2B) camera was located at head height, but at an angle and away from the subject so as not to directly interfere with the visual task presented. No specific prior training or “test” runs were performed, and the camera was started with the examiner seated behind the subjects who were simply and quietly asked to maintain their direction of gaze toward the target and not to talk. The subjects were not specifically told at the start of the videography that their eyeblinks were being monitored, but subjects were recruited to participate in a study on eyeblinking (see discussion). The videography was done for a period of a little more than 5 min, and the subjects were not given clues as to the passage of time.
Kinetic Analyses of Videographic Data of Spontaneous Eyeblink Activity
The videotape from each of the subjects was analyzed using an event marker essentially as previously detailed. 4 The resultant video image included head and neck of the subject and was played back on a 13-cm television monitor. While the video was being played in real time, a button on the event marker was depressed every time an eyeblink event was considered to have occurred. Based on pilot studies, 4 the interobserver agreement for this type of assessment on young adults is at least 91%. For all records, with the event marker trace running at 4.2 mm/s on a stepping motor, the trace was then used to assess the time (within ±0.1 s) at which the start of an eyeblink event occurred. These measures of the eyeblink timing were done manually with a rule, starting with the first completed eyeblink event. From the sequence on the trace, the total number of blinks in 5 min was counted, the overall blink rate and the minute-by-minute blink rate were calculated, and the IEBIs were measured. For the IEBI values, an allowance of 0.1 s was made for the actual duration of the eyeblink. After the IEBI values were entered into a statistics software package (see below), histograms were generated of the distribution of the IEBI values 4,5 so that the eyeblink pattern could be categorized as irregular, J-type, or symmetrical (normal).
All data were entered into a software program (Systat, Version 8.0; Systat, Evanston, IL) that allowed descriptive statistics to be generated for the entire group and by category of age, gender, and eyeblink pattern. Correlation and regression analyses and comparisons between data sets were carried with routines available in the software, with the cut off for statistical significance being set at p = 0.05. Assessments for normality were carried out with the Lilliefors variant of the Kolmogorov-Smirnov one-sample test, and nonparameteric comparisons were made using the Kruskal-Wallis analysis of variance or Friedman test.
General Characteristics, Constancy, and Pattern of the Spontaneous Eyeblink Activity
A total of 65 individuals were initially recruited and videotaped, but records from four individuals were discarded from analysis. Two rejections were made because the subjects could not refrain from talking (even after two attempts), and two other traces were rejected because the subjects did not maintain their eye gaze toward the target during the 5-min period (i.e., they became restless). For the remaining 61 subjects, a successful recording was obtained. Short (1-min) excerpts from traces are shown in Fig. 2 for three different subjects.
The traces reveal that subjects, seated quietly and viewing a target in primary eye gaze, show a generally unremarkable sequence of eyeblink events. A cursory inspection of the event marker outputs during the 5-min recording periods generally revealed no obvious pattern or sequence to the spontaneous eyeblink activities. However, as illustrated in line A of Fig. 2, there are sometimes a few closely grouped eyeblink activities (i.e., several blinks occurring in relatively quick succession followed by a quieter period of activity). In other cases, the grouping is a little more widely spaced, and the interval to the next blink is a little longer (Fig. 2, line B). For some subjects, the pattern of blinking appears to be fairly regular (Fig. 2, line C). For the three excerpts, the SEBRs were 8, 11, and 13 eyeblinks/min, respectively.
From such excerpts, repeated during the entire 5-min recording period, the SEBR can be calculated. For the group of 61 subjects, the SEBR for each minute was calculated, and the results were averaged (Fig. 3). As the figure shows, there was no substantial time-dependent trend in the mean SEBR values during the 5-min videography, and the same result is found if either the median SEBR values are plotted in a similar way (not shown) or various types of regression analysis are undertaken with the individual data points (not shown, but including linear, logarithmic, exponential, and quadratic). The figure indicates that the mean SEBR values are marginally higher (mean ± SD, 11.3 ± 0.5 eyeblinks/min) for the first minute compared with minutes 2 (mean, 9.6 ± 4.7 eyeblinks/min) and 3 (mean, 9.8 ± 4.5 eyeblinks/min), but these differences were not found to unambiguously statistically significant. The distributions of SEBR values for each subject for each minute were approximately normal (i.e., p values close to 0.05), a paired t-test for minute 1 vs. minute 2 indicated statistical significance (p = 0.01), and a Friedman nonparametric rank analysis also indicated the difference to be just significant (p = 0.035). Comparing the SEBR values in the first minute with the overall SEBR value for 5 min (mean, 10.3 ± 3.1 eyeblinks/min) could also be shown to be just statistically significant using t-test (p = 0.025), but not on a Friedman analysis of variance (p = 0.055). However, all other comparisons (e.g., minute 1 vs. minutes 2, 3, 4, or 5) failed to reach statistical significance (p ≥ 0.05). It should also be noted that the intersubject variance (SD) is four or five times the net difference between SEBR values for either the first minute compared with the 5-min period or the second minute. The differences between minute 1 and minute 2 of the video sequences are therefore considered to be chance occurrences that do not reflect an important time-dependent change in SEBR (see Discussion).
General Characteristics of the Inter-Eyeblink Intervals
From the event marker traces, the actual IEBIs were measured. Across the 61 subjects and during the 5-min recording periods, a wide range of values for IEBIs was observed, namely 0.3 to 69.7 s. For each individual, a mean IEBI value was calculated, and these ranged from 3.8 to 12.7 s, with a group mean value of 6.4 ± 2.4 s (N = 61). The overall distribution of the mean IEBI values from each subject was substantially skewed (Fig. 4, top).
The coefficient of skewness on the distribution of mean IEBI values was 0.922, and the distribution was not normal (p < 0.001 test for normality). The median value for the IEBI values was slightly less than the mean value at 5.5 s. The overall variance (calculated as the normalized SD) was modest: 3.1 s in 10.3 s (i.e., 37.5%). The distribution of this variance shows that for many individuals, it is around this level (Fig. 4, bottom). However, the variance was skewed (coefficient of skewness, 0.814) and for some individuals was even larger than the mean value.
The skewness on the distribution of IEBI values and their variance means that the inverse correlation between the calculated SEBR values (based on the number of eyeblink events during the 5-min observation period) and the IEBI values is not perfect (Fig. 5), although the overall correlation is still good (r = 0.957; see Discussion).
Analysis of Gender-Related Differences in Spontaneous Eyeblink Activity
No substantial gender-related differences in either the rate or characteristics of the spontaneous eyeblink activity were detected. For men, the mean SEBR was 10.8 ± 2.7 eyeblinks/min (N = 30), whereas it was 9.7 ± 3.5 eyeblinks/min (N = 31) for the women; the difference was not statistically significant (p = 0.145). The mean IEBI value for men was 5.9 ± 1.9 s, whereas it was 6.9 ± 2.6 s for the women (p = 0.220). Similarly, neither the minimum (0.65 ± 0.64 vs. 0.56 ± 0.30 s) nor the maximum (22.7 ± 11.4 vs. 29.2 ± 15.7 s) inter-eyeblink intervals were statistically significant (p ≥ 0.1; see below for further analyses). The failure to identify a gender-related difference is, however, likely due to the overall intersubject variance in spontaneous eyeblink activity (see below and Discussion section).
Categorization of the Pattern and Variance of Spontaneous Eyeblink Activity
As indicated above, the spontaneous eyeblink activity of younger adults is somewhat variable. With the reasonably large cohort studied, it was considered a useful exercise to explore the causes of this variability, especially in terms of the eyeblink pattern.
Overall, the patterns of eyeblink activity were composed of a series of eyeblinks, and the overall output from any one individual can be characterized in terms of the shortest and longest inter-eyeblink intervals. The data on these for all of the subjects are shown in Fig. 6. The shortest inter-eyeblink intervals usually were ≤1 s, with a minimum of 0.3 s (Fig. 6, top). There were, however, a few individuals who did not show very short IEBI values (i.e., eyeblinks in very quick succession) and for whom the shortest interval between eyeblinks was >1 s (and even as much as 3.3 s). There was thus a 10-fold difference in the shortest IEBI values across such a group of younger individuals. A similar characteristic can also be noted for the longest inter-eyeblink values (Fig. 6, bottom), with a range from 8.0 to 69.7 s.
Measures of all of the IEBI values for each subject during the 5-min period and subsequent analyses of the distribution of the IEBI values for each subject indicated that approximately one-third of the subjects showed an irregular eyeblink pattern, one-third showed J-type pattern, and the remaining one-third of the subjects showed a symmetrical type. No example of a bimodal pattern for IEBIs was observed. Analyses to assess whether there were any time-dependent changes in SEBR for the three different eyeblink patterns is shown in Fig. 7. As with the data from all 61 individuals, analyses of each individual group revealed no obvious time-dependent changes in SEBR during the 5-min videography period. For those showing an irregular eyeblink pattern (as shown in the top analysis from Fig. 7), the SEBR values from the first and second minute could be shown to be just statistically significant (Friedman analysis of variance, p = 0.037), but none of the other comparisons were found to be statistically significant (p ≥ 0.05).
The traces provided in Fig. 2 are considered to be representative examples of irregular, J-type, and symmetrical eyeblink activity, respectively. The example shown in line A of Fig. 2 suggests that an irregular eyeblink pattern is perhaps composed of fairly long inter-eyeblink intervals interspersed with two (or more) rapidly repeated eyeblinks, although this was not always that obvious. Such rapidly repeated eyeblink events may be considered as being “flurries” (see Discussion). In comparison, those with a J-type pattern (Fig. 2, line B) showed some grouping of eyeblinks, and the pattern appeared a little more regular. Analyses of IEBIs in such a case show a set of shorter values and then a number of progressively longer values. The most regular type of pattern is shown in line C of Fig. 2 where it can be seen that most of the IEBI values are of a similar duration; this is the symmetrical pattern.
Overall, analyses of the SEBR and IEBI values (Fig. 8) clearly show substantial differences between the three groups of subjects exhibiting the different types of eyeblink patterns. As indicated from the box plots (Fig. 8), the median SEBR value for those with an irregular eyeblink pattern was substantially lower (6.3 eyeblinks/min) compared with those with a J-type (11.0 eyeblinks/min) or normal eyeblink pattern (12.6 eyeblinks/min). As would be expected, the mean SEBR value for those with the symmetrical eyeblink pattern is almost the same as the median value (12.3 eyeblinks/min). The expected inverse characteristics for the IEBI values was also evident (Fig. 8) in that the median IEBI value for subjects showing an irregular blink pattern is almost twice that for the other eyeblink patterns. The overall data are summarized in Table 1. Comparisons of SEBR values and the maximum IEBI values between the groups, by nonparametric statistics, showed all groups to be statistically different from each other (p < 0.05). For mean IEBI values, the group with irregular eyeblink patterns was different from those with J-type or normal patterns (p < 0.001), but the difference between the IEBI values for the J-type and normal group was only just statistically significant (p = 0.040). With the size of the cohort studied, there was insufficient data to be able to usefully and unambiguously ascertain whether there were differences between men and women in terms of the patterns of spontaneous eyeblink activity. No obvious differences were evident, however, i.e., approximately the same numbers of men and women exhibited the different types of eyeblink pattern.
The results from the present study, which was based on more detailed assessments of the primary gaze-spontaneous eyeblink activity in young adults, confirm that different types of eyeblink pattern can be routinely observed. The overall SEBR is clearly dependent on which pattern is usually observed. It is proposed that that pattern of spontaneous eyeblink activity where there are similar intervals between each eyeblink should be considered as the normal one, both from a physiological and statistical perspective. That such patterns exist and that some of them are presumably not “normal” must surely be taken into account when considering the function of spontaneous eyeblink activity. This is relevant to contact lens wear and to advice offered to overcome anomalies in eyeblink activity during contact lens wear. 8,9 Similarly, the impact of different patterns of eyeblink activity on SEBR needs to be considered to further develop our understanding of any relationships between symptoms of ocular discomfort and spontaneous eyeblink activity, 7,10 including those associated with dry eye symptoms. 7,11
In the present studies, an assessment of spontaneous eyeblink activity, in primary eye gaze and in silence, was made using nonconcealed videography of subjects during a 5-min period. This method is the same as that used in a previous pilot study, 4 and the general approach was adopted from that outlined by Yolton and colleagues. 12 The results for SEBR are generally consistent with most other studies using a similar paradigm (i.e., primary eye gaze and silence) and both confirm and extend previous observations on the characteristics of eyeblink patterns in humans. 6 The results of the present studies clearly show that a cause for intersubject differences in SEBR can be attributed to different patterns of spontaneous eyeblink activity despite the constancy of the visual task. Future studies should address the issues of different eyeblink patterns rather than attempting to simply provide mean values for either SEBR or inter-eyeblink intervals.
It is unknown at this time why different individuals exhibit clearly different patterns of spontaneous eyeblink activity, but an argument can be made that these are either due to an endogenous feature of the individuals or the circumstances under which the assessments were made. Different patterns of eyeblink activity have previously been observed both by electromyography and during reading 5 as well as by videography in primary gaze. 4,6 Therefore, the different patterns cannot be obviously attributed to eye position or an interference due to the recording system.
The subjects for this study were recruited on the basis of being “normal,” i.e., they were generally healthy, were not experiencing any common ailments (e.g., a cold), had no significant eye disease (especially that of the ocular surface), were not contact lens wearers, and were generally asymptomatic. Some of the individuals did report use of systemic medications, and a case might be made that these could affect spontaneous eyeblink activity. It has been reported that women taking oral contraceptives could have a substantially and statistically higher eyeblink rate, 12 and that women might have slightly higher (but not significantly different) eyeblink rate compared with men. However, in the present study, the women showed a slightly lower mean SEBR compared with men (but the difference was also not significant). With the variance (±SD) in the SEBR and the small difference observed, the chance of detecting a significant difference is low (28%). Stated another way and assuming the same variance between men and women and the same level of difference, a sample size of about twice the group studied would be needed to have even an 80% chance of detecting a difference. In addition to the gender-related trend, the same was also seen for the female subjects taking oral contraceptives, i.e., these subjects did not show a substantial increase in SEBR as previously reported. 12 It is unknown why this effect was not observed but, overall, much larger cohorts need to be recruited to assess the possible impact of these and other medications. In addition to oral contraceptives, seven of the subjects (four men and three women) reported occasional use of oral antihistamines for seasonal allergies, but analysis of various features of their eyeblink activity showed no remarkable differences, i.e., it was hypothesized that perhaps these medications might produce an irregular eyeblink pattern, but this was not observed.
A question that might be posed is whether the so-called spontaneous eyeblink activity of the subjects in this study might have been altered as a result of these subjects being volunteers and/or because they were aware that their eyeblink activity was being recorded. The data from the present studies, as well as those from a pilot study, 4 are presented as evidence that spontaneous eyeblink activity in primary gaze and in silence does not change appreciably during a 5-min assessment period. Such a constancy was observed for three different patterns of eyeblink activity, i.e., any change is not obviously pattern dependent, although individuals with a bimodal pattern for IEBIs might be different. As detailed previously, 1 it seems likely that far greater changes in SEBR can occur if the experimental paradigm changes, e.g., the subject changes their direction of gaze or starts talking. Notwithstanding, there are some issues that need to be addressed.
For the present studies on eyeblink activity, care was taken to try to avoid making the subjects too aware of the videography. A subject was seated in a high-back comfortable chair and simply asked to direct their gaze toward a high-contrast target on the white wallboard in front of them. The verbal exchange with the subjects before videography was hopefully such that they did not feel obliged to be rigidly attentive. The language used when instructing a subject will, of course, be different for different individuals, but it is hoped that the subjects simply relaxed and looked toward the target without trying to “stare” or consciously “fixate” the target 13 nor suppress their habitual eyeblink activity. However, they were not specifically told to blink normally to hopefully not make them conscious of their eyeblink activity. So, instructions were given to the effect of “I need to video you for a few minutes while you silently look at that target on the wall.” Neither the subject nor the investigator spoke during the videography, and while the investigator monitored the time with a stopwatch, no cues were provided to the subject as to the passage of time. The latter decision was made to try to limit the chance of any anticipatory reaction, e.g., to the impending end of the recording. The chair height was adjusted so that the target was aligned at the height of the primary gaze position. It was considered important that the subject was comfortable and that the visual task was minimal to limit the chance of fatigue. Some data have been presented to indicate that SEBR will progressively increase over time, 14 but the task duration here (5 min) was much shorter than when fatigue was observed (i.e., during 10 min in assessment periods of up to 110 min). Similarly, data have been presented that the mean SEBR might be higher late in the evening compared with during the day. 15 The magnitude of the fatigue-related or evening-related change was only 4 or 5 eyeblinks/min (12 to 16 or 11 to 16 eyeblinks/min, respectively), and these changes are larger than the minor time-dependent changes during the 5-min assessment used in the present studies. It has also been suggested that flurries of eyeblink activity are indicators of fatigue, 14 and so it might be argued that those with irregular eyeblink activity might be simply tired. Further studies are needed to characterize such an effect. The target selected was deliberately large (35 mm high) and of high contrast. A simple character was chosen to hopefully avoid a significant element of cognition and deciphering of a target (e.g., a certain letter on a high-contrast Snellen chart). The target was placed at a shorter distance than a Snellen chart (i.e., 2 m) to hopefully ensure that even subjects with a poor visual acuity (including those with a low-to-modest refractive error) would be able to see it adequately enough to be able to direct their gaze toward it. It was not considered important, however, that the subjects had to be able to see the target clearly.
The results from the present videography studies confirm and extend those of a previous pilot study 4 and earlier observations of Carney and Hill 6,8,9 in that at least three different patterns of spontaneous eyeblink activity can occur in primary eye gaze. Perhaps the most significant point to be made is that approximately 30% of individuals can be expected to show a symmetrical eyeblink pattern. 6 This pattern of eyeblink pattern does not therefore appear to be unusual; although Ponder and Kennedy 5 report a much very much lower incidence of this type of blinking pattern (i.e., 3.6%), it should be noted that their assessments were made with subjects attached to an electromyographic unit. This issue of the pattern of eyeblink activity is important from the perspective of the types of statistical analyses that can be reasonably performed. As pointed out in previous methodological studies, 4 a perfect concordance between the calculated SEBR and the inter-eyeblink interval (r > 0.995) would usually only be expected if the distribution of IEBI values was Gaussian and when the mean and median IEBI values were the same. This close correlation (r = 0.970) was observed for subjects in the present study that exhibited such a symmetrical eyeblink pattern. Because this statistical correlation does exist, it is proposed that the term symmetrical be replaced by normal, with the deliberate intent of both indicating that the distribution of IEBI values was statistically normal and that this should be considered as the benchmark for spontaneous eyeblink activity, at least in primary eye gaze.
In terms of the physiological function of the spontaneous eyeblink, 16 it might be argued that a regularly repeated eyeblink event is more likely to ensure an even distribution of the tear film and thus, perhaps, promote stability of the precorneal tear film. The present studies, as well as those of Carney and Hill, 6 also indicate that the eyeblink rate is higher if a symmetrical pattern of eyeblink activity occurs. It has been proposed that a weak relationship can be observed between spontaneous eyeblink activity and tear film stability. 17,18 However, it needs to be noted that these studies used the observed SEBR values to calculate an IEBI value (rather than actually measuring IEBI values) and, as a result, provide no data on the pattern of eyeblink activity. It can be argued that these calculations are sufficiently unreliable so as to make their use of questionable value. 1,4 A useful outcome should come from specific studies designed to assess possible interrelationships between tear film stability and the pattern of eyeblinking. The formation and stability of the tear film is widely considered to be dependent on eyeblink activity, with the eyelid velocity and the lipid layer being considered critical to the rapid formation of a stable tear film layer. 19–22 Changes in the distribution of the lipid layer have been observed when comparing incomplete with forced eyeblinks, 21 and it would be useful, therefore, to also assess the overall characteristics of the lipid layer in subjects with different types of eyeblink patterns; studies to this effect are currently in progress. Another aspect of the inter-eyeblink interval has also been considered important as a useful measure of tear film instability, namely the maximum blink interval. 23 The measure taken was that for when “a subject opened his or her eyes until he or she could no longer comfortably keep them open.”23 It can surely be argued that this is not a spontaneous activity and so is fundamentally different from the longest IEBI values recorded in the present study. Indeed, the mean for the longest IEBI value in the present studies (22.7 ± 11.4 s; N = 61) is almost twice that reported for a forced maximum IEBI (11.4 ± 1.1 s; N = 50) reported by Nakamori et al., 23 albeit for Oriental rather than white subjects. Further studies are therefore also needed in this area and are in progress.
1. Doughty MJ. Consideration of three types of spontaneous eyeblink
activity in normal humans: during reading and video display terminal use, in primary gaze, and while in conversation. Optom Vis Sci 2001; 78: 712–25.
2. Stern JA, Walrath LC, Goldstein R. The endogenous eyeblink
. Psychophysiology 1984; 21: 22–33.
3. Bentivoglio AR, Bressman SB, Cassetta E, Carretta D, Tonali P, Albanese A. Analysis of blink rate patterns in normal subjects. Mov Disord 1997; 12: 1028–34.
4. Zaman ML, Doughty MJ. Some methodological issues in the assessment of the spontaneous eyeblink
frequency in man. Ophthalmic Physiol Opt 1997; 17: 421–32.
5. Ponder E, Kennedy WP. On the act of blinking
. Q J Exp Physiol 1927; 18: 89–110.
6. Carney LG, Hill RM. The nature of normal blinking
patterns. Acta Ophthalmol (Copenh) 1982; 60: 427–33.
7. Doughty MJ, Blades K, Ibrahim N. Assessment of the number of eye symptoms, and the impact of some confounding variables, for office staff in non-air conditioned buildings. Ophthal Physiol Optics 2002; 22: 143–55.
8. Hill RM, Carney LG. The effects of hard lens wear on blinking
behavior. ICLC 1984; 11: 242–8.
9. Carney RG, Hill RM. Variation in blinking
behavior during soft lens wear. ICLC 1984; 11: 250–3.
10. Acosta MC, Gallar J, Belmonte C. The influence of eye solutions on blinking
and ocular comfort at rest and during work at video display terminals. Exp Eye Res 1999; 68: 663–9.
11. Tsubota K, Hata S, Okusawa Y, Egami F, Ohtsuki T, Nakamori K. Quantitative videographic analysis of blinking
in normal subjects and patients with dry eye. Arch Ophthalmol 1996; 114: 715–20.
12. Yolton DP, Yolton RL, Lopez R, Bogner B, Stevens R, Rao D. The effects of gender and birth control pill use on spontaneous blink rates. J Am Optom Assoc 1994; 65: 763–70.
13. Telford CW, Thompson N. Some factors influencing voluntary and reflex eyelid responses. J Exp Psychol 1933; 16: 524–39.
14. Stern JA, Boyer D, Schroeder D. Blink rate: a possible measure of fatigue. Hum Factors 1994; 36: 285–97.
15. Barbato G, Ficca G, Muscettola G, Fichele M, Beatrice M, Rinaldi F. Diurnal variation in spontaneous eye-blink rate. Psychiatry Res 2000; 93: 145–51.
16. Doane MG. Interactions of eyelids and tears in corneal wetting and the dynamics of the normal human eyeblink
. Am J Ophthalmol 1980; 89: 507–16.
17. Prause JU, Norn M. Relation between blink frequency and break-up time? Acta Ophthalmol (Copenh) 1987; 65: 19–22.
18. Yap M. Tear break-up time is related to blink frequency. Acta Ophthalmol (Copenh) 1991; 69: 92–4.
19. Holly FJ. Formation and rupture of the tear film. Exper Eye Res 1973; 15: 515–25.
20. Forst G. Structure of the tear film during the blinking
process. Ophthalmic Physiol Opt 1987; 7: 81–3.
21. Korb DR, Baron DF, Herman JP, Finnemore VM, Exford JM, Hermosa JL, Leahy CD, Glonek T, Greiner JV. Tear film lipid layer thickness as a function of blinking
. Cornea 1994; 13: 354–9.
22. Owens H, Phillips J. Spreading of the tears after a blink: velocity and stabilization time in healthy eyes. Cornea 2001; 20: 484–7.
23. Nakamori K, Odawara M, Nakajima T, Mizutani T, Tsubota K. Blinking
is controlled primarily by ocular surface conditions. Am J Ophthalmol 1997; 124: 24–30.
Keywords:© 2002 American Academy of Optometry
eyeblink; blinking; eyeblink rates; eyeblink patterns; human