VisionRes.Vol. 31,No. 7/8,pp. 1439-1443, 1991 Printed in Great Britain. All rights reserved

0042-6989/91 $3.00 + 0.00 Copyright 0 1991 Pergamon Press plc

RESEARCH NOTE INCREASE IN SACCADIC INCREASED FREQUENCY C. Department

PEAK VELOCITY WITH OF SACCADES IN MAN

J. LUECK,’ T. J. CRAWFORD,’ H. C. HANSEN~ and C. K~~NNARD’* of Neurology, The London Hospital, Whitechapel, London El IBB, U.K. and Weurologische Universititsklinik, D-2000, Hamburg 20, Germany (Received 20 September 199@ in revised form 6 December 1990)

Abstract-Twelve normal subjects (aged 22-80 yr, mean 47 yr) performed three blocks of 20 saccades made to LED targets stepped back and forth. The first and last blocks were performed at a (slow) rate of 0.18 Hz, while the middle block was performed at the faster rate of 1.15 Hz. Mean saccadic amplitude was unaffected by saccade rate, but latency and duration became shorter at the higher frequency. Most interestingly, the peak velocity increased by approx. 6% when saccades were performed at the higher rate. This increase was statistically significant, even after normalization for saccade amplitude. That saccadic frequency may affect saccadic peak velocity must be considered as a potential variable when analysing saccades. Saccades

Main sequence

Saccade metrics

Saccadic peak velocity increases with saccadic amplitude according to a relation now generally known as the “main sequence”, a term first coined by Bahill, Clark and Stark (1975). For a given saccadic amplitude, peak velocity is wellrecognized to vary considerably from one individual to another (Boghen, Troost, Daroff, Dell’Osso & Birkett, 1974; Baloh, Sills, Kumley & Honrubia, 1975a; Bahill, Brockenbrough & Troost, 1981), and has often been reported to vary with factors such as age (Spooner, Sakala & Baloh, 1980; Warabi, Kase & Kato, 1984; Sharpe & Zackon, 1987), though not all studies find this (Henriksson, Pyykkii, Schalen & Wennmo, 1980; Abel, Troost & Dell’Osso, 1983; Abel, Troost 8z Dell’Osso, 1987). Another major factor affecting saccadic peak velocity is neurological disease (see Leigh & Zee, 1983). Even for data from one individual, the main sequence has been shown to be significantly affected by fatigue (Riggs, Merton & Morton, 1974; Bahill & Stark, 1975; Schmidt, Abel, Dell’Osso & Daroff, 1979; Fuchs & Binder, 1983), and drugs such as alcohol (Franck & Kuhlo, 1970; Wilkinson, Kime & Purnell, 1974; Bittencourt, Lloyd, Richens, Smith, Toseland & Wade, 198 1; Katoh, 1988), benzodiazepines *To whom correspondence

should be addressed.

(Aschoff, 1968; Gentles & Llewellyn-Thomas, 197 1) and antiepileptics (Tedeschi, Casucci, Allocca, Riva, Di Costanzo, Quattrone, Baruzzi & Bonavita 1989). Abducting saccades are usually (Robinson, 1964; Fricker, 1971; Baloh, Konrad, Sills & Honrubia, 1975b; Sonderegger, Meienberg & Ehrengruber, 1986; Collewijn, Erkelens & Steinman, 1988), but not always (Hyde, 1959; Ishikawa & Terakado, 1973; Boghen et al., 1974), reported to be faster than adducting saccades; this may depend in part on the method of recording (Boghen et al., 1974; Hess, Miiri t Meienberg, 1986). Similarly, centripetal saccades are faster than centrifugal ones (Hyde, 1959; Cook, Stark & Zuber, 1966; Abel, Dell’Osso, Daroff & Parker, 1979). Saccades made in the dark are usually found to be slower (Becker & Fuchs, 1969; Riggs et al., 1974; Kiimer, 1975; Henriksson et al., 1980), as are saccades to auditory targets (Zahn, Abel & Dell’Osso, 1978; Zambarbieri, Schmid, Magenes & Prablanc, 1982). The type of saccade is also important. Peak velocities are lower in antisaccades, remembered and predictive saccades (Smit, van Gisbergen & Cools, 1987; Bronstein & Kennard, 1987; Lueck, Tanyeri, Crawford, Henderson & Kennard, 1990), and it is possible under certain circumstances to alter saccadic velocity by voluntary control

1439

1440

Research Note

(Crawford, 1984). It is usually assumed, however, that for data for one individual and for a given type of saccade, the relation between peak velocity and amplitude is fixed. We here report that the rate at which saccades are made also has a significant effect on peak velocity, saccades made at a higher rate having relatively larger peak velocities than those made at a lower rate. Twelve normal subjects performed the experiment, their ages ranging from 22 to 80 yr (mean 47). None was on medication known to affect cerebral function or eye movement. Eye movements were measured using infra-red electrooculography (IROG) (Reulen, Marcus, Kroops, de Vries, Tiesinga, Boshuizen & Bos, 1988), and recorded using a Minograph chart recorder (bandwidth 5OOHz) at a paper speed of 25 mm/set. Eye position was differentiated online (bandwidth 100 Hz) to yield velocity which was also recorded on the chart recorder. Stimuli were controlled by a PDP 1l/73 computer, and the saccade latency, amplitude, duration and peak velocity of all initial saccades were measured using a Tectronix 4025A digitising tablet accurate to kO.02 in. (0.15 mm). Any saccade occurring within + 50 msec of a blink was not included in subsequent analysis. All subjects performed the same paradigm which involved following target steps back and forth between two light-emitting diodes (LEDs) located at +_15 deg on either side of the vertical midline. The experiment was performed in the dark. The LEDs could be alternated at two different frequencies: 0.18 Hz and 1.15 Hz. Subjects were presented with three blocks of 20 target jumps, the first and third at the lower rate, the second at the higher rate (i.e. 0.18 Hz, then 1.15 Hz, then 0.18 Hz again). Subjects were asked to follow the target LEDs quickly and accurately, and not to anticipate target movement. No spectacle correction was provided. Analyses of variance using a one-way repeated measures design with three levels of saccade frequency were performed on saccadic latency, amplitude, peak velocity and duration. (The two blocks at 0.18 Hz were treated separately in order to observe any effects of practice or fatigue.) If significant effects were observed, post-hoc c-tests were performed to determine which of the conditions was responsible. Since it could be argued that small alterations in saccade amplitude were responsible for any effects observed, the data were then normalized for amplitude. Duration has frequently been shown to increase linearly with saccade ampli-

tude (Hyde, 1959; Robinson, 1964), and therefore normalization was achieved by dividing the durations of the individual saccades by their respective amplitudes. The main sequence is usually accepted as being most approximately described by a negative exponential of the form Peak velocity = A [1 - exp( - B. amplitude)] where A and B are constants (Baloh et al., 1975a). Normalization of peak velocity was therefore achieved by dividing the peak velocities of individual saccades by the natural logarithm of their respective saccadic amplitude. Analyses of variance were then performed on the normalized data, using a design identical to that used for absolute data, followed by c-tests when appropriate. The mean results are shown in the table. As there was no statistically significant difference between the effect on abducting and that on abducting saccades, these were combined for the purpose of further analysis. The latency of saccades performed at the lower frequency was typical of reflex saccades (about 200 msec), but the latency at the higher frequency was, on average, about 80 msec, indicating that some of the saccades were predictive. The difference in latencies was significant [F(d.f. = 2, 11) = 4.27; P < 0.051. The effect of saccade frequency on mean saccade duration was also significant [F(d.f. = 2, 11) = 11.82; P < 0.011, though this the case after normalization was not [F(d.f. = 2,ll) = 3.48, NS]. For latency and absolute duration, post-hoc t-tests indicated a significant difference between the 1.15 Hz condition and both 0.18 Hz conditions. The mean initial saccade amplitude was approx. 27 deg in all three blocks, and there was no significant difference after analysis of variance. There was, however, a significant effect of saccade frequency on mean peak velocity which Table 1. Mean saccadic measurements at different saccade frequencies. Latencies and durations are given to the nearest 10 msec; peak velocities to the nearest 10 deg/sec Slow (0.18Hz) Latency (msec) (fSEM) Amplitude (deg) (& SEM) Duration (msec) (ItSEM) Peak velocity (deg/sec) (*SEM) Duration/amplitude (iSEM) Peak velocity/ln(amplitude) (zbSEM)

27.16 (0.81) 100 (4)

Fast Slow (1.15Hz) (0.18Hz)

26.65 (0.69)

(f7qo)

:; 490 (25.2)

(G3) 141 (4.9)

(fz) 150 (7.0)

170 (34) 26.77 (1.07) 100 (4) 450 (21.2) 3.9 (0.12) 137 (5.7)

Research Note

y=138+0.04x,

R10.15

y-156+0.02x,

RaO0.18

._-

g 4 *

‘i

‘Jo

=

c

e

zt 250

200 160 100 SO Ol

-200

t

0

,

I

200

400

600

400

6

C y=140+0.02~,

A-0.15

350

-200

0

200

0

Istmcy Fig. I. Normalized peak velocity plotted against saccade latency for each of the three blocks of target frequency, 0.18 Hz (A); 1.15Hz (B); and the second 0.18 Hz (C). For comparison purposes, the scales are identical and the linear regression lines extended to the edges of the graphs. As can be seen, the saccades performed at the higher rate (B) had a much greater scatter of iatency, some being anticipatory. However, the regression lines show that the effect of Jatency on normalized peak velocity is such that saccades with longer Iatencies tend to have slightly higher peak velocities. Latency therefore does not account for the effect of saccade frequency on peak veIocity.

1442

Research Note

increased by approx. 67% [F(d.f. = 2, 11) = 6.55; P < 0.051. The significance of the effect was still evident after normalization [F(d.f. = 2,11) = 7.77; P < O.Ol]. Post hoc paired r-tests showed that for velocity there was a significant difference between the fast (1.15 Hz) and both the initial and final slow frequencies (0.18 Hz). In fact, when assessed by analyses of variance applied to individual data, 9 of the 12 subjects showed a highly significant increase in peak velocity of saccades performed at the higher frequency. This study demonstrates that saccades performed at a higher rate (1.15 Hz) have a higher peak velocity for a given amplitude than saccades performed at a lower rate (0.18 Hz). There was no significant difference between the lower rate saccadic parameters before and after performing at the higher rate, and this argues very strongly against there being a spurious effect of, e.g. practice, attention, fatigue or arousal. Similarly, the fact that the effect could be observed to be individually significant in 9 of the 12 subjects suggests the phenomenon is not an artefactual effect of averaging. The mean latency of the saccades at the higher frequency (1.15 Hz) was only 80 msec. As mentioned above, this is much shorter than the typical latency of a reflex saccades and suggests that there may have been an element of prediction. Predictive saccades typically have lower peak velocities than reflex saccades (Bronstein & Kennard, 1987; Smit & van Gisbergen, 1989). The latency difference is therefore very unlikely to be responsible for the velocity effect described here since the saccades with the higher velocities had the shorter latencies. Nevertheless, to investigate latency as a potentially significant factor, normalized peak velocity was plotted against latency for all saccades at each of the three target frequencies (see figure). The plot confirms that the increase in normalized peak velocity observed at the higher frequency was independent of saccade tatency. The rate at which saccades are executed must therefore be added to the list of factors which affect saccadic peak velocity. Unlike many of the factors mentioned above, such as age, drugs, saccade direction and type, saccade frequency may well be an important consideration when grouping together multiple saccades in a given experiment for analysis, and conceivably may account for some differences observed between the results of different experiments. Characteristics of the effect such as whether peak velocity

increases linearly with saccade frequency require further investigation. As yet we have no adequate explanation for the phenomenon. Acknow1edgements-C.J.L.

was in receipt of a Wellcome Training Fellowship. This work was funded by grants from the Medical Research Council and the Wellcome Trust. We are grateful to Dr ‘I’. H. Koexe for the use of his digitizing tablet, and for statistical advice. REFERENCES Abel, L. A., Dell’Osso, L. F., Daroff, R. B. and Parker, L. (1979). Saccades in extremes of lateral gaze. Inuesrigaiiue Ophthalmology and Visual Science, 18, 324-321.

Abel, L. A., Troost, B. T. & Dell’Osso, L. F. (1983). The effects of age on normal saccadic characteristics and their variability. Vision Research, 23, 33-37. Abel, L. A., Troost, B. T. & Dell’Osso, L. F. (1987). The effects of age on saccadic eye movement characteristics. Investigative Oph~~a~moloigy(Suppl.), 20, 26.

Aschoff, J, C. (1968). Ve~nder~gen rascher BIi~k~wegungen (Saccaden) beim Menschen unter Diazepam ~~iurn~. Archiv fir Psychiatrie und ~e~e~ra~~i~en 211, 325-332.

Bahill, A. T. & Stark, L. (1975). Overlapping saccades and glissades are produced by fatigue in the saccadic eye movement system. Experimenral Neurology, 48, 95-106. Bahill, A. T., Clark, M. R. & Stark, L. (1975). Dynamic overshoot in saccadic eye movements in caused by neurological control signal reversals. Experimental Neurology, 48, 107-122.

Bahill, A. T., Brockenbrough, A. & Troost, B. T. (1981). Variability and development of a normative data base for saccadic eye movements. Investigative Ophfhalmology and Visual Science, 21, 116-125. Baloh, R. W., Sills, A. W., Kumley, W. E. & Honrubia, V. (1975a). Q~nti~tive measurement of saccade amplitude, duration and velocity. Neuroiogy, 25, 1065-1070. Baloh, R. W., Konrad, H. R., Sills, A. W. % Honrubia, V. (1975b). The saccade velocity test. Neurology, 25, 1071-1076. Becker, W. & Fuchs, A. F. (1969). Further properties of the human saccadic system: Eye movements and correction saccades with and without visual fixation points. Vision Research, 9, 1247-1258.

Bittencourt, P., Lloyd, D., Richens, A., Smith, A. T., Toseland, P. A. & Wade, P. (1981). Differential effect of alcohol on eye movement measures in man. Journal of Physiology (Proceedings),

3Z2, 48P.

Boghen, D., Troost, 8. T., DarotT, R. B., DeIl’Osso, L. F. & Birkett, J. E. (1974). Velocity characteristics of normal human saccades. Znvestigative Ophlhaimo~ogy, 1.3, 619-623.

Bronstein, A. M. & Kennard, C. (1987). Predictive eye saccades are different from visually triggered saccades. Vision Research, 27, 517-520.

Collewijn, H., Erkelens, C. J. & Steinman, R. M. (1988). Binocular co-ordination of human vertical saccadic eye movements. Journal of Physiology, London, 404, 183-197. Cook, G., Stark, L. % Zuber, B. L. (1966). Horizontal eye movements studied with the on-line computer. Archives of Ophfha~mology, 76, 589-595.

Crawford, T. J. (1984). The modification of saccadic trajectories. In Gale, A. G. & Johnston, F. (Eds.), Theoretical

Research Note and applied aspects ofeye movement research (pp. 95-102).

Amsterdam: Elsevier. Franck, M. C. & Kuhlo, W. (1970). Die Wirkung des Alkohols auf die raschen Blickzielbewe8ungen (Saccaden) beim Menschen. Archiv fir Psychiatric Nervenkrankheiten, 213, 238-245.

Riggs, L. A., Merton, P. A. & Morton, H. B. (1974). Suppression of visual phosphenes during saccadic eye movements. Vision Research, 14, 997-1011. Robinson, D. A. (1964). The mechanics of human saccadic eye movements. Journal of Physiology, London, 174, 245-264.

Fricker, S. J. (1971). Dynamic measurements of horizontal eye motion-I. Acceleration and velocity matrices. Zneestigative ~~th~mo~ogy,

1443

14 724-?32.

Fuchs, A. F. & Binder, M. D. (1983). Fatigue resistance of human extraocular muscles. Journal of N~rophysiology, 49, 28-34.

cine, 50, 393395.

Sharpe, J. A. & Zackon, D. H. (1987). Senescent saccades. Acta Otoiaryngo~ogy, Stockhoim, I&, 422-428.

Gentles, W. & Llewellyn-Thomas, E. (1971). Effect of benzodiazepines upon saccadic eye movements in man. Clinical and Pharmacological Therapy, 12, 563-574.

Henriksson, N. G., Pyykk6, I., Schalbn, L. & Wenmo, C. (1980). Velocity patterns of rapid eye movements. Acta Otolaryngology, 89, W-512. Hess, C. W., Miiri, R. & Meienberg, 0. (1986). Recording

of horizontal saccadic eye movements. Methodological comparison between electro-oculography and infrared reflection oculography. Neuro-ophthalmoiogy, 6,189-198. Hyde, J. E. (1959). Some characteristic of humau voluntary ocular movements in the horizontal plane. American Journal of Opht~~moiogy,

Schmidt, D., Abel, L. A., Dell’Osso, L. F. & Daroff, R. B. (1979). Saccadic velocity characteristics: Intrinsic variability and fatigue. Aviation Space and E~iro~nta~ Medi-

48, 85-94.

Brain Research,

76, 64-74.

Smit, A. C., van Gisbergen, J. A. M. & Cools, A. R. (1987). A parametric analysis of human saccades in different experimental paradigms. Vision Research, 27, 1745-l 762.

Sonderegger, E. N., Meienberg, 0. & Ehrengruber, H. (1986). Normative data of saccadic eye movements for routine diagnosis of ophthalmoneurolo~~ disorders. Nero-o~tha~rno~ogy~

Ishikawa, S. & Terakado, R. (1973). Maximum velocity of saccadic eye movements in normal and strabismic subjects. Japanese Journal of Ophthalmology, 17, 11-21. Katoh, Z. (1988). Slowing effects of alcohol on voluntary eye movements. Aviation Space and Environmental Medicine, 59, 606-610. Uirner, F. H. (1975). Non-visual control of human saccadic eye movements. In Lennerstrand, G. & Bach-y-Rim P. (Eds.), Basic mechanisms of ocular motility and their clinicat implications (pp. 565-569). Oxford: Pergamon. Leigh, R. J. & Zee, D. S. (1983). The neuroIogy of eye mavements. P~adelp~a: Davis. Lueck, C. J., Tanyeri, S., Crawford, T. J., Henderson, L. & Kennard, C. (1990) Antisaccades and remembered saccades in Parkinson’s disease. Journal of Neurology, Neurosurgery and Psychiatry, 53, 284-288.

Reulen, J. P. H., Marcus, J. T., Koops, D., de Vries, F. R., Tiesinga, G., Boshuizen, K. & Bos, J. E. (1988). Precise recording of eye movement: The IRIS technique, part 1. Medical and Biological Engineering 20-26.

Smit, A. C. & van Gisbergen, J. A. M. (1989). A shortlatency transition in saccade dynamics during squarewave trackin and its significance for the differentiation of visually-guided and predictive saccades. Experimental

and Computers, 26,

6, 257-269.

Spooner, J. W., Sakala, S. M. & Baloh, R. W. (1980). Effect of aging on eye tracking. Archives of Neurology, 37, 575-576.

Tedeschi, G., Casucci, G., Allocca, S., Riva, R., Di Costanzo, A., Quattrone, A., Baruzzi, A. & Bonavita, V. (1989). Computer analysis of saccadic eye movements: Assessment of two different carbamazepine formulations. European Journal of CIintcal Pharmacology, 37, 513-516.

Warabi, T., Kase, M. & Kato, T. (1984). Effect of aging on the accuracy of visually guided saccadic eye movement. Anna& of Neurology, 16, 449-454. W~inson, I. M. S., Kime, R. & Purnell, M. (1974). Alcohol and human eye movement. Brain, 97 785-792. Zahn, J. R., Abel, L. A. & Dell’Osso, L. F. (1978). The audio-ocular response characteristics. Sensory Processes, 2, 32-37. Zambarbieri, D., Schmid, R., Magenes, G. & Prablanc, C. (1982). Saccadic responses evoked by presentation of visual and auditory targets. Experimental Brain Research, 47, 417427.