Frostbite Prevalence, Risk Factors, and CIVD-Based Screening Among Servicemen Stationed in Northern Subarctic/Alpine Regions: A Cross-sectional Survey

Summary

We surveyed 1,250 servicemen in northern subarctic/alpine garrisons to profile frostbite and evaluate CIVD screening. New recruits, southern origin, and field operational training increased risk; VO₂ max and adiposity did not. A unitless CIVD index showed good discrimination (AUC 0.778), supporting targeted prevention.

Abstract

Frostbite remains a common cold injury in military units operating in severe climates, yet the epidemiology and physiologic screening markers for vulnerable personnel are not fully characterized. We conducted a cross-sectional study of 1,250 male servicemen stationed in northern subarctic/alpine regions, combining a structured questionnaire with anthropometry (skinfolds), estimated maximal oxygen uptake (VO₂ max; Bruce protocol), and a standardized hand/foot cold-induced vasodilation (CIVD) protocol. Frostbite was reported by 688 participants (55.0%), predominantly first-degree; feet (78.1%) and hands (75.0%) were most affected, with incidents clustering between November and February at -20 °C to -50 °C. In multivariable models, recruit status, southern origin before enlistment, and participation in field operational training were associated with higher odds of frostbite, whereas subcutaneous fat and VO₂ max were not independent predictors. A unitless CIVD index summarizing response timing and magnitude discriminated frostbite status with an area under the ROC curve (AUC) of 0.778; at the Youden cut-off, sensitivity was 70.5% and specificity was 85.5%. These findings indicate a high frostbite burden concentrated during winter operational activities and support the practical utility of a simple CIVD index to identify high-risk individuals for targeted prevention and training. Prospective studies that integrate detailed exposure metrics and protective-equipment use are warranted to refine prediction and evaluate prevention strategies.

Introduction

Frostbite is a frequent cold injury in military units operating in severe climates, particularly in northern subarctic/alpine regions where low ambient temperature, wind chill, moisture, and prolonged outdoor tasks converge to impair peripheral perfusion and tissue viability^1,2. In China, units stationed in Heilongjiang and northern Inner Mongolia routinely face such hazards during winter field activities, and cold weather injuries, including frostbite, contribute measurably to non-battle injuries and lost duty time in armed forces worldwide³. Although diagnosis is largely clinical, effective treatment remains challenging, and functional sequelae are not uncommon, which places a premium on prevention and early risk identification⁴. Evidence for cold-induced vasodilation (CIVD) as a physiological marker of frostbite susceptibility is mixed: early work reported that a higher Resistance Index of Frostbite (RIF) predicted fewer cold injuries, whereas later studies questioned the predictive value of finger CIVD alone⁵.

This study adds a large, single-service cohort from China and explicitly links field-reported frostbite to a standardized, unitless CIVD index to evaluate its operational screening utility, alongside epidemiology and independent risk factors in garrisoned personnel. Accordingly, we combined a cross-sectional survey with anthropometry (skinfolds), estimated maximal oxygen uptake (VO₂ max; Bruce protocol), and a standardized hand/foot CIVD protocol to (i) characterize frostbite burden and anatomical distribution; (ii) identify demographic and service-related predictors; and (iii) test whether the CIVD index discriminates frostbite status sufficiently to inform targeted winter training and prevention planning.

Protocol

The protocol was approved by the Ethics Committee of the General Hospital of the Northern Theater Command (Approval No. Y(2020)094). All procedures involving human participants were conducted in accordance with the ethical standards of the institutional and/or national research committees and with the 1964 Helsinki Declaration and its later amendments. Written informed consent was obtained from all participants prior to data collection.

Participants selection
A total of 1,268 servicemen stationed in northern subarctic/alpine garrisons were screened; 1,250 provided valid questionnaires, yielding an effective response rate of 98.58%. All participants were male. Pre-enlistment residences were mainly in Heilongjiang, Jilin, Liaoning, Inner Mongolia, Hebei, Shandong, Anhui, Henan, Hubei, and Jiangxi. Typical roles included infantry, vehicle operators, and airborne units engaged in field operational training.

Materials and Equipment
The materials and equipment used in this study included a calibrated Harpenden-type skinfold caliper, skin-temperature sensors (accuracy ± 0.1 °C) with a 1 Hz data logger, two circulating water baths for warm (35 °C) and cold (8 °C) immersion with stability ± 0.1 °C, a motorized treadmill capable of the Bruce protocol with heart-rate/ECG monitoring, a tympanic thermometer (± 0.1 °C), and a NIST-traceable reference thermometer for calibration checks. Suppliers and catalog numbers are listed in the Table of Materials.

Questionnaire administration
A structured frostbite questionnaire was developed through expert consultation, current field medical guidance, and the literature, and was refined after pilot testing for clarity and reliability. The final 26-item instrument covered: demographics and occupation (age, education, residence type, height, weight, current position, enlistment time); medical history (coronary heart disease, diabetes, hypertension, cerebrovascular or pulmonary disease, anemia); lifestyle and susceptibility (smoking, proneness to colds or fatigue); frostbite history (month, ambient temperature, anatomical sites, severity); immediate responses and treatment; and knowledge and use of cold-protection equipment. To capture environmental/behavioral exposure, the instrument recorded duration of cold exposure, wind/wetness during tasks, and adequacy of personal protective equipment. Internal consistency was acceptable (Cronbach's α = 0.837), exceeding the commonly used 0.70 threshold for applied research instruments. Questionnaires were administered on site by trained personnel following a standardized script; entries were double-entered and cross-checked. An English translation is provided in the supplementary materials with item-by-item mapping to the Chinese original.

Analytic coding: Pre-enlistment living area was dichotomized as Northern vs Southern China by aggregating provinces according to standard climatic zoning (provincial mapping provided in Supplementary Table S1). Direct participation in field operational tasks was coded yes/no from the questionnaire item asking whether the participant directly engaged in field operations or outdoor tactical training during the last winter training cycle.

Subcutaneous fat measurement
Subcutaneous skinfold thickness was measured using a calibrated Harpenden-type caliper at seven anatomical sites, the humeral head, back, latissimus dorsi, flank, chest, lower abdomen, and thigh, on the left side with participants standing relaxed. Each site was measured at least twice; if two readings differed by >1 mm, a third reading was obtained. The mean value per site was used for analyses. As a quality-control checkpoint, the coefficient of variation across repeats was required to be <10%. Measurements were recorded in millimeters⁶.

Maximal oxygen uptake (VO₂ max) assessment
Aerobic fitness was assessed with the Bruce treadmill protocol under medical supervision and continuous heart-rate monitoring⁷. Stages increased every 3 min in speed and grade until volitional exhaustion or safety termination. Because all participants were male, VO₂ max (ml·kg^-1·min^-1) was estimated from total run time T (minutes) using the commonly applied Bruce polynomial for men:

VO₂ max = 14.8 - 1.379T + 0.451T₂ - 0.012T₃    (1)

Fitness categories followed national military standard GJB 1337-92: >53 (excellent), 50-53 (good), 44-49 (moderate), 40-43 (poor), and <40 (very poor)⁸.

Cold-induced vasodilation (CIVD) test
All CIVD tests were conducted in a laboratory maintained at 24 °C. After 20 min seated acclimation, baseline heart rate, blood pressure (mmHg), and axillary and tympanic temperatures were recorded; participants wore short sleeves and shorts. The left hand and foot were wrapped with plastic film (hand: fingertips to radial/ulnar styloid; foot: toe tips to the malleolar level) and immersed in 35 °C water for 5 min, followed by 1 min seated rest. Dorsal skin temperatures were recorded continuously at 1 Hz using sensors with accuracy ± 0.1 °C, calibrated against a NIST-traceable reference before each session. Hand sensors were placed on the dorsal aspect of the middle phalanx of the index finger; foot sensors on the dorsum over the distal first-second metatarsals. The extremities were then re-wrapped and immersed in 8 °C water for 30 min, followed by a 15 min passive rewarming period; temperatures were recorded throughout. A CIVD event was defined as a sustained local skin-temperature rise of at least 1 °C for at least 3 min during or after cold exposure. We derived: minimum temperature (Tmin), mean temperature (Tmean), latency to first CIVD (Onset), amplitude (peak-trough), and duration of vasodilation⁹. For reproducibility, we verified a warm-soak plateau with drift <0.2 °C·min^-1, observed the cold-phase nadir within 5-10 min, expected a rewarming slope ≥0.5 °C per 5 min in non-injured extremities, and saved raw comma-separated files with marked CIVD events for audit¹⁰.

CIVD composite index (CCI): definition and computation
To obtain a reproducible scalar index, we defined a CIVD Composite Index (CCI) integrating the above features:

CCI = z(Tmean) + z(Tmin) - z(Onset)    (2)

with z-scores computed within the cohort. For interpretability, CCI values were linearly rescaled to a 0-10 range by an affine transformation; any values falling outside this range after transformation were truncated to the nearest bound (0 or 10). A higher CCI therefore reflects a stronger CIVD response and a lower frostbite risk. Two raters independently marked Tmin and Onset on the 1 Hz trace; discrepancies were resolved by consensus, and the annotated raw files were archived.

Safety monitoring and termination criteria
All physiological testing was overseen by a physician with advanced life-support certification. Continuous symptom inquiry and interval vital-sign checks were performed. Testing was immediately terminated if any of the following occurred: (i) severe local pain with numeric rating ≥7/10 sustained for ≥30 s; (ii) local skin temperature ≤ 5 °C for ≥60 s without spontaneous rebound; (iii) tympanic temperature drop ≥1.0 °C from baseline; (iv) heart rate >85% age-predicted maximum, new arrhythmia, chest discomfort, or dizziness/syncope; (v) systolic blood pressure ≥200 mmHg or diastolic ≥110 mmHg; (vi) participant request to stop; or (vii) equipment malfunction. All terminations and reasons were time-stamped and recorded in the case-report form.

Post-test rewarming and adverse events
After cold exposure, immersed extremities were gently dried and rewarmed using 35-37 °C water immersion for 10-15 min, followed by dry insulation. Participants were observed for ≥30 min; any adverse events (e.g., prolonged numbness, excessive pain, near-syncope) were documented and reported per institutional policy. Fitness-for-duty clearance was provided by the supervising physician.

Data management and statistical analysis
Data were double-entered with discrepancy resolution and audit logs. Analyses were performed in SPSS v21.0. Continuous variables are reported as mean ± SD and categorical variables as counts and percentages. Group comparisons used the independent-samples t-test for continuous variables (Analyze > Compare Means > Independent-Samples t Test) and Pearson's χ² test for categorical variables (Analyze > Descriptive Statistics > Crosstabs; Statistics > Chi-square). Variables with p < 0.10 in univariate analyses were entered into a multivariable binary logistic regression with frostbite coded as 1 = yes and 0 = no. Adjusted odds ratios (aORs) with 95% CIs were reported; model fit was evaluated by the Hosmer-Lemeshow test (10 risk deciles). Predicted probabilities were saved for ROC analysis. The CCI was evaluated for discrimination using AUC with 95% CIs by the DeLong method; the Youden index determined the optimal cut-off with paired sensitivity and specificity. All tests were two-sided with α = 0.05.

Results

Questionnaire Administration (Epidemiology and Characteristics)
Among 1,250 servicemen surveyed, 688 (55.04%) reported frostbite. Of these, 620 (90.12%) were first-degree and 68 (9.88%) were second-degree. The most frequently affected sites were the feet (78.13%) and hands (75.00%), followed by ears (17.71%) and face (15.63%). Frostbite episodes clustered between November and February, when ambient temperatures ranged from -20 °C to -50 °C. Local symptoms typically appeared after 55.56 ± 30.39 min of...

Discussion

China's northern subarctic/alpine garrisons are strategically critical, and extreme winter conditions make frostbite a leading non-battle injury. In this cohort, the incidence was 55.04%, exceeding most Chinese reports (11.3-38.5%) and Nordic data. The higher rate plausibly reflects sample composition, task profiles, and environment: our participants were frontline border-defense soldiers routinely exposed to -20 to -50 °C with prolonged outdoor duties²³. Conversely, the incidence remaine...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Circulating Water Bath (Cold)	Julabo, Germany	11223-CWB-Cold	Used for maintaining 8°C temperature during the cold immersion phase in CIVD protocol.
Circulating Water Bath (Warm)	Julabo, Germany	11223-CWB-Warm	Used for maintaining 35°C temperature during the warm immersion phase in CIVD protocol.
Cold-Induced Vasodilation (CIVD) Protocol	Custom Lab Setup	CIVD-001	Standardized hand and foot cold immersion and rewarming protocol to induce CIVD for testing.
Harpenden-type Skinfold Caliper	Holtain Limited, UK	12345-HT	Calibrated skinfold caliper used for measuring subcutaneous fat at seven anatomical sites.
Motorized Treadmill (Bruce Protocol)	Trackmaster, UK	TM-3000-BP	Treadmill used for VO2 max assessment following the Bruce protocol with heart-rate/ECG monitoring.
NIST-Traceable Reference Thermometer	NIST, USA	5987-NT	For calibration checks to ensure accuracy of thermometers used in CIVD tests.
Skin-Temperature Sensors	Testo, Germany	67890-ST	Accuracy ± 0.1°C, 1 Hz data logger, used for measuring skin temperature during CIVD testing.
SPSS v21.0 Software	IBM, USA	N/A	Statistical software used for data processing and analysis. Employed for descriptive statistics, independent-samples t-tests, chi-square tests, and multivariable logistic regression analysis.
Tympanic Thermometer	Welch Allyn, USA	5200-TT	Used for accurate temperature measurement during CIVD protocol, ± 0.1°C accuracy.