Foot immersion with and without neck cooling reduces self-reported environmental symptoms in older adults exposed to simulated indoor overheating

Abstract

While foot immersion and neck cooling have been recommended for protecting heat-vulnerable groups, recent evidence does not support their efficacy for mitigating increases in physiological heat strain in older adults. However, their influence on self-reported environmental symptoms and mood-state remains unclear. Seventeen older adults (nine females, median [interquartile range] age: 72 [69-74]) completed three randomized heat exposures (6-h; 38°C, 35% relative humidity) with no cooling (control), foot immersion to mid-calf in 20°C water for the final 40-min of each hour (foot immersion), or foot immersion with a wet towel (20°C) around the neck (foot immersion with neck cooling). Core temperature, skin temperature, and heart rate areas under the curve (AUC) were assessed as indicators of cumulative physiological strain. Environmental symptom scores (68-item environmental symptoms questionnaire) and mood disturbance (40-item profile of mood states questionnaire) were evaluated at end-heating (adjusted for pre-exposure). Core temperature AUC was not different between conditions (p = 0.418). However, the skin temperature and heart rate AUCs were 11.8°C · h [95% confidence interval: 8.1, 15.5] and 12.5 bpm · h [0.1, 24.8] lower for foot immersion and 16.6°C · h [12.9, 20.3] and 19.6 bpm · h [7.2, 32.0] lower for foot immersion with neck cooling compared to control (p ≤ 0.032). Environmental symptom scores were 0.8-fold [0.6, 1.0] lower for both foot immersion with and without neck cooling, compared to control (both p = 0.036). Mood disturbance was not different between conditions (both p ≥ 0.275). Foot immersion with and without neck cooling reduces self-reported environmental symptoms in older adults despite having little effect on physiological heat strain.

Keywords: Cooling interventions; heat stress; heat wave; psychometric stress; thermoregulation.

Figure 1.

Influence of cooling condition on the perceptual responses for older adults (69-74 y; n = 17) at the end of three randomized 6-h extreme heat exposures (38°C, 35% relative humidity). Conditions differed only in the applied cooling intervention: no cooling (control), submersion of the feet to mid-calf in 20°C water for the last 40 min of each hour (foot immersion), or foot immersion with a wet towel (20°C) draped around the neck (foot immersion with neck cooling). Total symptom score (top panel) was derived from the 68-item environmental symptoms questionnaire, version 4 (ESQ-IV). Total mood disturbance (middle panel) and Energy Index (lower panel) were both derived from the 40-item profile of mood states questionnaire (POMS-40). Data are presented as modeled means (large points) and 95% confidence intervals (error bars) with individual scores (small points), boxplots (median and interquartile range), and density plots (distribution) for each condition. End-heating symptom scores were evaluated using a negative binomial generalized linear mixed model with a log link function, except for energy index (POMS-40; difference score) which was analyzed with a traditional linear model (identity link). Cooling condition (categorical predictor), as well as pre-exposure score (Score_pre; continuous predictor) were included as fixed effects. To account for repeated measurements due to the crossover design, a random intercept was modeled for each participant. Between-condition contrasts are presented as fold-differences [95% confidence limits] for total symptom score and total mood disturbance after back transformation from the log scale and as mean differences [95% confidence limits] for Energy Index. Significance was set at p < 0.05 (two-sided). p values were adjusted for multiplicity using the Holm-Bonferroni procedure.

Figure 2.

Effect of cooling condition on total symptom scores after controlling for cumulative effects of mean skin temperature. Responses were evaluated in older adults (69-74 y; n = 17) at the end of three randomized 6-h extreme heat exposures (38°C, 35% relative humidity). Conditions differed only in the applied cooling intervention: no cooling (control, blue), submersion of the feet to mid-calf in 20°C water for the last 40 min of each hour (foot immersion, gray), or foot immersion with a wet towel (20°C) draped around the neck (foot immersion with neck cooling, yellow). The red points and error bars represent the mean (95% confidence intervals) responses for each condition. Total symptom scores were derived from the 68-item environmental symptoms questionnaire, version 4 (ESQ-IV). End-heating symptom scores were evaluated using a negative binomial generalized linear mixed model with a log link function. Cooling condition (categorical predictor), as well as pre-exposure score (Score_pre; continuous predictor) and skin temperature area under the curve (Tskin_auc, hours 0-6; continuous predictor) were included as fixed effects. To account for repeated measurements due to the crossover design, a random intercept was modeled for each participant. Significance was set at p < 0.05 (two-sided).

Figure 3.

Rank order of individual item responses at end-heating from the 68-item environmental symptoms questionnaire, version 4 (ESQ-IV). Grouped raw score responses are reported for older adults (69-74 y; n = 17) at the end of three randomized 6-h extreme heat exposures (38°C, 35% relative humidity). Conditions differed only in the applied cooling intervention: no cooling (control), submersion of the feet to mid-calf in 20°C water for the last 40 min of each hour (foot immersion), or foot immersion with a wet towel (20°C) draped around the neck (foot immersion with neck cooling). For each item, participants were asked to rank their response based on a 6 level likert scale (0 = not at all, 1 = slight, 2 = somewhat, 3 = moderate, 4 = quite a bit, 5 = extreme). Only items with group mean ratings ≥ 1 (“slight”) are presented here, representing a minimum stimulus response. Individual items for all conditions were rank ordered by group mean, then by prevalence of responses for the no-cooling control condition.

Figure 4.

Rank order of individual item responses at end-heating from the 40-item profile of mood states questionnaire (POMS-40). Grouped raw score responses are reported for older adults (69-74 y; n = 17) at the end of three randomized 6-h extreme heat exposures (38°C, 35% relative humidity). Conditions differed only in the applied cooling intervention: no cooling (control), submersion of the feet to mid-calf in 20°C water for the last 40 min of each hour (foot immersion), or foot immersion with a wet towel (20°C) draped around the neck (foot immersion with neck cooling). For each item, participants were asked to rank their response based on a 5 level likert scale (0 = not at all, 1 = a little, 2 = moderately, 3 = quite a lot, 4 = extremely). Only items with group mean ratings ≥ 1 (“A little”) are presented here, representing a minimum stimulus response. Individual items for all conditions were rank ordered by group mean, then by prevalence of responses for the no-cooling control condition.

Figure 5.

Influence of cooling condition on the physiological responses of older adults (69-74 y; n = 17) at the end of three randomized 6-h extreme heat exposures (38°C, 35% relative humidity). Conditions differed only in the applied cooling intervention: no cooling (control), submersion of the feet to mid-calf in 20°C water for the last 40 min of each hour (foot immersion), or foot immersion with a wet towel (20°C) draped around the neck (foot immersion with neck cooling). The cumulative physiological effects of each condition were evaluated by assessing the areas under the curve (AUC) for body core temperature (Tre, top panel), mean skin temperature (Tskin, middle panel), and heart rate (HR, lower panel). Data are presented as modeled means (large points) and 95% confidence intervals (error bars) with individual scores (small points), boxplots (median and interquartile range), and density plots (distribution) for each condition. Between-condition contrasts are presented as mean differences [95% confidence limits]. Significance was set at p < 0.05 (two-sided). p values were adjusted for multiplicity using the Holm-Bonferroni procedure.