Markovits AS, Reddix MD, O'Connell SR, Collyer PD. Comparison of bifocal and progressive addition lenses on aviator target detection performance. Aviat. Space Environ. Med. 1995; 66:303-8.
The objective of this project was to determine if the type of presbyopic correction worn by aviators, conventional bifocal versus progressive addition lenses (PAL's), differentially affects aviator visual search performance. Experienced aviators, most with tactical fighter aircraft experience, searched for high-contrast targets under simulated dawn/dust (mesopic) lighting conditions (~3.0 cd/m2) while wearing either a standard bifocal (ST-25) or PAL spectacle correction. Latency of locating high-contrast targets under these viewing conditions was differentially affected by the type of presbyopic correction used. Specifically, compared to a standard bifocal (ST-25), a PAL correction (Varilux Infinity®) significantly lowered the time needed to locate static targets at a cockpit instrument viewing distance (83 cm). Accuracy of target location responses was not affected by the type of correction used. In addition, 7 months post-experiment, 7 of the 12 participants (58%) indicated that they used their PAL correction exclusively when flying the T-39 Sabre Liner. Three subjects (25%) used their PAL correction intermittently (primarily at night) when flying and two subjects preferred not to use the PAL's. These results suggest that relative to bifocals, speed of responding to static targets at intermediate viewing distances may be improved by wearing PAL's, and that subjects were able to adapt to PAL lenses quickly in a laboratory setting, using them later in a functional aviation environment.
From the Naval Aerospace and Operational Medical Institute, Pensacola, FL (A. S. Markovits); Naval Aerospace Medical Research Laboratory, Pensacola, FL (M. D. Reddix and P. D. Collyer); and Department of Ophthalmology, Naval Hospital, San Diego, CA (S. R. O'Connell).
This manuscript was received for review in September 1993. It was revised in March and May 1994 and accepted for publication in May 1994.
Address reprint requests to: CAPT Andrew S. Markovits, M.D., B.S., Naval Aerospace and Operational Medical Institute, 220 Hovey Road, Pensacola, FL 32508-1047.
Aviators become presbyopic beginning as early as age 40, as do the vast majority of the population (7). Between 3 and 4% of the U.S. Navy pilots (M. Mittelman, Personal Communication, February, 1993) and 12.4% of U.S. Air Force pilots (20) are presbyopic and require a multifocal-spectacle correction such as bifocals. One would expect the percentage of bifocal wearers to be even greater for aviators in the reserve forces, as the average age of reserve aviators is greater than that of active duty aviators (18). The underlying cause(s) of presbyopia (conceptualized as an age-dependent loss of ocular accommodative ability) remain unclear (3,4,13). Possible causes include, but are not limited to the following: a) changes in the elastic properties of the crystalline lens (8,23,24); b) liquefaction of the vitreous (5); c) alterations in anterior segment geometry (10,14,15,16,21); and d) a combination of lens growth and concomitant anterior chamber shallowing (17).
Bifocals and trifocals are authorized by the USAF and Navy as an acceptable correction for presbyopic aviators. In this project our objective was to determine if the type of presbyopic correction worn by aviators, conventional bifocal versus progressive-addition lenses (PAL's), differentially affects aviator visual search performance. Human performance data, such as speed and accuracy of target detection/identification, provide additional objective measures for use in assessing the suitability of PAL's to an aviation environment. Of course human performance data should not be viewed in isolation of other salient issues such as modulation transfer, psychophysical observations, human factors (including spectacle-induced head movement), and patient adaptability to inherent PAL astigmatism.
Aviators are required to view critical information at a minimum of three accommodation distances (approach plate, cockpit instruments, and infinity). Bifocals, however, have only 2 focal lengths; 20 feet (infinity) and, usually, 40 cm, a fairly standard reading distance. Trifocals on the other hand, although offering a correction for three viewing distances, necessarily have a relatively small intermediate segment (vertically), preventing a full view of the cockpit instruments (because trifocals are almost never user by military aviators, they were not included for study here). In addition, some pilots are uncomfortable with the head movements that are needed to accommodate changes in focal length when wearing multifocal lenses. These conditions have created a reluctance in pilots to wear presbyopic corrective eyewear, and potential flight hazards associated with inefficient or difficult vision.
PAL's overcome some of the inherent shortcomings of many bifocal and trifocal lenses, but not without tradeoffs. For example, compared to a trifocal correction. PAL's increase the combined zone of near and intermediate correction and eliminate visible lines in the lens (Fig. 1). This intermediate zone of changing power (narrow, clear area in PAL lens, Fig. 1), is referred to as the transition channel. The sphericity of the lens is maintained in this region, resulting in high quality image modulation. However, outside the transition zone (shaded area in PAL lens, Fig. 1) image quality suffers because of unavoidable spherical aberrations. The add power of the correction combined with the lens manufacturing technique determine the extent of the transition zone and the magnitude of peripheral aberrations (Fig. 2).
Few objective studies have examined the performance of subjects wearing a PAL correction for presbyopia. Previous research comparing PAL wearers to nonpresbyopic or multifocal controls (i.e., bifocal or trifocal wearers) in simple target detection (1,2,22) and reading tasks (12) suggests that, for near vision (» 40 cm):
PAL's have been given favorable ratings by airline pilots; however, the only study to address the effects of progressive lenses on pilot performance did not use objective measures and did not employ a bifocal or emmetropic control group (6). If progressive lenses improve, relative to bifocals, some aspects of aviator performance (e.g., speed and accuracy of target detection) then the flight community may want to conduct a comprehensive, and objective, investigation of their applicability to in an aeronautical environment compared to other types of presbyopic correction (e.g., bifocals).
The present study was designed to compare bifocal and PAL corrections for presbyopia using a target detection task and three accommodation distances common to the cockpit environment. Speed and accuracy of target detection for both types of presbyopic corrections were compared.
One active duty and eleven retired military aviators participated in this study. The age of subjects ranged from 43 to 60 years (mean = 53, SEM = 1.5). The retired military aviator group are currently flying T-39 Sabre Liners as pilots in command. Ten of the retired military aviators group have piloted tactical fighter aircraft. Ten subjects reported their total logged flight hours. Total flight hours ranged from 4000 to 11500 (mean = 8970, SEM = 771). Volunteer subjects were recruited, evaluated, and employed with the procedures specified in Department of Defense Directive 3216.2 and Secretary of the Navy Instruction 3900.39 series. These instructions are based on voluntary informed consent and meet or exceed the provisions of prevailing national and international guidelines.
Each subject was given a complete ophthalmological examination at the Naval Aerospace and Operational Medical Institute (NAMI). Only subjects showing presbyopia that would normally be corrected with bifocal lenses were considered for participation. Because lens opacity (the clarity of an eye's lens) may cause decrements in visual performance independent of visual acuity (11), the clarity of the lens in both eyes of each subject was assessed before their participation in the study using an Opacity Lensmeter (model 701: Interzeag AG; Schlieren, Switzerland). None of the subjects showed signs of pathological opacity of the lens (cataract).
Cockpit environment: Subjects participated while seated in an A/4 ejection seat located behind an F/15 aircraft windscreen assembly located in a separate room, isolated from the experimenters and data collection equipment. Subjects were visually monitored using a low-light sensitive closed-circuit television camera (Model 6415-2000/0000, Cohu Electronics Division, San Diego, CA). An automated intercom system located near the cockpit allowed the experimenter to maintain voice contact with the subject at all times. Ambient lighting was provided by the projection system and video monitors. Mesopic light levels (» 3 cd · m-2) were maintained.
Visual stimulus array: Subjects viewed computer-generated, visual stimulus arrays (Fig. 3) projected at 3 distances within their forward line of sight. Each projected display consisted of 119 randomly placed distractor rectangles and one target rectangle (60% the size of the distractors). This computer-generated visual array was converted to an analog video signal and then either: a) rear-projected onto a diffused projection screen, using a High Resolution, High Brightness Monochrome Projection Monitor (Model 38-B025030-71, Electro-home Limited, Ontario, Canada) placed 280 cm from the subject; or b) displayed on a 30.5-cm (12-in) video monitor (Sony, Model PUM-1271Q), placed 83 cm in front of the subject; or c) displayed on a 22.8-cm (9 in) video monitor (Burle, Model TC1910A), placed 40 cm in front of the subject. The 30.5 and 22.8-cm video monitors were located inside the windscreen assembly (15º to the left and right, respectively, of the subjects forward line of sight).
There were 40 unique visual arrays generated, each containing 1 target. A small crosshair was located at the center of each display, dividing the display into equal quadrants. Targets occurred equally often in each of the four quadrants at each of five eccentricities (1, 2.4, 3.8, 4.3, and 5.3º) measured from the center of the display. In one experimental session each visual array was presented once, in random order, on each visual display device (VDD: back-projection screen, 22.8- or 30.5-cm video monitor). Only one VDD was illuminated at any time. In addition, visual arrays were never displayed successively on the same VDD, forcing the subject to accommodate to a new focal distance to view the next visual array. Thus, subjects viewed 120 visual arrays in a single experimental session.
A Pritchard Photometer with 6' arc aperture (Model PR-1980A, Photo Research, Burbank, CA) was used to measure the luminance of each target rectangle, the distractor rectangle nearest the target, and the background midway between the target and its closest distractor. These measurements were made for visual arrays appearing on the rear projection screen and both video monitors and were used to compute target-background and distractor-background brightness contrast [(LMax – LMin)/(LMax + LMin)]. Target-background and distractor-background brightness contrast (mean = 0.72, SEM = 0.01; mean = 0.74, SEM = 0.91, respectively) was constant across the three VDD's. The overall luminance of the back-projection screen, 30.5- and 22.8-cm video monitors was approximately equal (2.99, 2.87, and 3.07 cd · m-2, respectively).
Experimental control and data acquisition: Experimental control and data acquisition was under micro-computer control (Compaq Deskpro 386/20, model 60). An analog-to-digital I/O board (model DASCON-1, Metrabyte Corporation, Taunton, MA), multifunction timer (Model CTM-5, DASCON-1, Metrabyte Corporation, Taunton, MA), and solid-state controllers (BRS/LVE, Inc.) were used to monitor subject responses and control the onset and duration of the VDD's, and auditory feedback. A compiled algorithm written in Quick-BASIC source code (Microsoft Corp., Redmond, WA) provided control over the function of these peripheral devices.
Corrective spectacles: Each participant received a pair of corrective PAL's at no cost (Varilux Infinity, Varilux Corporation, Foster City, CA). Each subject was also issued a new pair of bifocal lenses with standard 25-mm segments (ST-25, American Optical). Bifocal lenses offered near correction for reading distance (» 40 cm) but not for intermediate panel distance. Add powers ranged from 1 to 2.25 (mean = 1.917, SEM = -0.11). Due to operational constraints (e.g., flight schedules of most participants conflicted with a strict adaptation protocol) subjects were allowed between 3 d and 4 wk to adapt to their new bifocal and PAL lenses.
Subjects were tested separately. They participated in three practice and two experimental sessions over 5 successive days (one session per day). Subjects sat in the A-4 ejection seat located behind an F.15 cockpit windscreen assembly in a completely darkened room for the first 5 min of each experimental session. At the completion of this adaptation period, the center area of each VDD was illuminated by the word 'GO.' Subjects were told that pressing the display-advance button, held in their nondominant hand, would reveal a visual array on one of the three VDD's in front of them and that their task was to identify the location of a single target rectangle as quickly as possible (without sacrificing accuracy) by pressing one of four response keys. Each response key corresponded to a different quadrant of the visual display. The keys were placed in a 3.5-cm wide by 2.5-cm long grid on an aviator's knee-board. Subjects responded with their dominant hand.
On each trial the display remained on until the subject responded, or for 2.8 s, whichever occurred first. On days 1 and 2 (the first 2 of 3 practice sessions), however, the display remained on for a longer period of time (3.0 and 3.2 s, respectively). Longer display times were needed on these days to facilitate practice. After the subject responded, the word 'GO' reappeared in the center of each VDD indicating that the response had been recorded and the next trial was ready to begin. Displays appeared in quasi-random order such that on the following trial the visual array was displayed on one of the VDD's not viewed on the previous trial. Correct target-location responses were immediately followed by a high-pitched tone, whereas incorrect responses were followed by a low-pitched tone. Subjects viewed two display sets (120 trials each) each day, one while wearing bifocal lenses, and the other while wearing PAL's. The order in which the corrective lenses were worn (bifocal first or PAL first) were counter balanced. Subjects participated for 5 successive days. Each experimental session duration was 15-20 min in length.
Recording of subject response time to locate a target was time-locked to visual display onset. Subjects were shown their performance after each session. Further, on the following day, each subject was shown how his previous day's performance compared to that of the other 11 subjects.
The independent variables in this study were as follows: a) type of presbyopic correction (2 levels: bifocal and PAL); b) target eccentricity (5 levels: 1.0, 2.4, 3.8, 4.3, and 5.3º); and c) accommodation distance (40, 83, and 280 cm). The dependent variables were response accuracy and response latency.
A completely within-subjects repeated-measures analysis of variance (ANOVA) design was used to evaluate the effects of the experimental treatments on the accuracy and latency of target-location responses. Post-hoc pairwise comparisons among means were carried out using Tukey's HSD test at the 0.05 probability level. Speed and accuracy-of-responding data from days 4 and 5 (non-training days) were considered for analysis below. Only correct target-location responses were used in the analysis. The effect of presbyopic correction on target-location performance was examined in a 2 x 3 x 5 way repeated-measures ANOVA (correcting, bifocal and PAL; accommodation distance, 40, 83, and 280 cm; target eccentricity, 1.0, 2.4, 3.8, 4.3, and 5.3º). Type of presbyopic correction and accommodation distance interacted (Fig. 4) to significantly affect latency of target-location responses [F(2,22) = 5.88, p<0.01]. Post-hoc pairwise comparisons between presbyopic correction (bifocal vs. PAL) at each accommodation distance revealed that mean response latency when wearing PAL's was significantly lower (mean = 1992, SEM – 56) compared to bifocals (mean = 2103, SEM = 57) at the intermediate (83-cm) viewing distance. The slightly elevated response latency of the PAL lens at the 40- and 280-cm viewing distances were not statistically significant. No other significant effects involving type of presbyopic correction were observed for latency or accuracy of target-detection responses.
Subjective responses from a post-7-month questionnaire were revealing. Of the 12 participants, 7 (58%) indicated that they used their PAL correction exclusively when flying the T-39 Sabre Liner. Three subjects (25%) used their PAL correction intermittently (primarily at night) when flying, and two subjects preferred not to use the PAL's when flying. There were insufficient data to determine of a common factor such as age, ametropia, or strength of bifocal was responsible for an individual accepting or rejecting PAL's.
High-contrast targets viewed under dawn/dusk lighting conditions (» 3.0 cd · m-2) were located with equal accuracy when wearing bifocal or PAL corrections. However, latency of responding to high-contrast targets under the same viewing conditions was differentially affected by the type of presbyopic correction used. Specifically, compared to a standard bifocal (ST-25), a PAL correction (Varilux Infinity) significantly lowered the time needed to locate high-contrast targets at an intermediate viewing distance (83 cm). These results suggest that subjects were able to adapt to PAL lenses quickly, and that, relative to bifocals, latency of responding to static targets at intermediate viewing distances may be reduced by wearing PAL's.
These results do not address each mechanism by which PAL's may reduce (improve) the latency of target detection at an intermediate viewing distance, or allow generalizing to the presbyopic aviator population at large. Previous research (22) suggests that a combination of eye-head coordination factors, including saccadic gain control, could be involved in the adaptation process. In addition, the clear field of view required for target detection in this study (± 5º) may not have forced subjects to use, or compensate for, the nonspherical portion of the PAL lens. Use of a wider cockpit display would be helpful for addressing this issue. Finally, it should be noted that extent and shape of the area of a PAL lens compromised by nonspherical surfaces varies as a function of both manufacturing process and add power. Consequently, these results can be generalized only to individuals wearing the Varilux Infinity PAL with no greater than 2.25 D add.
Finally, the wearing of any type of corrective spectacle poses a unique set of problems for aviators that must not be ignored when considering applicability to the aviation environment. These include but are not limited to, obstructed field of view, fogging, nasal and ear discomfort, reflections (day or night), excessive frame movement due to G-forces and vibration (19), and increased mean target detection times (25).