RESEARCH ARTICLE

Resistance Index of Frostbite as a Predictor of Cold Injury in Arctic Operations

Hein A. M. Daanen and Norbert R. van der Struijs

DAANEN HAM, VAN DER STRUIJS NR. Resistance index of frostbite as a predictor of cold injury in Arctic operations. Aviat Space Environ Med 2005; 76:1119 22. Introduction: Cold-induced vasodilation (CIVD) is mentioned as a mechanism that may prevent the occurrence of local cold injuries. The magnitude of the CIVD reaction differs considerably between subjects and there were some indications that those subjects with a fast CIVD reaction with high amplitude had a reduced risk for cold injuries. The purpose of this investigation was to determine the magnitude of the nger CIVD reaction in subjects prior to operation in cold areas and to relate these scores to the occurrence of cold injuries. Methods: In order to evaluate the magnitude of the CIVD response, 206 subjects immersed their left middle nger in ice water for 30 min. The Resistance Index for Frostbite (RIF) according to Yoshimura was determined on the basis of the nger skin temperature response. This index ranges from 3 (high risk) to 9 (low risk) depending on the response time and response magnitude. Later, most of the screened subjects deployed as part of a group of 1080 marines for winter operations in Norway. Results: The Caucasian subjects in this study had higher RIF scores than the non-Caucasians (7.0 1.6 vs. 6.1 2.1). The mean RIF was relatively high as compared with a reference group of Japanese male soldiers (6.9 1.7 vs. 5.7 1.7). Unexpectedly, smokers had a higher RIF score than non-smokers did. The RIF score was inversely related to pain during the test. There were 54 marines who suffered cold injuries during training in Norway. Of those, 11 were in the measured group of 206 marines. These subjects had a RIF of 5.2 1.6, as compared with 7.0 1.6 for the remaining subjects, which was signicantly different. Conclusions: The RIF shows considerable differences between subjects. The RIF, determined in a simple lab test, may be related to the risk for cold injuries during operations in the eld. Keywords: frostbite, cold injuries, screening, cold induced vasodilation.

Resistance Index for Frostbite (RIF) was calculated from the CIVD response. To our knowledge no studies are currently available that validate the RIF as an indicator for the risk of getting cold injuries during operations in the cold. Therefore, this study was performed, in which results of a RIF prospective screening test were compared with the possible occurrence of cold injuries during operation in a cold area. Also, for the screening group, we investigated the relation between RIF-score and other factors such as ethnicity and smoking behavior. METHODS Subjects A total of 206 subjects participated in the screening procedure. The subjects were recruits of the Royal Netherlands Navy with a mean stature of 182 7 cm, bodyweight 77 7 kg, and age 20 2 yr. Of the 21 subjects with other ethnicity, 5 were Moroccan, 8 Indonesian, 4 Afro-Caribbean, and 4 had mixed ethnicity. The local ethics committee of TNO Human Factors approved the protocol and the subjects gave their written consent prior to the immersion test. Protocol The measurements were performed in January and February 2000 and 2001 indoors. The mean ambient temperature in the room was 18.7 1.1C (range 16.3C to 20.9C). The subjects reported in groups of no more than 14. They were dressed in combat suits. Each subject received a long cup with ice cubes and cold water. The subjects sat on the oor or on a chair with the cup at the level of the heart. They immersed their left middle nger in ice water for 30 min after application of a type T thermocouple at the ngertip (T-T-28M, TempFrom TNO Defence, Security and Safety, Business Unit Human Factors, Soesterberg (H. A. M. Daanen), and the Royal Netherlands Navy, Ofce of the Surgeon General (N. R. van der Struijs), The Hague, The Netherlands. This manuscript was received for review in July 2005. It was accepted for publication in October 2005. Address reprint requests to: Hein Daanen, Ph.D., TNO Defence, Security and Safety, Business Unit Human Factors, P.O. Box 23, 3769 ZG Soesterberg, The Netherlands; daanen@tm.tno.nl. Reprint & Copyright by Aerospace Medical Association, Alexandria, VA.

URING THE OCCUPATION of Manchuria by Japanese forces in the Second World War, it became clear that some soldiers were more prone to local cold injuries than others. Some Japanese scientists related the occurrence of cold-induced vasodilation (CIVD) to the cold injury risk (18). CIVD is the paradoxical vasodilatory reaction of blood vessels to cold ( 15C), during which a considerable amount of heat is released that may prevent the occurrence of local cold injuries (18). The CIVD-reaction is attributed to paralysis of the muscle wall around the arterio-venous anastomoses (4,6). In 1950, Yoshimura and Iida developed a simple test to determine the risk for local cold injuries using the CIVD reaction (18). The left middle nger was immersed in ice water for 30 min. The ngertip temperature was determined each minute using a small thermocouple at the ngertip. The water was well stirred and ambient temperature was set at about 20C. A

Aviation, Space, and Environmental Medicine Vol. 76, No. 12 December 2005

1119

PREDICTION OF COLD INJURYDAANEN & VAN DER STRUIJS

TABLE I. SCORING SYSTEM OF YOSHIMURA AND IIDA (18) TO DETERMINE THE RESISTANCE INDEX OF FROSTBITE (RIF). Number of points Tmin Tonset Tmean 1 1.5C 12 min 4.0C 2 1.64.0C 811 min 4.17.0C 3 4.1C 7 min 7.1C 1 degree 2nd degree TOTAL
st

TABLE II. SEVERITY AND LOCATION OF THE 57 COLD INJURIES IN 54 MARINES. Hands 7 7 14 Feet 30 11 41 Head 2 0 2 TOTAL 39 18 57

Tmin stands for the minimal nger skin temperature, Tmean equals the average nger skin temperature, and Tonset corresponds to the time at which cold-induced vasodilation (CIVD) starts. All points are added so that the RIF score ranges from 3 (weak reaction to cold) to 9 (strong reaction to cold).

Data Processing The relation between RIF and ethnicity, body mass index, and smoking behavior was investigated using regression analysis (15). A t-test was used to identify differences on the RIF-index between the subjects with cold injuries and the remaining group. RESULTS In 16 out of 206 subjects, the CIVD response was too ambiguous to determine the onset of CIVD. This may have been due to continuous heat transfer to the water (14) or because the nger was accidentally removed from the water in the cup. The mean RIF-score of the remaining 190 subjects was 6.9 1.7 SD. Since one subject did not ll in his smoking behavior, the number of subjects for the analysis was 189. Surprisingly, the 106 non-smokers had an average RIF score of 6.6 1.7, while the 83 smokers had a higher RIF score of 7.2 1.5 (t-test, t 2.6, df 187, p 0.0098). The Caucasians had a higher RIF score (7.0 1.6) than the other ethnicities (6.1 2.1) (t-test, t 2.2, df 188, p 0.031). The pain score was maximal 5 min after immersion and decreased with time (Fig. 1). Pain was well correlated to the RIF score: the lower the RIF, the higher the pain score (p 0.001). The correlation was strongest after 10 min of immersion (0.46), followed by 15 min after immersion (0.36). The average oral temperature of the measured 94 subjects was 36.8 0.2C. There was no signicant relationship between oral temperature and RIF score (r 0.08). Of the 54 marines who suffered cold injuries during training in Norway, 11 of them were in the measured pool of 190 subjects with a clear CIVD response (Fig. 2). These 11 subjects had a RIF of 5.2 1.6, as compared

control, Voorburg, The Netherlands). The subjects were asked to stir the ice water with their immersed nger. Ice was added to the cup as soon as the cubes became too small to maintain a constant water temperature of 0C. Every 5 s the skin temperature of the ngertip was measured and stored in a PC. Since core temperature is known to inuence the CIVD-response (5,7), the oral temperature was measured in a subsample of 94 subjects using a thermistor with an RD-temp datalogger (Omega Technologies Ltd, Broughton Astley, UK). The subject put the thermistor in the sublingual pocket and was asked not to talk or to breathe through the mouth. Every 5 min the subjects rated their pain experience on the Borg-scale (1). Scores are rated 0 or more: 0 (nothing at all), 0.5 (extremely weak), 1 (very weak), 2 (weak), 3 (moderate), 5 (strong), 7 (very strong), and 10 (extremely strong). The RIF was calculated according to Yoshimura and Iida (18). Three variables were used: the lowest nger skin temperature prior to the CIVD response (Tmin); the onset time of CIVD (Tonset); and the mean nger skin temperature from minute 5 to 30 (Tmean). Yoshimura and Iida (18) constructed a scoring system for each variable: 2 points if the score was close to the average of their reference group of 100 Japanese subjects; 1 point if the reaction was considerably worse; and 3 points if the reaction was considerably better (Table I). This system is also used in the current experiment. The values are derived from each individual CIVD response. The onset time was determined from the temperature chart by two investigators and omitted if disagreement occurred. Follow-Up Study In 2002 a group of 1080 marines went to Norway for training in the cold. A part of this group consisted of subjects who had performed the nger immersion test. When cold injuries occurred during training, a medical doctor checked the subjects and the location and severity of the cold injury were recorded. The RIF values of the subjects that suffered cold injuries and participated in the immersion test were compared with the RIF values of the other subjects in the immersion test. A total of 57 cold injuries were observed in 54 marines, which was exceptionally high and probably related to the extreme cold during the training period. The severity and location of the cold injuries of the total group is shown in Table II. 1120

Fig. 1. Pain scores (mean SEM) on the Borg scale vs. time for immersion of a nger in ice water; n 205 for times 525 and n 191 for time 30 min.

Aviation, Space, and Environmental Medicine Vol. 76, No. 12 December 2005

PREDICTION OF COLD INJURYDAANEN & VAN DER STRUIJS training in the cold. Smoking is known to lead to instantaneous vasoconstriction, and we speculate that the vasoconstriction reaction due to cold may be blunted in smokers. The pain ratings during the cold-water immersion test correlated well with the RIF score. This was surprising because the pain was scored in a group and social control was expected to dominate. Marines do not easily call events painful. Pain is known to depend on the phase of the CIVD response (11). The vasoconstriction phase is most painful, and during vasodilation pain decreases. Most subjects are expected to be in the vasoconstriction phase 5 min after immersion, and, therefore, their pain score is high. During immersion, core temperature generally decreases slightly due to heat loss to the water and the CIVD magnitude subsequently decreases. Regression analysis showed that subjects experiencing strong pain after 10 min of immersion had a worse CIVD reaction and thus may have a higher risk for local cold injuries. Thus, pain can be seen as an effective warning signal. It is extremely unlikely that the cold-water immersion test itself might cause any damage to the immersed nger. The nger may feel tender after immersion for several hours, but 3 h of immersion are needed to result in overt non-freezing cold injuries (12). Campbell and Jewitt (2) reported a case of cold-induced angioedema after 4.5 min of hand immersion in 0C water, but propanolol may have triggered this exaggerated response to the cold stress. This case is very rare since the cold pressor test is considered to be quite safe. It has been administered to more than 1700 children ages 4 to 18 yr without known adverse physical or psychological effects (16). Core temperature is well known to affect CIVD. When core temperature drops, CIVD magnitude also decreases. In the experiment, no relation was observed between sublingual temperature and RIF. This observation may be explained by the facts that oral temperature is just a global estimator of core temperature and that

Fig. 2. Schematic overview of the number of involved subjects. Most subjects in the screening test participated in the Norway training. However, the exact number could not be determined.

with 7.0 1.6 for the remaining 179 subjects (Fig. 3). This was signicantly different (t-test, t 3.6, df 188, p 0.001). Three subjects had second-degree injuries of the feet and three others had second-degree injuries of the hands. Three subjects had rst-degree injuries of the hands, and two had rst-degree injuries of the feet. One subject had rst-degree injuries of both the hands and the feet. The injured subjects did not differ from the uninjured group in body mass index (23.1 vs. 23.2 kg m2, t 0.1, df 188, p 0.91) or smoking behavior (40% vs. 44%, t 0.3, df 187, p 0.80). Of the 11 subjects in the injury group, 10 were Caucasians. DISCUSSION The mean RIF-value of the Japanese soldiers was 5.7 1.7 (18) compared with the value of 6.9 1.7 of the mixed Dutch population in the current experiment. Contrary to our observations, Hirai et al. (8) observed that the Japanese had a faster and more pronounced CIVD-reaction than Caucasians. The difference disappeared when subjects were living in the same area, showing not only genetic effects but also the effect of the living environment. The comparison between the old Japanese data and the current experiment suggests that the Dutch marines have relatively good protection against local cold injuries. This may be due to excellent physical condition, health, and food status or differences in history of cold exposure or cold history (4). In our screening group 11% of the subjects were non-Caucasians. Generally, Afro-Caribbeans have a weaker CIVD-response as compared with Caucasians (9). The RIF values of the Caucasians in our study were higher than the other ethnicities (7.0 vs. 6.1). Four out of nine subjects with a RIF value of 3 were non-Caucasian: an African, an Indonesian, and two Moroccans. These ethnic differences cannot be fully subscribed to ambient factors; they partly remain when people who are born in hot climates move to colder areas. The general experience in the Royal Netherlands Navy is that smokers have a higher risk for local cold injuries. The results, however, show that smokers have a better CIVD-reaction than non-smokers. It is important to realize that smoking was not allowed during

Fig. 3. Resistance Index of Frostbite (RIF) values for 11 subjects who suffered cold injuries compared with the other 179 measured subjects. The percentages denote the percentage of cold injuries that occurred for each RIF value. A RIF of 3 to 4 is classied as a weak reaction, 5 to 7 as average, and 8 to 9 as strong (18).

Aviation, Space, and Environmental Medicine Vol. 76, No. 12 December 2005

1121

PREDICTION OF COLD INJURYDAANEN & VAN DER STRUIJS core temperature threshold for hand blood ow varies considerably between subjects. Generally, only comparisons within subjects show a relation between core temperature and CIVD parameters. Scientic evidence for the validity of RIF scores as an indicator for cold injury risk is scarce. Iida (10) and co-workers immersed a ngertip in 10C cooling liquid for 10 min. Freezing occurred only in absence of vasodilation. The best validation is to determine the RIF scores of a large group of people and to check whether people with cold injuries had different RIF scores than those without. This study was a rst attempt. Only 11 subjects were both in the experiment and suffered cold injuries. Even though this group is small, the hypothesis that cold injuries are related to CIVD response could not be rejected. The reference group consisted of the other 179 subjects in the prospective screening test. It is estimated that almost every subject in the screening test went to the cold training in Norway. However, it cannot be excluded that some of them dropped out during the training and did not go. In the discussions during the experiments we noticed that subjects with low RIF values were generally aware of their increased susceptibility to local cold injuries. It is likely that these subjects behave more carefully in cold areas than subjects with a good CIVD response. Still, we observed in our study that the group with cold injuries had a worse CIVD reaction (lower RIF values) than the reference group. In extreme cold and windy climates, freezing cold injuries occur extremely fast, even before CIVD can occur (17). In this case RIF values cannot be considered as a good indicator of cold injury risk. During the operation in Norway, the winds were not extreme, so the occurrence of freezing injuries before CIVD onset was not very likely. In order to be able to use the nger immersion test as an indicator for the risk of frostbite, the test has to be reproducible. Yoshimura and Iida (18) investigated the reproducibility of the RIF score by measuring the CIVD reaction of ve subjects every 2 to 4 d. The maximal variation was 1 point (SD 0.26). This can be considered good reproducibility. Daanen (3) investigated the CIVD-reproducibility in eight subjects and observed an SD in Tonset of about 1 min and an SD in Tmin of 0.7C. OBrien recently observed that the within-subject coefcient of variation is only 19% for onset time, 21% for Tmin and 17% for Tmax (13). She observed that the nail bed showed better reproducibility than the nger pad, and thus may be a better place to measure. However, even for the ngertip, the variation within subjects was always smaller than the variation between subjects. The results indicate that subjects with RIF-scores of 3 or 4 in the screening test have a higher risk of sustaining cold injuries (Fig. 3). If we assume that 1000 subjects were to be tested prior to deploying and sustained an expected 50 cases of cold injuries, 95 subjects would be expected to have RIF values of 3 4 and thus be stopped from going. In those, the screening would be expected to prevent 21 cases of cold injuries. However, 29 cold injury cases would still be expected in the remaining 905 subjects (false negatives). This means that 42% of the cold injuries would be prevented by the screening test. On the other hand, screening would have prevented 74 subjects from going who would not have sustained cold injuries (false positives). In conclusion, we demonstrated that the RIF, as calculated from the nger skin temperature, was higher in Caucasians than in the other ethnicities. Unexpectedly, smokers had a better RIF score than non-smokers. Subjects experiencing pain after 10 min of immersion had worse RIF scores and are thus thought to have higher risks for cold injuries. Of our subjects, 11 acquired cold injuries during an operation in the cold. These subjects had signicantly worse RIF scores than the other tested subjects. This study, therefore, shows that the RIF, determined in a simple lab test, may be related to the risk for cold injuries during operations in the eld.

REFERENCES 1. Borg G. Borgs perceived exertion and pain scales. Champaign, IL: Human Kinetics; 1988. 2. Campbell SC, Jewitt DE. Cold-induced immersion injury: a potential hazard of the cold pressor test. Am J Cardiol 1985; 55:585 6. 3. Daanen HAM. Central and peripheral control of nger blood ow in the cold [Thesis]. Amsterdam, The Netherlands: Vrije Universiteit; 1997. 4. Daanen HAM. Finger cold-induced vasolidation: a review. Eur J Appl Physiol 2003; 89:41126. 5. Daanen HAM, Ducharme MB. Finger cold-induced vasodilation during hypothermia, hyperthermia and at thermoneutrality. Aviat Space Environ Med 1999; 70:1206 10. 6. Daanen HAM, Ducharme MB. Axon reexes in cold exposed ngers. Eur J Appl Physiol 2000; 81:240 4. 7. Daanen HAM, Van de Linde FJG, Romet TT, Ducharme MB. The effect of body temperature on the hunting response of the middle nger. Eur J Appl Physiol 1997; 76:538 43. 8. Hirai K, Horvath SM, Weinstein V. Differences in the vascular hunting reaction between Caucasians and Japanese. Angiology 1970; 21:50210. 9. Iampietro PF, Goldman RF, Buskirk ER, Bars DE. Responses of Negro and white males to cold. J Appl Physiol 1959; 14:798 800. 10. Iida T. Studies of vascular reaction to cold (part 1). Physiological signicance of vascular reaction to cold (in Japanese). J Physiol Soc Jap 1949; 11:73. 11. Kreh A, Anton F, Gilly H, Handwerker HO. Vascular reactions correlated with pain due to cold. Exp Neurology 1984; 85:533 46. 12. Oakley EHN. Heat and cold. In: Hunter D, Adams PH, Aw T-C, et al, eds. Hunters diseases of occupations. London, UK: Hodder & Stoughton Educational; 2000. 13. OBrien C. Reproducibility of the cold-induced vasodilation response in the human nger. J Appl Physiol 2005; 98:1334 40. 14. Purkayastha SS, Ilavazhagan G, Ray US, Selvamurthy W. Responses of Arctic and tropical men to a standard cold test and peripheral vascular responses to local cold stress in the Arctic. Aviat Space Environ Med 1993; 64:11139. 15. STATISTICA for Windows [Computer program manual]. Tulsa, OK: StatSoft, Inc.; 2000. 16. Von Baeyer CL, Piira T, Chambers CT, et al. Guidelines for the cold pressor task as an experimental pain stimulus for use with children. J Pain 2005; 6:218 27. 17. Wilson O, Goldman RF. Role of air temperature and wind in the time necessary for a nger to freeze. J Appl Physiol 1970; 19:5937. 18. Yoshimura H, Iida T. Studies on the reactivity of skin vessels to extreme cold. Part I. A point test on the resistance against frostbite. Jap J Physiol 1950; 1:14759.

1122

Aviation, Space, and Environmental Medicine Vol. 76, No. 12 December 2005