PLEASE SCROLL DOWN FOR ARTICLE. Full terms and conditions of use:

Transcription

1 This article was downloaded by: [Swets Content Distribution] On: 19 January 2010 Access details: Access Details: [subscription number ] Publisher Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: Registered office: Mortimer House, Mortimer Street, London W1T 3JH, UK Ergonomics Publication details, including instructions for authors and subscription information: Effects of wearing aircrew protective clothing on physiological and cognitive responses under various ambient conditions Hilde Færevik a ; Randi Eidsmo Reinertsen a a Department of Health and Work Physiology SINTEF Unimed N-7465 Trondheim Norway. To cite this Article Færevik, Hilde and Reinertsen, Randi Eidsmo(2003) 'Effects of wearing aircrew protective clothing on physiological and cognitive responses under various ambient conditions', Ergonomics, 46: 8, To link to this Article: DOI: / URL: PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

2 ERGONOMICS, 2003, VOL. 46, NO. 8, Effects of wearing aircrew protective clothing on physiological and cognitive responses under various ambient conditions HILDE FÆREVIK* and RANDI EIDSMO REINERTSEN Department of Health and Work Physiology, SINTEF Unimed, N-7465 Trondheim, Norway Keywords: Protective clothing; Heat stress; Cognitive performance; Pilots. Heat stress can be a significant problem for pilots wearing protective clothing during flights, because they provide extra insulation which prevents evaporative heat loss. Heat stress can influence human cognitive activity, which might be critical in the flying situation, requiring efficient and error-free performance. This study investigated the effect of wearing protective clothing under various ambient conditions on physiological and cognitive performance. On several occasions, eight subjects were exposed for 3 h to three different environmental conditions; 08C at 80% RH, 238C at 63% RH and 408C at 19% RH. The subjects were equipped with thermistors, dressed as they normally do for flights (including helmet, two layers of underwear and an uninsulated survival suit). During three separate exposures the subjects carried out two cognitive performance tests (Vigilance test and DG test). Performance was scored as correct, incorrect, missed reaction and reaction time. Skin temperature, deep body temperature, heart rate, oxygen consumption, temperature and humidity inside the clothing, sweat loss, subjective sensation of temperature and thermal comfort were measured. Rises in rectal temperature, skin temperature, heart rate and body water loss indicated a high level of heat stress in the 408C ambient temperature condition in comparison with 08C and 238C. Performance of the DG test was unaffected by ambient temperature. However, the number of incorrect reactions in the Vigilance test was significantly higher at 408C than at 238C (p = 0.006) or 08C (p = 0.03). The effect on Vigilance performance correlated with changes in deep-body temperature, and this is in accordance with earlier studies that have demonstrated that cognitive performance is virtually unaffected unless environmental conditions are sufficient to change deep body temperature. 1. Introduction Human error is regarded as a contributing factor in 85% of all aviation crashes (Li et al. 2001). Identifying and reducing factors that have detrimental effects on human performance in the aircraft cockpit environment can therefore improve flight safety. The more specific task of helicopter flying includes a number of factors that degrade human performance; these include vibration, noise, gravitation forces, uncomfortable seats and high temperatures. Several studies have demonstrated an association between ambient heat stress and pilot error outside the laboratory. A study of 500 Israel military helicopter accidents and incidents found that pilot errors were fewest at ambient temperatures of C, with an increased risk at C and the *Author for correspondence. Hilde.Ferevik@sintef.no Ergonomics ISSN print/issn online # 2003 Taylor & Francis Ltd DOI: /

3 Heat stress problems of wearing aircrew protective clothing 781 highest risk at 358C (Froom et al. 1993). Field trials of pilots flying the FR-4C at Shaw AFB in South Carolina, USA also demonstrated that pilot error increases during flights in hot weather (Bollinger and Carwell 1975). Studies of fighter pilots flying the A-10 fighter jet at Davis-Montham Air Force Base in Arizona have demonstrated lower G-tolerance and increased general fatigue in low-level flying on warmer flights (Nunneley and Flick 1981). These studies investigated problems that arise when aircraft are operating in hot areas. This is not a problem in northern Europe. Sea King helicopters operating on the Norwegian coast have a relatively low mean ambient cockpit temperature in winter (18.48C), but which occasionally rises to 408C (Færevik and Reinertsen 1998). This high ambient temperature is caused by the helicopter s large canopy, which produces a greenhouse effect, combined with a heating system, which is located by the pilots feet. In normal aircrew clothing assemblies, the resulting thermal strain might be physiologically acceptable. However, Sea King pilots are obliged to wear survival suits all year around because they are operating in areas with low sea temperatures. Wearing a survival suit results in higher discomfort ratings and significant rises in skin temperatures and sweat rates in helicopter pilots during the winter (Færevik and Reinertsen 1998). Laboratory studies have demonstrated that an ambient temperature of C is needed for thermoneutrality in subjects wearing survival suits (Færevik et al. 2001). Even though heat stress, protective clothing and performance have been thoroughly investigated, the relationships between environmental parameters, physiological status and cognitive performance are still not fully understood. Studies that have looked at the effects on cognitive performance of wearing protective clothing at different ambient temperatures often lack information about how the physiological status of the subject correlates with decrements in performance (Fine and Kobrick 1987, Thornton and Caldwell 1993, Reardon et al. 1998b). Other studies have focused primarily on the physiological consequences of wearing protective clothing in different ambient temperatures, or on limitations on the ability to perform physical work, rather than on mental performance as such (Thornton et al. 1985, White et al. 1991, Scott et al. 1994, Rissanen and Rintama ki 1997). Reviews of cognitive performance under hot conditions have also revealed several shortcomings, such as inadequate pretraining of the subjects on the experimental tasks, the use of unrealistic tasks for assessing performance, poor methodological designs for statistical implications and insufficient duration of exposure to the heat (Kobrick and Fine 1983, Ramsey 1995). On the background of these shortcomings and the problem of Norwegian Sea King pilots described above, the aims of this study were to investigate the effects of wearing aircrew protective clothing under various ambient conditions on physiological and cognitive parameters. Any observed changes in performance should be related not only to the ambient conditions involved but also to changes in physiological measures of heat strain. Such information will aid assessments of aircrew efficiency under various ambient conditions while wearing protective clothing. It was hypothesized that typical cockpit temperatures cause heat stress, which will have detrimental effects on cognitive performance of pilots wearing aircrew protective clothing. 2. Materials and methods Eight male volunteers (including three RNAF pilots) participated in the study. The mean age, weight, stature, percent body fat and surface area (A Du ) of the subjects were years, kg, cm, % and

4 782 H. Færevik and R. E. Reinertsen m 2, respectively. The subjects were asked to retire to bed at their usual time on the night before each exposure. They abstained from taking exercise and from consuming caffeine, alcohol or snuff for 24 h before exposure. None of the subjects were smokers. All subjects were in good health and had recently undergone an electrocardiogram test. The Ethical Review Committee of the Faculty of Medicine at the Norwegian University of Science and Technology approved the experimental procedure. The subjects were free to withdraw from the chamber environment at any time. The experiment was cancelled if either of the following physiological safety limits were exceeded: core temperature above 398C or skin temperature above 428C Experimental protocol On three different occasions, the subjects were exposed for 3 h to three different ambient conditions in an environmental chamber; 08C and 80% relative humidity (RH), 238C and 63% RH, 408C and 19% RH, respectively. The subjects were exposed in a random order to the three ambient conditions so as to avoid any effects of order. The tests were carried out at the same time of the day on different days with at least a 1-day pause between the tests. Total time of exposure to the test conditions was 3 h. Subjects were not permitted to drink or eat during the experiment. Subjects reported to the preparation room at least 1 h before testing. They were equipped with thermistors and heart rate recorder, and then dressed as they normally do for flights, in long legged/long sleeved underwear (200 g Ullfrotte), a woollen whole body cover all, helmet and the uninsulated British Mark 10 survival suit (Beaufort). This survival suit consists of a double-layer of cotton Ventile which permits the transmission of water vapour; once wetted, the fibres expand such that the interfibre spaces no longer transfuse liquid. The insulation value of the whole clothing ensemble was 2.20 Clo, measured on a thermal manikin. The subjects sat quietly outside the climatic chamber for 20 min in order to provide baseline measurements of oxygen consumption (VO 2 ), skin and rectal temperature and subjective evaluations. In order to identify changes associated with the development of thermal stress, performance was measured at intervals throughout the exposure. During the 3 h of exposure in the climatic chamber, subjects carried out an experimental schedule according to the timetable described in table Cognitive variables It was not possible to devise a test procedure capable of fully simulating the actual working conditions of helicopter pilots in an operational situation. Cognitive performance was therefore assessed using the computerized Vienna Test Battery (Schufried 1990) (figure 1). This test battery has been designed to test vigilance and multiple-choice reactions, and is used in the selection of pilots for the Royal Norwegian Air Force. The subjects were required to perform two cognitive tests; the Vigilance test and the Vienna Determination Unit test (DG test) to measure reactive stress tolerance. The Vigilance test requires subjects to exercise continuous vigilance over a lengthy period of time. A bright dot moves in small jumps along a circular path. Occasionally, the dot jumps twice the usual distance, to which event subjects must react by pressing a reaction button. The following variables were scored; increase/decrease in reaction times, total number of incorrect/correct reactions and missed reactions. The DG test is intended to evaluate the ability of subjects to perform multiple-choice reactions to rapidly changing stimuli and to detect

5 Heat stress problems of wearing aircrew protective clothing 783 Table 1. Experimental schedule followed during exposure to the three ambient conditions, 08C at 80% RH, 238C at 63% RH and 408C at 19% RH. Session Elapsed time (min) Event Before start 7 20 Subjective assessment of thermal comfort Baseline values of skin and rectal temperatures 0 Subjects enter the environmental chamber DG test Subjective evaluation Oxygen consumption Vigilance test Subjective evaluation Oxygen consumption DG test Subjective evaluation Oxygen consumption Vigilance test Subjective evaluation Oxygen consumption DG test Subjective evaluation Oxygen consumption Vigilance test Subjective evaluation Oxygen consumption impairments in attentive capacity. Subjects are required to respond to visual or acoustic stimuli by pressing a button or stepping on a foot pedal. The visual stimuli consist of white, yellow, red, green and blue signals, the acoustic stimuli of high and low tones. Performance is measured in terms of the number of correct (timely/ delayed), incorrect and missed reactions as well as reaction time. The test lasts for 13 min and consists of four subtests with the same stimuli but presented in different orders and at different speeds. Before the actual study in the environmental chamber, all subjects did 10 pilot test runs on the Vienna test battery in a thermoneutral environment to ensure that a learning plateau was reached and minimize the learning effects during the three environmental exposures Physiological variables Rectal temperature (T re ) was measured by a thermistor probe (YSI-700, Yellow Springs Instrument, USA, accuracy C) inserted 10 cm beyond the anal sphincter. Skin temperatures were measured using thermistors (YSI-400 Yellow Springs Instrument, USA, accuracy C) at six different locations (forehead, chest, upper arm, left hand, front thigh and front leg). Mean skin temperatures (MST) were calculated using the formula of Ramanathan (1964). Heart-rate (f c ) was recorded by a Polar Sports Tester (Polar Electro, Finland). All parameters were recorded at 1-min intervals, and graphically and numerically displayed on a computer screen to have continuous information about the thermal state of the subject while the experiment was running (software: Templog 3.1). The means of 10 min were used for statistical analysis. Oxygen consumption (VO 2 ) was logged for 5 min using a Cortex MetaMax Portable Metabolic Test System (Cortex Biophysic GmbH, Germany). A questionnaire developed by Nielsen et al. (1989) was used to

6 784 H. Færevik and R. E. Reinertsen Figure 1. Set-up of the Vienna test battery in the laboratory. obtain information about overall thermal sensation, sensation of skin and clothing wetness, ambient temperature preference and thermal comfort. Total body weight loss was obtained by weighing the subjects before and after the experiment. Body surface in square metres (A Du ) was calculated using the following formula (DuBois and DuBois 1916): A Du = W b H b 0.725, where W b is the body weight in kg, and H b is the body height in m. Percentage body fat was calculated using the Durnin and Wommersley 4-site skinfold thickness measure (Durnin and Wommersley 1974). Ambient conditions (air temperature, radiation and relative humidity) were measured continuously with an Indoor Climate Analyser T1213 (Bru el & Kjær A/S, Denmark). Microclimate inside the clothing was measured using sensors (type HIH-3605-B-CP, Honeywell, USA,) located in the upper back region. The sensors were attached inside the inner layer of clothing, between the inner and the middle layer and between the middle and the outer layer. All logged data were transferred to a computer and the results were displayed every minute on the computer screen to have continuous information about the microclimate of the clothing while the experiment was running (software: Templog 3.1) Statistical analysis The difference in performance response under the different ambient conditions was assessed by two-way analyses of variance (ANOVA) for repeated measures. SPSS

7 Heat stress problems of wearing aircrew protective clothing (SPSS Inc. Chicago, USA) was used to process the statistical material. A within group study design was used. All performance data were tested for effects of time (session 1,2,3), ambient condition, and the interaction between these two. When ANOVA revealed a significant main effect, a contrast test was used as a post hoc test to locate significant differences between temperatures. For the DG test there were four subtests in each session. The subtests were collapsed in the analysis, since any time effect from test duration of only 13 min was not expected. Time-dependent changes in rectal temperature, mean skin temperatures, oxygen consumption and heart rate were also evaluated by two-way analyses of variance (ANOVA) for repeated measures, using a contrast test as a post hoc test. Differences in sweating rates and the individual ratings of thermal comfort, thermal sensation, degree of shivering or sweating, sensation of skin wetness, temperature and humidity in the clothing and ambient temperature preference were assessed by Student s t-test for paired samples. The Shapiro Wilk s test was used to test for normal distribution. Spearman s test was used to find correlations between changes in performance and the physiological measurements/subjective ratings. Correlation tests were performed within each series. Results are presented as means + SD for eight subjects. All differences reported are significant at the p level. 3. Results 3.1. Rectal temperature (T re ) The rectal temperature (T re ) fell slightly at both 08C and 238C ambient temperature (T a ), by 0.68C and 0.38C respectively (figure 2) when the start and end values of the Figure 2. Development of mean rectal temperatures in the three test conditions: 08C, 238C and 408C. Þ * Indicates significantly higher rectal temperature at 408C thanat08c after 40 min, and after 70 min than at 238C. The difference lasted throughout the test period (n=8).

8 786 H. Færevik and R. E. Reinertsen whole test period were compared. At 408C (T a )T re rose significantly by 1.28C to 38.48C. The time-dependent change in T re revealed a significant effect of ambient condition. When the time courses of the changes in T re were compared, these were significantly higher in the 408C (T a ) condition than at 08C (T a ) after 40 min. After 70 min the T re at 408C was also significantly higher than at 238C (T a ). These differences lasted throughout the period of testing. There were no differences in rectal temperature between 08C and 238C Mean skin temperature (MST) The change in MST from the initial value on entering the climatic chamber was greater when subjects were exposed to 408C than to 238C or08c (figure 3). MST was significantly different after only 20 min and throughout the rest of the test in all three series. After 3 h in the climatic chamber MST rose by 1.98C from its initial value when subjects exposed to 408C (T a ). MST was almost unchanged (0.38C) at 238C (T a ), and fell significantly at 08C (T a ), by 2.98C Heart rate (f c ) Heart rate also demonstrated significant effects of both time and ambient condition. The increase in mean f c was significantly higher at 408C(T a ) than at 238Cor08C(T a ) (figure 4). Heart rate rose significantly in the whole 408C series, ending at 120 beats min 71. In both 23 and 08C (T a ), f c decreased slightly, ending at values of 76 and 65 beats min 71. The heart rate was significantly higher after 30 min and throughout remainder of the test at 408C (T a ) than at either 0 or 238C (T a ). There were no differences in f c between 08C and238c. Figure 3. Development of mean skin temperature in the three test conditions. Þ*Indicates significantly higher MST at 408C after 20 min than at 238C or08c. The differences lasted throughout the test period (n=8).

9 Heat stress problems of wearing aircrew protective clothing 787 Figure 4. Development of heart rate in the three test conditions. Þ*Indicates significantly higher heart rate at 408C after 30 min than at 238C or08c. The observed differences lasted throughout the test period (n=8) Sweat production and body water loss Losses of body water through sweat and evaporation were considerably higher at 408C (T a ) than 238C or 08C, with values of g, g and g, respectively (figure 5). The subjects were significantly dehydrated in the 408C condition, with a mean weight loss of 1.2 kg, equivalent to 1.5% of their total body weight Oxygen consumption Mean oxygen consumption (VO 2 ) under the three ambient conditions was on average l min 7 1 at 08C, l min 7 1 at 208C and l min 7 1 at 408C(T a ). This corresponds to mean heat production rates of W m 7 2, W m 7 2 and W m 7 2 respectively. The change in VO 2 from the initial value was significant only at 08C and 408C(T a ). VO 2 rose by 0.16 l min 7 1, 0.03 l min 7 1 and 0.10 l min 7 1 at 08C, 208C and 408C (T a ), respectively. The change in VO 2 from the baseline level at 08C (T a ) was after 137 min and for the remainder of the test. The change in VO 2 from baseline that was observed at 408C(T a ) was not significant until 177 min (the last measure). The time-dependent rise in oxygen consumption did not differ between the three exposure conditions Subjective assessments and microclimate in the clothing After only 15 min, subjects felt significantly warmer when exposed to 408C (T a ) than at 238C or and 08C. While they felt only slightly warm when exposed to

10 788 H. Færevik and R. E. Reinertsen Figure 5. Total body water loss during the 3 h of exposure in the three test conditions 08C, 238C and 408C. Þ*Indicates considerably higher water loss at 408C than at 08C or 238C (n=8). 238C, they stated that they were hot at 408C (T a ). Thermal sensation of the feet, hands and head were significantly warmer at 408C (T a ) than at 238C or08c (T a ) after 15 min and during the remainder of the test. The subjects felt significantly colder at 08C than at 238C (T a ), and were thermally more uncomfortable at 408C than at either 08C or238c (figure 6). This was noted after only 15 min in the climatic chamber. They remained more uncomfortable throughout the test. Subjects stated that they started sweating after only 15 min when exposed to 408C; after 55 min they were sweating moderately, and their clothing felt damp. At 155 min the clothing felt wet and they stated that they were sweating heavily. This is in contrast to the 238C (T a ) condition, where it was not until 135 min that the subjects stated that they were sweating slightly, and that their clothing felt slightly damp. At 08C (T a ) the clothing felt dry throughout the tests, and no sensation of sweating was felt at all. This is in accordance with the results of the clothing moisture measurements. In the 408C series, the inner and middle layers of the clothing were 100% saturated with moisture after 110 min, and this condition persisted throughout the rest of the experiment. The outer layer reached 80% saturation after 121 min. In the 238C series, less moisture accumulated in the inner layer, but even so 80% was reached after 130 min. In the 08C series the moisture was transported from the inner layer outwards in the clothing ensemble, resulting in a fall in moisture content in the inner layer from 60% at the beginning of the test to 30% at the end Performance data The mean scores of the performance tasks are summarized in table 2. In the Vigilance test ANOVA showed no effects of time or ambient condition or interaction

11 Heat stress problems of wearing aircrew protective clothing 789 Figure 6. Time course of subjective ratings of thermal comfort. Þ*Indicates significant differences in comfort between 408C and 238C conditions, and between 408C and 08C after 15 min and for the remainder of the test period (n=8). between them in the responses for reaction time, number of correct reactions or number of missed reactions. The only significant effect of ambient condition in the Vigilance test was in the total number of incorrect reactions for the whole time series (figure 7). The average number of incorrect reactions was 1.44 in the 408C series; this was significantly higher compared to 0.80 at 08C (p = 0.03) and 0.89 at 238C (p = 0.006). There was no difference between 08C and 238C (p = 0.62). Incorrect reactions were unaffected by time and there was no interaction between time and ambient condition. In the more complex cognitive DG test there was no significant effect of ambient condition, but a clear effect of time on some parameters (table 2 and figure 8). There were no interactions between temperature and time. The total effect of time on performance parameters was observed in terms of fewer incorrect reactions in Session 1 than in Sessions 2 (p = 0.032) or 3 (p = 0.009), and faster reaction times in Session 1 than in Session 2 (p = 0.007) or 3 (p = 0.012). Performance measured as missed and delayed reactions in the DG test was unaffected by time. Performance scores were examined not only as they were affected by ambient condition and time, but also in terms of how changes in performance correlated with physiological and subjective changes. The only significant correlation was in the Vigilance test. Spearman s test showed that the increase in the number of incorrect reactions in the Vigilance test in the 408C series clearly correlated with the change in core temperature (1.28C increaseint re ) (Spearman s correlation coefficient: 0.907, p = 0.002) (figure 9). There was no such correlation at 238C or 08C. The increase in incorrect reactions did not correlate with any of the other physiological parameters or with subjective assessments of thermal sensation and comfort.

12 790 H. Færevik and R. E. Reinertsen Table 2. Cognitive performance in the Vigilance and DG tests. Mean number of reactions and mean reaction time are shown for each session (n = 8). * 1 Indicates an effect of ambient condition, observed in a higher total of incorrect reactions in the Vigilance test at 408C than at 08C(p = 0.03) or 238C (p = 0.006). * 2 Indicates a total effect of time in the DG test; fewer total incorrect reactions in Session 1 than Session 2 (p = 0.032) or 3 (p = 0.009). * 3 Indicates a slower reaction time in Session 1 than Session 2 (p = 0.007) or Session 3 (p = 0.012). Ambient temperature and relative humidity 08C, 80% RH 238C, 63% RH 408C, 19% RH Task Test variable Session 1 Session 2 Session 3 Session 1 Session 2 Session 3 Session 1 Session 2 Session 3 Vigilance Mean number of correct reactions Mean number of incorrect reactions * * * 1 Mean number of missed reactions Mean reaction time (second) DG test Mean number of correct in time reactions Mean number of correct delayed reactions Complex Mean number of incorrect reactions * * * Cognitive Mean number of missed reactions tasks Mean reaction time (second) * * *

13 Heat stress problems of wearing aircrew protective clothing 791 Figure 7. Total number of incorrect reactions, missed reactions and reaction time in the Vigilance test. Þ*Indicates significantly more incorrect reactions at 408C than at 08C (p = 0.03) or 238C (p = 0.006) (mean + SD) (n = 8). Figure 8. Total number of delayed, incorrect and missed reactions and reaction time in the DG test. (mean + SD) (n=8).

14 792 H. Færevik and R. E. Reinertsen Figure 9. Correlation of total incorrect reactions with change in rectal temperature. Spearman s Correlation Coefficient (0.907) indicates a clear correlation (p = 0.002). (n =8). 4. Discussion The aim of this study was to investigate the effect of wearing aircrew protective clothing under different ambient conditions on physiological and cognitive responses. There was a clear effect of wearing protective clothing in the hot (408C) condition on physiological parameters, as was demonstrated by raised rectal and skin temperatures, heart rate and degree of dehydration. The subjective discomfort was also high at 408C. Wearing protective clothing at 408C also lowered performance in some tasks. The study aimed to offer insight into the physiological or subjective parameters with which such performance decrement correlates. In support of current theories, it was demonstrated that performance does not deteriorate with time unless there is a significant change in deep-body temperature. The actual physiological thermal state of the subject s body seems to be of more importance than the thermal sensation of the body or thermal comfort. Furthermore, decrements in performance were dependent on the type of task performed. In the following paragraphs, decrements in performance are discussed in terms of the physiological status of the subject, subjective evaluations, the clothing worn, subject skills, the nature of the task, and the relevance of the test results to real life conditions. Higher ambient temperature in the 408C series is capable of statistically explaining the increase in incorrect reactions seen in the Vigilance test, but more interesting is the question of the physiological mechanisms on which this performance decrement is based. Wearing protective clothing tends to hinder evaporative cooling, resulting in an increase in heat storage in the body (Sullivan and Mekjavic 1992). The

15 Heat stress problems of wearing aircrew protective clothing 793 physiological responses of the body of this uncompensable heat stress situation were clearly demonstrated in raised rectal and skin temperatures, greater cardiac output, high rate of sweating and dehydration. The mean metabolic rate varied between 68.9 and 82.9 W m 7 2 in the study, but was not significantly different between the ambient conditions. The contribution of metabolic rate to the increase in core temperature during the experiment is regarded as being relatively slight. Metabolic rate measured during flight varies, depending on the different phases of flight and type of aircraft, between 48 and 116 W m 7 2 (Thornton et al. 1984). Mean metabolic rates in Sea King helicopter pilots have been measured at 88.8 W m 7 2 (Færevik and Reinertsen 1998). Rectal temperature rose to 38.48C when subjects were exposed to 408C, and the upper limit of core temperature recommended for pilots (388C) was exceeded after only 125 min in the climatic chamber. Even though the test lasted for 3 h, ANOVA revealed no effect that could be exclusively attributed to time on performance in the Vigilance test, but did reveal a clear effect of ambient condition. The correlation between the change in T re in the 408C ambient condition and the total number of incorrect reactions in the Vigilance test suggests that it is the dynamic change in T re, not the duration of the test, which causes performance decrements. This observation confirms the suggestion that the value of the ambient temperature is not the key factor that dictates whether a breakdown in performance will occur. Rather, it is the combination of ambient temperature with exposure time, sufficient to change deep body temperature, which affects performance (Grether 1973, Hancock 1981). Hancock (1981) claimed that higher levels of heat stress are required before any decrements in performance can be observed when exposure is less than 1 h. In support of this, Grether (1973) found that the ambient temperature at which individuals can maintain adequate performance is very close to the threshold temperature at which the body could compensate physiologically for the thermal strain. The relationship between thermoregulatory changes and performance is complex, and several hypotheses have been advanced that try to explain reductions in mental performance on the basis of thermoregulatory changes. Reviews of early studies have explained the effects of heat on performance on the basis of heat balance theory (Hancock 1986). According to this theory, performance should be better in a neutral environment when the body is in heat balance and there is a high gradient between core and skin temperatures. When this gradient is shallow, as in hot conditions, performance would be poor. Carlson (1961) proposed a theory of information overload when skin and core temperatures are both high, and suggests that physiological and psychological sources of input sum a single structure through a common channel, and that high skin and core temperature causes information overload. This is in agreement with the present study; in which rises in the number of incorrect reactions in the Vigilance test were observed only when core and skin temperatures were high in the 408C ambient condition. Other studies have demonstrated that both the rate and direction of change in deep body temperature affect performance (Allan et al. 1979, Gibson and Allan 1979). These studies demonstrated that decrements in performance were greater when body temperatures were high and rising than when they were low and falling. This is in agreement with the present study, which demonstrated no decrements on performance when T re fell by 0.38C or 0.68C, but clear decrements when T re rose by 1.28C. Although vigilance decrement is a common finding, in this study vigilance was not affected by time, subjective reports of thermal discomfort or dehydration. This absence of decrement in vigilance might indicate that thermal discomfort is not a very good measure of

16 794 H. Færevik and R. E. Reinertsen distraction, or that the test subjects are a selected cohort of highly motivated persons trained to be more able to withstand such distraction over time. The effects of increased thermal strain at 408C (T a ) did not have the same detrimental effect on performance in the more complex cognitive DG test as was observed in the Vigilance test. This might indicate that change in body temperature represents a form of distraction affecting simpler tasks, but increasing the level of arousal in more complex tasks (DG-test). The cognitive load of the DG battery is regarded as high, and a long period of training was required to achieve a consistent level of competence. This may have contributed to the relatively minor performance changes found. It is possible that the performance of such a well-learned test is difficult to disrupt, and this is supported by the fact that the total number of missed, incorrect and delayed reactions was generally very low. Hancock et al. (1995) have indicated that a key element in performance during heat stress is that the more experienced subjects are with the task they perform, the less likely they are to be disturbed by thermal stress. It is also possible that subjects may be able to put extra efforts into maintaining adequate performance, particularly since they were aware of the duration of the experiments. Another possible explanation is that the higher level of arousal and concentration required in this test might counteract the effect of increased body temperature. However, further studies will be needed to deal with these questions more thoroughly. Even though there were no effects of ambient condition in the DG test, there was a clear effect of time in some parameters. Fewer incorrect reactions and a slower reaction time in Session 1 than in Sessions 2 and 3 might be explained by the initial arousal effect of entering the climatic chamber. Such arousal or stimulating effects may very well cause an initial improvement in performance (Ramsey 1995). The hypothesis of an initial arousal effect is supported by that time-dependent changes were only observed in the DG test performed immediately on entering the chamber and not in the Vigilance test which was performed 20 min later. Much of the literature is concerned with the effects of thermal stress on mental performance in subjects not wearing protective clothing, and indices of heat stress in the surroundings (air temperature, radiant temperature, and relative humidity and air velocity) are frequently used to estimate the heat stress experienced by workers. However, wearing protective clothing will lower the external limits for safe ambient conditions. For a fully suited worker with only the head exposed (relative surface area of the head is approximately 7%) as much as 93% of the body surface area will be exposed to the suit microenvironment (Sullivan and Mekjavic 1992). This greatly impedes the exchange of moisture between the environment and the skin. Heat and moisture accumulate inside the clothing and, Sullivan and Mekjavic (1992) emphasized, the microenvironment inside the clothing is therefore of more importance than ambient conditions only. Sullivan and Mekjavic (1992) compared four different protective garments for helicopter pilots. The important finding of this study was that regardless of the heat load imposed, the true level of heat stress was dependent on the evaporative properties of the clothing, as was demonstrated by different increases in T re for the four different protective garments (varying from 0.28C to 1.28C). This further emphasizes the importance of the design and evaporative properties of the protective clothing worn. The double layer of cotton Ventile fabric in the British Mark 10 used in the present study is intended to permit the transmission of water vapour. When exposed to 08C the moisture was transported outwards in the underclothing. This moisture transport is favourable,

17 Heat stress problems of wearing aircrew protective clothing 795 in that it lessens the thermal sensation of wetness of the skin (Bakkevig and Nielsen 1994). However, the evaporative heat loss at 408C was not sufficient, resulting in 100% saturation in the inner and middle clothing layers, heat accumulation and increased physiological stress. It is not unlikely that a protective garment with better evaporative properties would have alleviated the thermal stress and lowered T re. This in turn could have had consequences for the cognitive performance, with fewer incorrect reactions in the Vigilance test. Detrimental effects of heat stress when wearing protective clothing on flight performance have been demonstrated in the UH-60 helicopter simulator (Reardon et al. 1998a). Caldwell et al.(1997) showed that aviators could not safely fly a single standard mission in the UH-60 helicopter simulator wearing protective clothing at 418C, without some type of cockpit or individual cooling garment. Most of the body water loss in the 408C ambient condition accumulated in the clothing layers. The total mean weight loss of 1.2 kg represented 1.5% of the subjects mean body weight, a level of dehydration that is associated with a reduction in physical work capacity (Saltin 1985). More important for the terms of this study, a 1.2% level dehydration has been shown to decrease G tolerance and significantly affect pilot performance (Bollinger and Carwell 1975, Gillingham and Winter 1976, Nunneley and Stribey 1979, Nunneley et al. 1995). In contradiction to these results, this study found no association between the dehydration level and performance effects. A possible explanation for this could be differences in the study design and the tasks to be performed. Even though it was not found that dehydration had an effect on performance, the discomfort caused by sweating or increases in the subjective sensation of warmth might well induce performance decrements. If the performance measures were a measure of comfort, a correlation between performance scores and ratings of environmental comfort would be expected. Spearman s correlation coefficient showed no such association. Enander (1984) stated that the relationship between subjective sensation, comfort and performance is by no means clear, and there is little reason to expect the optimum levels of any of these measures to coincide. This is in accordance with the findings of Griffiths and Boyce (1971), who specifically studied the relationship between performance, thermal sensation and comfort. They found that correlations between performance and subjective assessments were lower than between performance and temperature. They conclude that there is no evidence of a linear relationship between performance and subjective assessment of thermal comfort. This finding suggests that even though the feeling of thermal discomfort and sensation of warmth during Sea King helicopter flights is high (Færevik and Reinertsen 1998), this is not a critical factor for flight performance. However, the transfer of results from laboratory testing of cognitive tasks during heat stress to real life situations must not be done without due consideration. Several authors have emphasized the difficulty of comparing studies on cognitive performance because of differences in the tasks performed, prior training, severity and duration of exposures, and differences in the characteristics and motivation of subjects (Kobrick and Fine 1983, Ramsey 1995). Enander and Hygge (1990) claimed that setting a standard for limitations to heat stress must be done on the basis of individual differences. The subjects used for this study were highly motivated medical students or pilots with 3 to 6 years of experience. This group may be an extremely select cohort that was better able than most to withstand the detrimental

18 796 H. Færevik and R. E. Reinertsen effects of heat stress. Still, the choice of test subjects is relevant to the actual working situation. Even though a high level of motivation may to some extent counteract the detrimental effects of heat stress, performance motivation might well be reduced over several days of testing. Differences in levels of motivation between individual test subjects on different days were counteracted by the randomized test design. Furthermore, all subjects performed the test at the same time of the day on each occasion in order to counteract any effect of the time of the day the tests were performed. This study demonstrated that performance deficits were related not only to time and temperature conditions, but also to the type of task that subjects were asked to perform. The Vigilance test turned out to be more vulnerable to heat than the DG test. This is consonant with other findings that have demonstrated that certain cognitive tasks are more profoundly affected by high temperature than others; so called heat stress selectivity (Grether 1973, Hancock 1982). Performance decrement effects of heat on simple cognitive and perceptual motor tasks are rarely reported, even with rises in body temperature (Ramsey 1995). More complex tasks, particularly involving vigilance, have proved to be more vulnerable to thermal stress (Hancock 1986). Vigilance tasks have traditionally been characterized as tedious; however investigations has revealed that such tasks can be quite demanding and induce considerable stress on those who perform them (Hancock and Warm 1989). The Vigilance test used in the present study required the subjects to monitor a display for the appearance of a critical signal to initiate an important response. This low activity monitoring is much like the issue in the context of helicopter flying, where monotonous, low-level, sustained periods of attention are interspersed with highly stimulating, high-activity periods. Vigilance tasks represent an important class of functions in aviation and are especially important for helicopter pilots (Hancock et al. 1995). Sea King helicopters pilots have search and rescue operations that may last for up to h, and this requires continued sustained attention. The decrement in Vigilance response could possible be more pronounced in the Sea King helicopter, where pilots face a combination of interacting factors such as vibration, noise, gravitation forces, uncomfortable seats and high temperatures. Although Vigilance has been shown to be of relevance to helicopter flying, the next question is whether real-life ambient conditions are sufficient to cause the same physiological heat stress demonstrated in the 408C ambient condition in the laboratory. In a winter field study of helicopter pilots, T re of the pilots fell by 0.38C during min-long flights when wearing the same clothing as in the present study at an ambient temperature of C (Færevik and Reinertsen 1998). It is not likely that ambient temperature in the Sea King cockpit during winter flights is sufficient to cause a 1.28C rise in T re. The results of this study are therefore more relevant to flights in warmer climates and seasons. This study suggests that tolerance limits for heat stress ought to take into account the specific skills required in a given working situation, as well as the maximum changes in physiological and psychological responses that can be regarded as harmless. Furthermore, this study suggests that an ambient condition of 238C while wearing protective clothing does not cause any detrimental changes in psychological or physiological parameters and therefore could be used as a safety guideline for helicopter flight operations.

19 Heat stress problems of wearing aircrew protective clothing Conclusions This study demonstrated clear effects of protective clothing during 3 h exposures to ambient condition of 408C, on physiological parameters, as shown by increases in rectal and skin temperature, heart rate and degree of dehydration. This physiological heat stress caused decrements in vigilance, observed in terms of a rise in incorrect reactions to the test stimulus. The dynamic change in core temperature was the only physiological parameter that correlated with decreased performance, which supports the hypothesis that performance is largely unaffected unless there is an increase in deep-body temperature. Furthermore, performance deficits were dependent on the complexity of the performance test, and the Vigilance test proved to be the most vulnerable to heat stress. The results are important in the context of flying, where sustained attention over long periods of time is required, and suggest that a rise in core temperature as a result of wearing protective clothing in hot ambient conditions might mean an increased risk of pilot error. The results of this study challenge designers and producers of protective equipment to put greater efforts into reducing thermal stress during flight in order to prevent a rise in core temperature. Acknowledgements The authors wish to thank Vigdis By Kampenes for her help in statistical analyses. We also wish to thank the Royal Norwegian Airforce for their co-operation in this project and the Norwegian Research Council for the financial support. References ALLAN, J. R., GIBSON, T. M. and GREEN, R. G. 1979, Effects of induced cyclic changes in deep body temperature on task performances, Aviation, Space and Environmental Medicine, 50, BAKKEVIG, M. K. and NIELSEN, R. 1994, Impact of wet underwear on thermoregulatory responses and thermal comfort in the cold, Ergonomics, 37, BOLLINGER, R. R. and CARWELL, G. R. 1975, Biomedical cost of low-level flight in a hot environment, Aviation, Space and Environmental Medicine, 46, CALDWELL, J. L., CALDWELL, J. A. and SALTER, C. A. 1997, Effects of chemical protective clothing and heat stress on army helicopter pilot performance, Military Psychology, 9, CARLSON, L. D. 1961, Human performance under different thermal loads Technical Report No 61-43, Brooks AFB, TX: USAF Aerospace Medical Center, School of Aviation Medicine. DUBOIS, D. and DUBOIS, E. F. 1916, A formula to estimate the approximate surface area if height and weight be known, The Archives of Internal Medicine, 17, DURNIN, J. and WOMERSLEY, J. 1974, Body fat assessed from total body density and its estimation from skinfold thickness: Measurements on 481 men and women aged from 16 to 72 years, British Journal of Nutrition, 32, ENANDER, A. E. 1984, Performance and sensory aspects of work in cold environments: A review, Ergonomics, 27, ENANDER, A. E. and HYGGE, S. 1990, Thermal stress and human performance, Scandinavian Journal of Work and Environmental Health, 16, FÆREVIK, H., MARKUSSEN, D., ØGLÆND, G. E. and REINERTSEN, R. E. 2001, The thermoneutral zone when wearing aircrew protective clothing, Journal of Thermal Biology, 26, FÆREVIK, H. and REINERTSEN, R. E. 1998, Thermal stress in helicopter pilots, evaluation of two survival suits used during flight, In J. A. Hodgon, J. H. Heaney and Buono M. J. (eds), Environmental Ergonomics VIII (San Diego, CA), FINE, B. J. and KOBRICK, J. L. 1987, Effect of heat and chemical protective clothing on cognitive performance, Aviation, Space and Environmental Medicine, 58,