RESEARCH ARTICLE. Postural Sway During Single and Repeated Cold Exposures

Transcription

1 RESEARCH ARTICLE Postural Sway During Single and Repeated Cold Exposures Tiina M. Mäkinen, Hannu Rintamäki, Juha T. Korpelainen, Ville Kampman, Tiina Pääkkönen, Juha Oksa, Lawrence A. Palinkas, Juhani Leppäluoto, and Juhani Hassi MÄKINEN TM, RINTAMÄKI H, KORPELAINEN JT, KAMPMAN V, PÄÄKKÖNEN T, OKSA J, PALINKAS LA, LEPPÄLUOTO J, HASSI J. Postural sway during single and repeated cold exposures. Aviat Space Environ Med 2005; 76: Introduction: Tissue cooling changes sensory and neuromuscular functions that are also involved in postural control. The purpose of the study was to determine how acute and repeated exposures to cold affect whole body postural control. Methods: Postural sway was measured from 10 subjects during standing with eyes open (EO) and closed (EC) using an inclinometer-based method. Sway was assessed at at 10 C on 10 consecutive days and at 25 C on days 1, 5, and 10. Sway path length, area, velocity, side-to-side and forward-backward movement were assessed. At the same time, rectal and skin temperatures, muscle tonus/ shivering, thermal sensations, and comfort were recorded. Results: Acute exposure to 10 C caused thermal discomfort, significantly lowered ( C) mean skin temperatures, slightly lowered rectal temperature (36.7 C) and increased ( %) muscle tone, increased sway path length (67 87%, p 0.05), velocity (63 71%, p 0.05), total sway area (42 67%, p 0.05), and forward-backward movement (35 57%, p 0.05) compared with 25 C. Side-to-side movements were not altered in the cold. Postural sway increased with EC, and further when exposed to cold, but the effect of cold was smaller compared with EO. Repeated exposures over the 10-d period decreased sway 10 40% both at 25 C and at 10 C (p ), suggesting motor learning. The difference in sway between 25 C and 10 C remained the same throughout the 10-d period, suggesting that the observed cold habituation responses do not affect sway. Conclusions: The results demonstrate that postural control is impaired in cold, which may affect physical performance in cold environmental conditions. Keywords: postural control, cold strain, cold acclimation, habituation, thermoregulation, human. POSTURAL CONTROL is an essential element of human daily activities. Sufficient postural control is important in dynamic activities, such as physically demanding occupations (21). An impaired balance may result in decreased performance and injuries resulting from slipping, tripping, or falling accidents. Control of human posture is a complex phenomenon. Maintaining postural stability in the field of gravity requires that the center of mass falls within the area of support. This area is relatively small, requiring constant fine-tuning of movements in the different joints to maintain posture. Sensory information of the body s posture is gained through visual, somatosensory, and vestibular systems. The afferent information is integrated at the spinal cord, medulla, midbrain, and cerebral cortex. Finally, postural control is obtained by preprogrammed anticipatory postural adjustments, muscle reflexes, peripheral elasticity of muscles and tendons, as well as preprogrammed and voluntary corrections (13). Cold exposure may affect postural control through a variety of mechanisms. The cold environment itself, with icy surfaces and a reduced amount of light during the winter, can endanger postural stability (4). Different physiological responses related to cooling may also affect postural control. For example, shivering may affect postural control due to increased muscle tone. It is not known whether this increased tension in muscles has a beneficial or disadvantageous effect on sway. When cooling progresses, the muscle tone is changed into tremor and associated with visible shaking or shuddering. It is possible that shivering causes perturbations in fine motor control (17), requiring more tuning of movements compared with a warm environment. Cooling also affects the sensory systems involved in postural control. For example, the ankle mechanoreceptors are important sensory components for maintaining balance. Previous studies examining the functional properties of the sole and ankle mechanoreceptors have demonstrated that local cooling of feet increases postural sway (15,16,24). The proprioceptors located in the muscles, tendons, and joints can also be affected by cooling, resulting in changes in neuromotor functions. Cooling may, for example, decrease the activity of the muscle spindles, leading to suppression of tendon-reflex amplitudes, consequently affecting neuromuscular control (19). The neural transmission of both afferent and efferent information may be slowed due to cooling From the Centre for Arctic Medicine, University of Oulu (T. M. Mäkinen, J. Hassi), the Department of Physiology, University of Oulu (H. Rintamäki, T. Pääkkönen, J. Leppäluoto), the Finnish Institute of Occupational Health (H. Rintamäki, J. Oksa), the Department of Neurology, University of Oulu (J. T. Korpelainen), the Microelectronics and Material Physics Laboratories and EMPART Research Group of Infotech Oulu, University of Oulu (V. Kampman), Oulu, Finland; and the Department of Family and Preventive Medicine, University of California, San Diego, CA (L. A. Palinkas). This manuscript was received for review in April It was accepted for publication in July Address reprint requests to: Tiina M. Mäkinen, Centre for Arctic Medicine, Thule Institute, University of Oulu, P.O. Box 5000, FIN University of Oulu, Oulu, Finland; tiina.makinen@oulu.fi. Reprint & Copyright by Aerospace Medical Association, Alexandria, VA. 947

2 of the nerves (23). It is also well known that cooling of a muscle impairs most of its functional properties. For example, an increased coactivation of muscle pairs may result in impaired muscular coordination (20). This could render the fine-tuning of movements when maintaining balance more difficult. Finally, cooling has been shown to increase the viscosity in the synovial joints (8). This increased stiffness may also affect regulation of body posture. It is well known that repeated exposures to cold result in acclimation. Depending on the type and intensity of cold exposure, acclimation responses may develop within a couple of weeks (14). Habituation, which is a form of cold acclimation, results in dampened physiological responses, e.g., less intense sensations of cold and discomfort, reduced vasoconstriction and BP, higher skin temperatures, delayed onset and reduced intensity of shivering, and diminished release of circulating stress hormones (27). It is possible that these changes may affect postural control. Because tissue cooling changes sensory and neuromuscular functions, the purpose of this study was to examine how single and repeated cold exposures affect whole body postural control. The hypothesis was that superficial cooling, especially of the peripheral areas, might impair postural control due to changes in sensory or neuromuscular functions. The second hypothesis was that repeated exposures to cold, causing changes in thermoregulation (cold habituation responses), may improve postural control. This could occur, for example, due to dampening of shivering and a reduced need for corrective movements. METHODS The experimental protocol was approved in advance by the Northern Ostrobothnia Hospital District. Test subjects were 10 young healthy men who volunteered. All of the subjects were students. They were informed of the nature, purpose, and possible risks/inconvenience caused by the experiment. A medical examination was conducted to confirm that they were healthy (e.g., excluding subjects who were hypersensitive to cold). A written consent to participate in the study was obtained from each subject before participating in the experiments. The anthropometric and physical characteristics of the study subjects were (mean SD): age yr (range 20 25), height cm (range ), weight kg (range 59 87), BMI (range 19 24), fat % (range 13 19), V O2 max ml min 1 kg 1 (range 44 65). On arriving for the tests, the subjects were equipped with thermistors. During the experiment they were lightly clad in shorts, socks, and athletic shoes. The experiments were conducted between 09:00 11:00 and 13:00 15:00 and two subjects were measured each day. The sway measurements were conducted at the same time of the day for each subject. To control for possible learning, the tests were performed under thermoneutral conditions on days 1, 5, and 10 in a climatic room (13 m 2 ) in which the temperature was C. The total duration of the stay at 25 C was 1.5 h and the control sway measurement was performed at the end of the stay in the thermoneutral conditions. After this the subjects were exposed to cold ( C) in a climatic chamber (27 m 2 )for2h d 1 on 10 successive days. The sway measurements were performed each day in the cold after 90 min of exposure. In both of these climatic chambers the relative humidity was 50 3% and the air velocity less than 0.2 m s 1. The acute effect of cold on postural control could be determined by comparing sway at 25 C to 10 C (days 1, 5, 10). By examining the magnitude of change in sway between 25 C and 10 C over the 10-d period it was possible to determine whether changes in thermoregulation affected sway. Sway A novel inclinometer-based method was used in this study (Body Sway Measurement System, Crea Research, Oulu, Finland). The advantage of this method was that it measured the absolute movements of the body, both momentary and cumulative values. By using this device it was also possible to detect the multiple simultaneous movements occurring in the different joints of the lower extremities. Furthermore, rotational artifacts could be avoided due to its special joint structure. This method has proven to be an accurate and repeatable method and useful in clinical applications (11,26). The device consisted of a belt fastened firmly at the level of the sacrum, an inflexible measuring rod, an inclinometric module, a special joint structure located on the ground, a power unit, and a PC. The measuring rod transmitted movements of the body to the detecting inclinometric module. The measurement resolution of this module was less than 0.5 mm and the high cut-off frequency was 25 Hz. The height of the measuring rod was adjusted according to the estimated center of mass (0.55 height). Prior to the experiments the subjects were familiarized with the equipment and measurement protocol. During the sway measurements the subjects stood at attention with their feet together and arms beside the body and keeping their eyes successively open (EO) or closed (EC). In the EO test subjects were asked to look straight ahead at a fixation point on the facing wall 4 m in front of them. The tests were performed under quiet and well-lit circumstances and repeated once. The duration of each measurement was 1 min. The recorded sway parameters were: the total path length of postural movements (at the level 0.55 subject s height), maximum deflection for x (side-to-side) and y (forwardbackward) sway (cm), velocity (cm s 1 ) and total sway area (cm 2 ). A more detailed basis for the calculation of these parameters is described elsewhere (11,26). Thermoregulation Skin temperatures were measured from 10 sites using thermistors (NTC DC 95, type 2252OHM, Digi-Key, Thief River Falls, MN): forehead, upper back, chest, abdomen, upper arm, lower arm, back of the hand, anterior thigh, and dorsal side of the foot and calf. The thermistors were calibrated in a temperature bath prior to use. Mean skin temperature (Tsk) was calculated as an 948

3 TABLE I. POSTURAL SWAY AT 25 C AND 10 C ON DAYS 1, 5, AND 10 (N 10, MEAN SE) WITH EYES OPEN (EO) AND EYES CLOSED (EC). Day 1 Day 5 Day C 10 C 25 C 10 C 25 C 10 C EO length (cm) a a b area (cm 2 ) a a delta y (cm) a delta x (cm) vel. (cm s 1 ) a a b EC length (cm) c a,c c c c area (cm 2 ) a,c c delta y (cm) delta x (cm) vel. (cm s 1 ) c a a c vel. velocity. a significantly different from 25 C, p 0.05; b significantly different from 25 C, p 0.01; c significantly different from EO, p area weighed average from the 10 different sites (6). Foot (Tfoot) and calf (Tcalf) temperatures were separately examined when analyzing the relationship between skin temperatures and sway. Foot temperature represents a temperature measured underneath the sock (thus not directly exposed to the ambient conditions). Rectal temperature (Trect) was measured 10 cm beyond the anal sphincter with an YSI401 probe (Yellow Springs Instrument Co., Yellow Springs, OH). Skin and rectal temperature were recorded at 1-min intervals with a datalogger (SmartReader Plus8, ACR Systems, Surrey, B.C., Canada). Means of skin and rectal temperatures were calculated for the time period when the sway measurements were performed (total duration 5 min). Muscle tonus was assessed by measuring surface EMG activity (model ME6000, Mega Electronics, Kuopio, Finland) from m. pectoralis. The measurements were conducted at the same time when the sway tests were performed. The electrodes were placed above the belly of the muscle. The distance between the recording contacts was 2 cm. Ground electrodes were attached above inactive tissues. Raw EMG signal (sampling rate 250 Hz) was recorded continuously throughout the exposure. The signal was amplified and the signal band between Hz was full wave rectified and averaged (aemg) with a 40-ms time constant. Butterworth filtering was used in the Hz measuring band. The common mode rejection rate of the amplifier was 110 db. Thermal perception for the whole body, trunk, hands, and feet were assessed using a 9-degree subjective judgment scale (ISO 10551) (9). Thermal comfort was assessed according to the same method. Statistical Analyses The effect of exposure period on sway, skin and rectal temperatures, and shivering were tested using a repeated measures ANOVA (within factor test time: days 1 10) separately for the two temperature conditions and for EO and EC. If a significant main effect was observed, means of separate days were compared with day 1. To control for multiple comparisons, the observed p-values were adjusted using the SAS Multitest procedure and the Bonferroni method. The effect of temperature (25 C vs. 10 C) was tested by paired samples t-tests. The change in sway between 25 C and 10 C was calculated and means of this change between day 1, 5, and 10 were compared by paired t-tests. A Bonferroni correction (SAS Multitest procedure) was applied to the observed p-values. Medians of thermal sensations were calculated. The effect of the temperature exposure on thermal sensations was examined by Wilcoxon s signed rank tests. The effect of exposure period on thermal sensations was examined using the Kendall s W test. Correlations between sway and thermoregulation were examined by Pearson s correlation tests. Significance was set at p RESULTS During acute exposure to cold sway was significantly higher compared with 25 C (Table I). For example, at 10 C, path length representing a cumulative value of sway increased by 62 87% (days 1, 5, and 10, EO) and 51 65% (days 1, 5, and 10, EC) compared with 25 C. The total sway area increased 42 67% (days 1, 5, and 10, EO) and 10 49% (days 1, 5, and 10, EC) at 10 C compared with 25 C. Velocity increased by 63 71% (days 1, 5, and 10, EO) and 50 64% (days 1, 5, and 10, EC) compared with 25 C. Furthermore, forward-backward movement increased 35 57% (days 1, 5, and 10, EO) and 18 39% (days 1, 5, and 10, EC) at 10 C compared with 25 C. Side-to-side-movement was not affected by cold exposure. Closing the eyes increased path length, sway area, and velocity of the movement both at 25 C and 10 C compared with EO (Table I). Furthermore, when the subjects closed their eyes only path length, velocity, and total area were significantly higher at 10 C compared with 25 C. This effect was abolished over the 10-d exposure period and no significant differences between 10 C and 25 C in any of the sway parameters were observed by day 10 with EC. When examining the main effect of repeated exposures at 10 C (EO), it was observed that forward-backward (F 2.165, p 0.05) and side-to-side (F 3.384, p 0.01) movements, as well as the total area of sway 949

4 Fig. 1. Sway during repeated exposures to 10 C (n 10, mean SE) when eyes are open. The plots represent A) total sway area (cm 2 ), B) path length (cm), C) side-to-side movement (cm), and D) forward-backward (cm) movement. * Significantly different from day 1, p (F 4.922, p 0.01), decreased significantly during the 10-d period (Fig. 1). The reduction in sway over the 10-d exposure period at 10 C was not so clear when eyes were closed. Only side-to-side (F 3.620, p 0.01) movements were significantly lower. In the cold the percent reduction in sway parameter over the 10-d exposure period (EO and EC, days 5 and 10 compared with day 1) was: total sway area 31 44%; mean path length 18 20%; mean forward-backward movements 18 21%; and mean side-to side movements 18 27%. When examining changes in sway at 25 C over the 10-d period, the reduction in sway was significant for forward-backward (F 4.952, p 0.05), total area of sway (F 4.176, p 0.05), and path length (F 5.479, p 0.05). When the eyes were closed, sway path length (F 5.702, p 0.05) and velocity (F 5,717, p 0.05) were significantly reduced. At 25 C the percent reduction in sway parameters over the 10-d exposure period (days 5 and 10, EO and EC compared with day 1) was: total sway area 21 45%; mean path length 16 21%; mean forward-backward movements 8 29%; and mean side-to side movements 9 20%. When comparing the change in sway between the two temperature conditions (25 C vs. 10 C) on days 1, 5, and 10, no significant differences were observed. Means of the measured Trect, Tsk, Tfoot, and Tcalf during the sway measurements (length of measurement period 5 min) are presented in Table II. At 10 C Tsk was C lower compared with 25 C (p 0.01). Tsk did not change significantly at 25 C during the 10-d period. However, at 10 C a significant increase (F 4.686, p 0.05) in Tsk was observed over time. Tsk was significantly higher during days 6 (p 0.05) and 10 (p 0.05) compared with day 1 (Fig. 2). Tsk correlated positively with path length (r 0.723, p 0.05) and with total sway area (r 0.764, p 0.05) on day 7. Trect was 0.3 C lower at 10 C compared with 25 C but did not change significantly over the 10-d exposure period either at 25 C or 10 C. Furthermore, Tfoot and Tcalf did not change significantly over the 10-d exposure period. Tcalf corre- TABLE II. TRECT, TSK, TFOOT, TCALF, AND AVERAGED EMG (aemg; V) ACTIVITY DURING SWAY ON DAYS 1, 5, AND 10 AT 25 C AND 10 C. THE DATA ARE MEAN SE (N 10). Day 1 Day 5 Day C 10 C 25 C 10 C 25 C 10 C Trect a a Tsk b b a,c Tfoot b b b Tcalf b b b aemg b b a a significantly different from 25 C, p 0.05; b significantly different from 25 C, p 0.01; c significantly different from day 1, p

5 lated with sway path length on days 9 (r 0.709, p 0.05) and 10 (r 0.748, p 0.05); with area on days 7 (r 0.729, p 0.05), 8 (r 0.705, p 0.05), and 10 (r 0.778, p 0.01); and with forward-backward sway on day 10 (r 0.756, p 0.01). Cold-related general thermal sensations became significantly less intense over the 10-d exposure period (Kendall s W 0.183, df 9, p 0.05). The general thermal sensations of the subjects were rated as cold (day 1) or cool (days 5 and 10) (Table III). The corresponding general thermal sensations at 25 C were neutral (day 1) or slightly warm (days 5 and 10), but the difference over the 10-d exposure period at 25 C was non-significant. The local thermal sensations in hands and feet did not change significantly either at 25 C or 10 C during the 10-d exposure period. In the cold, thermal comfort was rated as uncomfortable (day 1) or slightly uncomfortable (days 5 and 10) and did not change significantly during the 10-d period. TABLE III. THERMAL SENSATIONS AND COMFORT ON DAYS 1, 5, AND 10 AT 25 C AND 10 C (AFTER 90 MIN EXPOSURE). THE VALUES REPRESENT MEDIANS (N 10). Day 1 Day 5 Day C 10 C 25 C 10 C 25 C 10 C Thermal sensation* General 0 3 c c 1 2 a,c Trunk 0 3 c c 1 2 b Hands 0 3 c 1 3 c 1 3 c Feet c 1 1 b 1 1 b Thermal comfort 0 2 c 0 2 c 0 1 c *Thermal sensation: 1 slightly warm, 0 neutral, 1 slightly cool, 2 cool, 3 cold. Thermal comfort: 0 comfortable, 1 slightly uncomfortable, 2 uncomfortable. a Significantly different from day 1, p 0.05, Kendall s W 0.183; b significantly different from 25 C, p 0.05; c significantly different from 25 C, p Muscle Tonus/Shivering Cold increased the muscle tone by % compared with 25 C and aemg was significantly higher at 10 C compared with 25 C. In some of the subjects visible shivering was observed during the sway tests. The aemg amplitudes were 2 4 V at 25 C and 7 12 V at 10 C (Table II). Muscle tonus did not change significantly over the 10-d exposure period either at 25 C or 10 C. Muscle tonus correlated positively (EO) with path length on days 7 (r 0.648, p 0.05) and 8 (r 0.661, p 0.05). DISCUSSION The present study used a cold exposure which induced cold sensations and discomfort, a significantly lowered skin temperature ( C), increased muscle tonus ( %), and a slight reduction in rectal temperature (0.3 C). The main finding was that this level of cold exposure caused a significant increase in postural sway, indicating an impaired postural control. To our knowledge this is the first study to describe Fig. 2. Mean skin temperature (Tsk) during repeated exposures to 10 C (n 10, mean SE). * Significantly different from day 1, p changes in whole body postural control at quiet standing during single and repeated exposures to cold. Environmental stress and postural control were studied in a previous investigation (2) examining the effects of simultaneous multiple environmental stressors (e.g., cold, hypoxia, and fatigue) on postural control during a disabled submarine simulation. This study observed a decrement in postural stability after continued exposure (several days) to these stressors. However, the effect of cold exposure as an independent predictor of changes in sway was not possible to distinguish in this study. Other studies assessing cold and postural stability have focused on local effects such as the functional properties of ankle mechanoreceptors on postural stability (15,16,24). These studies showed that cooling of the feet to the level of numbness increased sway significantly, suggesting that the input from feet mechanoreceptors contributes significantly to postural control. Upper-body postural stability during cold exposure has also been examined separately, for example, during marksmanship (25). The effects on performance have varied due to different study protocols. Some of these studies have shown that cold strain has no effect (25), or even slightly improves (12) upper body stability. Upper body fine motor control during cold strain was also studied by Meigal et al. (17), who found a decrease in accuracy of a shoulder flexion task when the subjects were shivering. It is well known that visual information is essential for postural control. The present study showed that postural sway increased by fold both at 25 C and 10 C when the eyes were closed. This agrees with previous findings that the increase in sway can be almost two-to-threefold when the eyes are closed in women and men of different ages (3,5,22). The effect of cold exposure on postural sway was smaller in the EC compared with the EO test. Path length, velocity, and total sway area remained higher at 10 C compared with 25 C in the beginning of the 10-d exposure period. These parameters are reliable estimates of sway, as they represent the cumulative values compared with sideto-side and forward-backward movements, where momentary lapses may affect the overall results. However, 951

6 the additional effect of cold diminished over the acclimation period. Why the effect of cold on sway was lesser during EC might be due to the fact that the lack of visual information already increased sway, partially masking the effects of cold. However, the fact that closing the eyes further increased sway in the cold may be significant when considering activities performed outdoors in the winter, where darkness may significantly decrease visibility. We were interested in whether possible changes in thermoregulation occurring over the cold acclimation period, causing cold habituation, would affect postural control. The cold acclimation protocol used in this study is known to induce habituation responses (14). In fact, we observed a significantly higher Tsk ( 0.4 C) and less intense general sensations of cold at the end of the 10-d exposure period. These are characteristic habituation responses obtained after repeated exposures to cold (14,27). However, we did not observe a decrease in muscle tonus/shivering over the 10-d exposure period, which is related to cold acclimation and habituation of shivering (27). This can be partially explained by the fact that muscle tonus/shivering, which was measured in association with the sway tests, might have been momentarily suppressed due to focusing on the test. In fact, a reduction in EMG amplitude could be observed in some subjects (data not presented). It has been demonstrated that shivering may be voluntarily suppressed during short-term tasks, for example, breath-holding and arithmetical tasks (10). Sway decreased in the course of the 10-d exposure period. Depending on the sway parameter, the decrement in sway over the exposure period ranged between approximately 10 40% both at 25 C and 10 C. The difference in sway parameters between 25 C and 10 C remained relatively the same throughout the 10-d exposure period. This suggests that the observed changes in thermal responses (e.g., habituation) over the 10-d exposure period did not explain the decrease in sway under cold conditions, and implies that motor skill learning occurred to the same extent under both conditions. The effect of learning in repeated balance tests has shown divergent results. In some studies no learning could be observed (1). Other studies have demonstrated similar types of learning effects under thermoneutral conditions as we observed in our study (7,18). Previous studies (7), and our own experience (not published), show that there is a rapid reduction in sway after the first consecutive tests, after which no additional learning occurs. Also we observed that most of the sway parameters did not change significantly between the fifth and tenth day of exposure, indicating that no further learning occurred. One limitation of the present study was that we did not train the subjects in order to achieve a stable plateau before starting the experiments. This may have affected the absolute levels of sway to some extent, especially in the beginning of the 10-d period. However, the values at the end of the acclimation period represent a stable situation. Furthermore, by comparing sway between the two different temperature conditions during the same day of exposure allows us to determine the main effect of cold on postural control. Although cooling generally increased sway, individual changes in thermal responses over the 10-d exposure period were not consistently associated with changes in sway. When examining these individual responses on some of the exposure days (at the end of the acclimation period), a positive correlation between Tsk, Tcalf, Tfoot, and sway was observed, suggesting that skin cooling and a reduction in sway may be associated in a dose-dependent manner. This is against our hypothesis, but it could be hypothesized that during the end of the acclimation period a reduction in the sympathetic tonus could result in muscular relaxation (reduced muscle tone) and an increased sway. However, this causation remains speculative and needs further examination. Muscle tone and/or shivering did not consistently correlate with sway; a positive correlation was only observed on some of the days. This suggests that visible shivering may cause perturbations in balance, which results in more corrective movements to maintain posture. However, the use of the EMG results may again be limited, as muscle tonus could have been suppressed due to focusing on the postural stability test. The finding of increased sway in the cold may be important to recognize during leisure time or occupational activities performed in cold environmental conditions. Persons who are at a higher risk of falling may be especially susceptible to the effects of cooling due to changes in their postural control. These population groups are, for example, elderly people and/or persons having an impaired balance due to a neurological or musculoskeletal disorder (4,11). The risk of falling is further aggravated in cold conditions due to slippery surfaces and limited visibility (4). To conclude, we observed that cold exposure, which induces cold thermal sensation and thermal discomfort, a significantly lowered skin temperature, and increased muscle tone, increases postural sway considerably, indicating an impaired postural control. It is possible that subjects have to perform more corrective movements in the cold due to changes in their sensory functions or neuromuscular control. Postural sway increased when the eyes were closed, and even further when exposed to cold. However, the additional effect of cold on sway was lesser compared with when the eyes were kept open. Although repeated exposures decreased sway both at 25 C and 10 C, suggesting motor learning, the difference in sway between these two temperature conditions remained the same throughout the 10-d exposure period, suggesting that the observed cold habituation responses do not affect sway. Further studies are needed to examine the mechanisms related to the impaired postural control in the cold, especially the effects of cold on static and dynamic postural control in population groups who are at a higher risk of falling. ACKNOWLEDGMENTS For statistical analyses we consulted statisticians Mr. Jari Jokelainen (Department of Public Heath Sciences and General Practice, University of Oulu) and Mr. Jouko Remes (Finnish Institute of Occupational Health/Oulu), who we want to thank for their assistance. We would 952

7 like to thank the test subjects for their dedication to this study. The experiments performed during this study comply with the current laws of Finland. The study was supported by the Graduate School of Circumpolar Wellbeing, Health and Adaptation coordinated by the Centre for Arctic Medicine at the University of Oulu. REFERENCES 1. Black FO, Wall C, Rockette HE, Kitch R. Normal subject postural sway during the Romberg test. Am J Otolaryngol 1982; 3: Cymerman A, Young AJ, Francis TJ et al. Subjective symptoms and postural control during a disabled submarine simulation. Undersea Hyperb Med 2002; 29: Era P, Heikkinen E. Postural sway during standing and unexpected disturbance of balance in random samples of men of different ages. J Gerontol 1985; 40: Gao C, Abeysekera J. A systems perspective of slip and fall accidents on icy and snowy surfaces. Ergonomics 2004; 47: Gill J, Allum JH, Carpenter MG, et al. Trunk sway measures of postural stability during clinical balance tests: effects of age. J Gerontol A Biol Sci Med Sci 2001; 56:M Hardy JD, DuBois EF. The technic of measuring radiation and convection. J Nutr 1938; 15: Holliday PJ, Fernie GR. Changes in the measurement of postural sway resulting from repeated testing. Agressologie 1979; 20: Hunter J, Kerr EH, Rider RA. The relation of joint stiffness upon exposure to cold and the characteristics of synovial fluid. Can J Med Sci 1952; 30: ISO Ergonomics of the thermal environment: assessment of the influence of the thermal environment using subjective judgement scales. Geneva: International Standards Organization; Israel DJ, Wittmers LE, Hoffman RG, Pozos RS. Suppression of shivering by breath holding, relaxation, mental arithmetic, and warm water ingestion. Aviat Space Environ Med 1993; 64: Korpelainen R, Kaikkonen H, Kampman V, Korpelainen JT. Reliability of an inclinometric method for assessment of body sway. Technology and Health Care 2005; 13: Lakie M, Villagra F, Bowman I, Wilby R. Shooting performance is related to forearm temperature and hand tremor size. J Sports Sci 1995; 13: Latash ML. Neurophysiological basis of movement. Urbana, IL: Human Kinetics; Leppäluoto J, Korhonen I, Hassi J. Habituation of thermal sensations, skin temperatures, and norepinephrine in men exposed to cold air. J Appl Physiol 2001; 90: Magnusson M, Enbom H, Johansson R, Pyykko I. Significance of pressor input from the human feet in anterior-posterior postural control. The effect of hypothermia on vibration-induced body-sway. Acta Otolaryngol 1990; 110: Magnusson M, Enbom H, Johansson R, Wiklund J. Significance of pressor input from the human feet in lateral postural control. The effect of hypothermia on galvanically induced body-sway. Acta Otolaryngol 1990; 110: Meigal AY, Oksa J, Hohtola E, et al. Influence of cold shivering on fine motor control in the upper limb. Acta Physiol Scand 1998; 163: Nordahl SHG, Aasen T, Dyrkorn BM, et al. Static stabilometry and repeated testing in a normal population. Aviat Space Environ Med 2000; 71: Oksa J, Rintamäki H, Rissanen S, et al. Stretch- and H-reflexes of the lower leg during whole body cooling and local warming. Aviat Space Environ Med 2000; 71: Oksa J, Rintamäki H, Mäkinen T, et al. Cooling-induced changes in muscular performance and EMG activity of agonist and antagonist muscles. Aviat Space Environ Med 1995; 66: Punakallio A. Balance abilities of different-aged workers in physically demanding jobs. J Occup Rehabil 2003; 13: Røgind H, Lykkegaard JL, Bliddal H, Danneskiold-Samsøe B. Postural sway in normal subjects aged years. Clin Physiol Funct Imaging 2003; 23: Rutkove SB. Effects of temperature on neuromuscular electrophysiology. Muscle Nerve 2001; 24: Stal F, Fransson PA, Magnusson M, Karlberg M. Effects of hypothermic anesthesia of the feet on vibration-induced body sway and adaptation. J Vestib Res 2003; 13: Tikuisis P, Keefe AA, Keillor J, et al. Investigation of rifle marksmanship on simulated targets during thermal discomfort. Aviat Space Environ Med 2002; 73: Viitasalo MK, Kampman V, Sotaniemi KA, et al. Analysis of sway in Parkinson s disease using a new inclinometry-based method. Mov Dis 2002; 17: Young AJ. Homeostatic responses to prolonged cold exposure: human cold acclimatization. In: Fregly MJ, Blatteis CM, eds. Section 4: environmental physiology, vol 1, handbook of physiology. New York: Oxford University Press; 1996: