Improving Glycemic Control via Heat Therapy in older adults at risk for Alzheimer’s Disease (FIGHT-AD): a pilot study

Abstract

Impaired glycemic control increases the risk for type 2 diabetes (T2D) and Alzheimer’s Disease (AD). Heat therapy (HT), via hot water immersion (HWI), has shown promise in improving shared mechanisms implicated in both T2D and AD, like blood glucose regulation, insulin sensitivity, and inflammation. The potential for HT to improve brain health in individuals at risk for AD has not been examined. This pilot study aimed to assess the feasibility and adherence of utilizing HT in cognitively healthy older individuals at risk for AD due to existing metabolic risk factors. Participants underwent four weeks of HT (three sessions/week) via HWI, alongside cognitive screening, self-reported sleep characterization, glucose tolerance tests, and MRI scans pre- and post-intervention. A total of 18 participants (9 male, 9 female; mean age: 71.1 ± 3.9 years), demonstrating metabolic risk, completed the intervention. Participant adherence for the study was 96% (8 missed sessions out of 216 total sessions), with one study-related mild adverse event (mild dizziness/nausea). Overall, the research participants responded to a post-intervention survey saying they enjoyed participating in the study and it was not a burden on their schedules. Secondary outcomes of the HT intervention demonstrated significant changes in mean arterial pressure, diastolic blood pressure, and cerebral blood flow (p<0.05), with a trend toward improved body mass index (p=0.06). Future studies, including longer durations and a thermoneutral control group, are needed to fully understand heat therapy’s impact on glucose homeostasis and potential to improve brain health.

New & Noteworthy

Our pilot study demonstrated promising results for heat therapy (HT) via hot water immersion in older adults at risk for Alzheimer’s Disease due to metabolic factors. Despite a relatively short intervention, significant improvements in mean arterial pressure, diastolic blood pressure, and cerebral blood flow post-intervention were observed. High participant adherence, overall satisfaction, and minimal adverse events, suggest HT’s feasibility. These findings highlight HT’s potential as an effective alternative intervention for cardiometabolic dysfunction in at-risk populations.

Graphical Abstract

1.0. Introduction

Alzheimer’s Disease (AD) is the most prevalent neurodegenerative disorder and by 2060, the number of affected individuals is predicted to reach 13.8 million worldwide (1). While the etiology of AD is not fully understood, mechanisms such as impaired energy metabolism, cellular bioenergetic dysfunction, reduced intracellular protein homoeostasis, and inflammation are all potential contributors (2–6). The link between metabolic dysfunctions and AD (7), particularly impaired glucose metabolism (8), suggests a critical intersection between metabolic health and neurodegenerative disease progression. Elevated glucose levels, for example, correlate with increased cerebral amyloid deposition – a hallmark of AD – and reduced amyloid beta (Aβ) catabolism, indicating that interventions targeting metabolic health could play a crucial role in managing AD risk (9–13). A recently completed exercise prevention trial in cognitively healthy older adults at risk for AD showed that longitudinal increases in fasting glucose over one year were associated with regional increases in brain amyloid (10). This research suggests that glucose regulation may be an important therapeutic target for AD.

Traditional lifestyle interventions for AD focus on symptom management and slowing disease progression. Aerobic exercise has shown efficacy in improving functional abilities and potentially enhancing cognitive outcomes in individuals with AD by mitigating associated risk factors such as cardiovascular disease, obesity, and Type 2 diabetes (T2D) (14–18). The benefits of aerobic exercise training on brain health likely occur, at least in part, through increased cerebral blood flow (19), a critical vascular measure which declines with age (20, 21). However, the low adherence rates to exercise, particularly among older adults – with only 14% engaging sufficiently – limit its widespread applicability (22).

Heat therapy (HT) via hot water immersion (HWI) has emerged as a promising alternative therapeutic strategy. From the initial work by Dr. Hooper in 1999 demonstrating that three weeks of HWI significantly lowered blood glucose and glycated hemoglobin (HbA1c) in a small group of individuals (five men and three women, age range 43 to 68 years) (23), there are now a number of studies demonstrating benefits of HT on glucose regulation. Interventions ranging from 2–10 weeks of HT via HWI resulted in reductions in fasting glucose and/or glucose response to an oral glucose tolerance test in individuals with obesity and/or type 2 diabetes (23–25). Obese women with polycystic ovarian syndrome demonstrated decreased fasting glucose and improved glucose tolerance, as well as reduced sympathetic activity and improved cardiovascular risk profiles following repeated sessions of HWI (24, 26). HWI has demonstrated significant metabolic and cardiovascular benefits across various populations (24, 26–32) and has been shown to have anti-inflammatory effects, including acute increases in beneficial cytokines (33). HWI improves fasting glucose and insulin levels in ways similar to the effects of exercise (34), making it particularly attractive for individuals unable to participate in traditional physical activities (23, 33). Among the elderly population, a demographic at heightened risk for AD, HWI has been shown to be safe and well tolerated. In a 12-week study, elderly individuals with peripheral arterial disease who underwent HWI showed improvements in walking distance and resting blood pressure (28).

Considering the impact of systemic metabolic dysfunctions on AD pathogenesis, and the demonstrated benefits of HT on glucose metabolism, this pilot study aimed to evaluate the feasibility and adherence for a regimen of HWI heat therapy among cognitively healthy older adults at metabolic risk for AD. We hypothesized that HT via HWI would be a feasible intervention for this population with heat resulting in improvement to peripheral glucose metabolism, blood pressure, and anthropometric outcomes such as body weight and body mass index. Our findings suggest heat therapy may be a viable alternative to traditional exercise modalities for individuals at risk for AD.

2.0. Methods

2.1. Trial Design

This interventional pilot study characterized the feasibility of and adherence to HT in cognitively healthy older adults (65+) who are at risk for AD due to metabolic factors. Figure 1 presents the CONSORT diagram, which outlines the flow of participants through each stage of the study. All participants were required to have a study partner (someone who routinely interacts with the participant > 5 times a week) to be available to speak with the study team via telephone about the prospective participant’s functional performance during the Quick Dementia Rating System (QDRS) questionnaire (35). The study partner was consented over the phone. Following pre-screening, informed consent, and enrollment into the study, participants (n = 18) underwent a fasting blood draw, oral glucose tolerance test (OGTT), and a magnetic resonance imaging (MRI) scan. Following this pre-intervention evaluation, participants received HT three times a week for four weeks. After completion of heat treatments, subjects repeated the same assessments. Activities were condensed into the fewest possible visits to reduce participant burden and to allow the capture of key markers in their temporal relationship following a rigorous, time-sensitive protocol. The flow of the study is illustrated in Figure 2.

Figure 1. CONSORT diagram showing the flow of participants through the study, including recruitment, screening, enrollment, intervention, and follow-up.

Figure 2. Study flow diagram for the FIGHT-AD Pilot study.

Following consent, participants underwent pre-intervention evaluations, including a fasting blood draw, an oral glucose tolerance test (OGTT), and a magnetic resonance imaging (MRI) scan. Participants then received heat therapy three times a week for four weeks. Each heat therapy session began with the participant fully immersed in the hot water up to shoulder level until their core body temperature increased by 1°C or 30 minutes had elapsed, at which point immersion was reduced to waist level. Heart rate (HR), blood pressure (BP), and core temperature were measured before entering the hot tub, every 10 minutes during immersion, and 10 minutes after exiting. Upon completing the heat therapy regimen, participants underwent the same set of evaluations to assess post-intervention changes.

2.1.1. Outcome Measures

The co-primary outcomes of this pilot trial were feasibility of the intervention (number of HT sessions attended) and adherence to the intervention (number of HT sessions completed once started). Secondary outcomes included change in peripheral glucose metabolism, blood pressure, and anthropometric outcomes such as body weight and body mass index following the intervention. Exploratory outcomes included neuroimaging measures of brain blood flow and self-reported sleep characterization for preliminary characterization of intervention-related changes.

2.2. Recruitment and Participants

Participants were recruited through a registry of volunteers at the University of Kansas Alzheimer’s Disease Research Center (KU ADRC). Written informed consent was obtained from all participants after they received both a verbal and written briefing of all experimental procedures on the day of their first visit. Eligible participants were aged 65 years and older, had no reported history of cognitive impairment, and had history of or current metabolic impairment (i.e., a diagnosis of prediabetes or T2D, or at least two of the following criteria: hypertension, dyslipidemia, BMI ≥30). Additional inclusion and exclusion criteria are detailed in Table 1. This study was approved by the Institutional Review Board at the University of Kansas Medical Center (IRB STUDY00147446).

Table 1.

Study Inclusion/Exclusion Criteria

Inclusion Criteria	Exclusion Criteria
Informed Consent	Clinically significant chronic disease2
Men and women 65 years and older	ACSM risk score stratification of “High” unless cleared by their physician prior to participation
Stable medication dose (>1 month)	Diagnosis of Type 1 Diabetes Mellitus
Post-menopause (females)	Inability to undergo MRI scan
Diagnosis of metabolic impairment1	Diagnosis of recent myocardial infarction or symptoms of coronary artery disease (<2 years)
	Neurological disease impairing cognition or brain metabolism3
	Clinically significant depressive symptoms that may impair cognition
	Use of psychoactive and investigational medications
	Significant visual or auditory impairment
	Orthopedic complications that would preclude individuals from safely entering a hot tub
	Untreated hypothyroidism or diseases associated with heat intolerance4
	Prior diagnosis of cognitive impairment
	Contraindication for temperature pill ingestion5

¹This includes such diseases as metabolic syndrome (at least two of the following criteria: hypertension, dyslipidemia, BMI ≥30), prediabetes or Type 2 Diabetes.

²This includes diseases such as cancer, HIV, or acquired immunodeficiency syndrome.

³This includes diseases such as Alzheimer’s Disease, Parkinson’s Disease, stroke defined as a clinical episode with neuroimaging evidence in an appropriate area to explain the symptoms.

⁴This includes diseases such as Graves’ disease.

⁵This includes diseases such as inflammatory bowel disease or related.

2.3. Procedures

The list of procedures at each study visit is given in Table 2. Participants underwent a phone screening that included the QDRS questionnaire and the collection of basic demographic information. Once deemed eligible, participants completed all subsequent screening procedures and two in-person pre-intervention study visits (OGTT and MRI scan). After completing the pre-intervention visits, individuals attended approximately three HT sessions per week for four weeks, totaling 12 sessions, before completing two post-intervention study visits (OGTT and MRI scan). All visits were completed within approximately five weeks. Each study visit is detailed below.

Table 2.

Study Procedures

Assessments	Pre-Screening	Visit 1	Visit 2	Intervention (4 weeks)	Visit 3	Visit 4
Demographics	X
Telephone QDRS	X
Medications	X	X			X
Medical History	X	X			X
Oral Glucose Tolerance Test		X			X
Blood Draws		X			X
Vitals		X		X	X
MRI scan			X			X
Heat Therapy Sessions				X

Vitals include heart rate and blood pressure. Abbreviations: QDRS (Quick Dementia Rating System); MRI (Magnetic Resonance Imaging) scan

2.3.1. Pre-Intervention Visit 1

The OGTT visit was scheduled for the morning after participants had fasted overnight. Upon their arrival to the KU Clinical and Translational Science Unit (CTSU), fully informed, written consent was obtained for the in-person visits. Their medical history was then collected, current medications, and anthropometric measures, including height and weight, before starting the OGTT. Participants were asked to rest quietly and seated for five minutes, during which heart rate and blood pressure were measured. Subsequently, an IV catheter was inserted for blood collection throughout the study visit. To maintain consistency across all trials and accurately record any deviations from the planned schedule all visit activities were timed.

After placing the IV catheter, a baseline blood sample was taken. Participants were then given a commercially available oral glucose beverage, containing 75g of glucose, which they consumed within five minutes. Timing began immediately after they finished the beverage. Blood samples were collected at intervals of 30-, 60-, and 120-minutes post-consumption. The OGTT concluded with a 10-minute observation period following the final blood draw. The Sleep/Circadian (Pittsburgh Sleep Quality Index - PSQI) and Geriatric Depression Scale (GDS) surveys were administered to participants during both OGTT visits, before and after the HT intervention.

2.3.2. Pre-Intervention Visit 2

MRI scans were performed pre- and post-intervention (Skyra, Siemens, Erlangen, Germany). A T₁-weighted, 3D magnetization prepared rapid gradient echo (MPRAGE) structural scan was acquired for detailed anatomical assessment (TR/TE = 2300/2.95 ms, inversion time (TI) = 900 ms, flip angle = 9 deg, FOV = 253 × 270 mm, matrix = 240 × 256 voxels, voxel in-plane resolution = 1 mm², slice thickness = 1.0 mm, 176 sagittal slices, in-plane acceleration factor = 2, acquisition time = 5:12) A pseudo-continuous, background suppressed, 3D gradient and spin echo arterial spin labeling (pCASL) sequence protocol was acquired to measure cerebral blood flow (TE/TR = 22.4/4300 ms, FOV = 300 × 300 × 120 mm3, matrix = 96 × 66 × 48, Post-labeling delay = 2 s, 4-segmented acquisition without partial Fourier transform reconstruction, readout duration = 23.1 ms, total scan time 5:48, 2 M0 images). The pre-intervention MRI scan was performed prior to the use of ingestible heat sensors and the follow-up scan was performed at least seven days following the final HT session to allow for the final core temperature sensor to pass through the gastrointestinal tract (GI) tract (see Heat Therapy Intervention Visits below). The participant was checked using the respective receiving device for any ingested telemetric pill.

MRI analysis protocols have been described previously (36). Briefly, we created individualized gray matter regions of interest (whole brain, hippocampus, and cerebellum as a reference region) for each participant using the Statistical Parametric Mapping CAT12 (neuro.uni-jena.de/cat, r1059 2016–10-28). We motion corrected labeled and control pCASL images separately, realigning each image to the first peer image following M0 image acquisition. CBF was calculated with surround subtraction of each label/control pair without biopolar gradients which were then averaged. Subtraction images were then co-registered to the anatomical CSF segmented image and smoothed using a 6mm full width, half maximum Gaussian window.

2.3.3. Heat Therapy Intervention Visits (4 weeks)

Participants attended HT sessions three times weekly for four weeks. At each visit to the CTSU, core temperature was monitored using an ingestible temperature sensor (telemetric pill, HQ Inc, Palmetto Fl). Once ingested, the pill traveled through the GI tract and transmitted temperature to a receiving device outside of the participant’s body. The pill typically passed through the participant within 2–3 bowel movements. At each subsequent visit, the core temperature sensor’s accuracy was checked against the participant’s previous session data. If the sensor’s signal was intact, no new sensor was provided. When no signal was present, a new sensor was ingested, followed by a 30-minute rest period to ensure signal stability before proceeding. During the session, participants were immersed up to their shoulders in a 40°C water therapy bath until their core body temperature (Tc) increased by 1°C, or 30 minutes had elapsed. Although there is some variability in the increase in Tc between participants, the 30-minute maximum time for the initial heating maintained a maximal heat dose throughout the study. Participants then remained in the hot tub submerged to waist level to maintain their Tc for an additional 15 minutes, totaling 45 minutes. Following HT, participants exited the hot tub and were monitored for 10 minutes or until their Tc returned to baseline.

Signs and symptoms of heat-related illness were continuously monitored by study staff. Heart rate and blood pressure measurements were taken every 10 minutes during the intervention session to ensure adequate hydration and circulation. If a participant’s heart rate increased by more than 60 beats per minute above resting or increased by more than 20 beats per minute within a 5-minute period, the participant was moved to a seated position if they were previously fully submerged, or removed from the hot tub if they were already sitting up. If body temperature reached 39.5°C, or the participant experienced symptoms of heat-related illness, a ‘Rapid Cooling Protocol’ was implemented. Participants were allowed to drink water ad libitum during the session. If the participant’s Tc was affected by drinking water, repeat measurements were taken until the temperature normalized to ensure accurate Tc readings. Body weight was measured before and immediately after HT in all sessions to monitor hydration levels. Participants who did not drink enough water to compensate for sweat loss (body weight loss > 1%) were required to drink additional fluids to make up this difference before leaving the visit.

2.3.4. Post-Intervention Visits

Upon completion of the HT intervention, a post-intervention OGTT was performed as previously described. The post-intervention OGTT was performed within approximately three days of completing the final HT intervention visit. At this visit, participants provided feedback through a survey regarding their perception of the HT intervention (see Appendix Table 1 for survey questions and results). The MRI scan was performed on average nine days following the final HT intervention visit.

2.4. Blood Processing

At the baseline blood draw during visit 1 pre-intervention, whole blood was collected in an ACD yellow top vacutainer tube (8.5mL) specifically for apolipoprotein E4 (ApoE4) genotyping. ApoE genotyping was performed as described previously (36). Subsequently, at both baseline and three additional time points during each OGTT visit, blood was drawn into EDTA lavender top vacutainer tubes (10mL) used for plasma collection. Our previously reported optimized sample collection and processing procedures to ensure accurate and consistent collection of both platelet rich plasma (PRP) and platelet poor plasma (PPP) was used (37). PRP was generated by centrifuging whole blood at 1500 x g for 10 minutes at 4°C, followed by careful plasma removal. PPP was obtained from an aliquot of PRP through an additional centrifugation step at 1700 x g for 15 minutes at 4°C, performed during the baseline blood draw at each OGTT visit. Glucose levels were immediately analyzed in PRP using a YSI analyzer, while HbA1c levels were analyzed directly in whole blood from the ACD tube using the Siemens DCA Vantage Analyzer. After processing, aliquots of whole blood, PRP, and PPP were immediately frozen and stored at −80°C for future analyses.

2.5. Biomarker Analysis

The Simoa HD-X (Quanterix) was used to quantify markers of AD neuropathology, including plasma amyloid beta 42/40 ratio (Aβ42/40), phosphorylated Tau 181 (pTau181), neurofilament light (NfL), and glial fibrillary acidic protein (GFAP). Plasma Aβ42/40 is an established proxy for brain amyloid (38, 39), and plasma pTau181 is known to correlate with AD severity and brain imaging markers of tau pathology (40–42). NfL is a plasma marker of neurodegeneration (43) that tracks positively with age. Early astrocytosis secondary to amyloid-β pathology is detected by plasma GFAP (44). For this study we used the pTau181 (v2) and neuro 4 plex E (N4PE) kits for fluid biomarker analyses, with N4PE used specifically to analyze Aβ42/40, NfL, and GFAP (Quanterix). The goal of these analyses were to provide a baseline biomarker characterization of the cohort. All analyses followed the manufacturer’s instructions, including the use of appropriate standards and quality control samples (37). All samples were processed in duplicate, and the mean concentration of blood biomarkers was recorded for each sample. ELISA methodology is used to analyze metabolic markers (insulin).

2.6. Statistical Analyses

2.6.1. Primary Outcomes

Data was collected and analyzed for 19 subjects. One subject withdrew from the study for reasons unrelated to the study’s requirements, and the data collected on that subject prior to dropping out was not included in subsequent analyses.

The probability of a subject attending a single HT session was estimated and a Clopper-Pearson exact 95% confidence interval was constructed for the estimate. The probability of a subject attending all 12 HT sessions was estimated similarly. Wald 95% confidence intervals were also constructed for the estimates of these two outcomes.

2.6.2. Secondary and Exploratory Outcomes

One-sample paired t-tests were conducted for each secondary outcome to assess the differences between post-HT and pre-HT values. Change variables were derived by calculating the difference between post-HT and pre-HT values for each secondary outcome. Ordinary least squares regression was employed to model the change in each secondary outcome as a function of age and sex. Each change score was also modeled to obtain an estimated age- and sex-adjusted intercept for the model, using the average age and the proportion of males to females specific to that secondary outcome. All analyses were conducted using SAS Version 9.4. The full data set will be made available on Harvard Dataverse following publication.

3.0. Results

3.1. Participant Demographics

One of the initial 19 recruited participants, one withdrew from the study due to unrelated events after four visits. It is noteworthy that the withdrawn participant attended and successfully completed 100% of their visits. Subsequent analyses were conducted with this participant excluded from the dataset. Among the remaining 18 participants, the mean age was 71 years, and 50% were female. The cohort exhibited considerable diversity, with 28% identifying with an minoritized racial or ethnic community. All participants were classified as being at high metabolic risk, evident from average HbA1c values of 48.5 mmol/ml, average BMI of 29 kg/m², and 15 out of 18 participants using antihypertensive medications. In this cohort, ATN markers were measured to confirm baseline cognitive function, yielding average values of 0.07 pg/mL for Aβ42/40, 2.49 pg/mL for pTau181, and 18.30 pg/mL for NfL. All participants were initially identified as cognitively unimpaired during recruitment; however, after the administration of a Geriatric Depression Scale (GDS) questionnaire provided at the OGTT visit, two participants scored high on the scale. Subsequent exclusion analyses demonstrated that these outliers did not significantly impact the overall results. Detailed demographic information for the study participants is presented in Table 3.

Table 3.

Demographic and clinical characteristics of study participants (n = 18).

Patient Characteristics	Value
Age (years), mean ± SD	71.06 (3.87)
Male sex, n (%)	9 (50.0)
Race, n (%)   Asian   Black or African American   Another Racial Identity   White	2 (11.1) 2 (11.1) 1 (5.6) 13 (72.2)
Hispanic Ethnicity n (%)	2 (11.1)
Education (years), mean ± SD	16.28 (1.99)
ApoE4 carrier, n (%)	8 (44.4)
Hypertension Medication, n (%)	15 (83.3)
Diabetes medication (#,%)	11 (61%)
HbA1c 6.5% or greater	9 (50%)
Total QDRS Score, mean ± SD	0.72 (0.84)
NfL (pg/mL), mean ± SD	18.30 (7.84)
pTau181 (pg/mL), mean ± SD	2.49 (2.02)
Aβ42/40 (pg/mL), mean ± SD	0.07 (0.01)

This table presents comprehensive demographic and health-related characteristics of the study participants. Mean age, education level, Total QDRS (Quick Dementia Rating System) score, NfL (neurofilament light), pTau181 (phosphorylated Tau 181), Aβ42/40 (amyloid β 42/40) are reported along with corresponding standard deviations (SD) or percents. Additionally, the table includes information on the distribution of participants based on sex, race, ethnicity, ApoE4 carrier status, and the use of hypertension medication.

3.2. Heat Therapy Adherence and Feasibility

Out of the 12 scheduled HT sessions, the estimated probability of participants attending any given HT session was 96.3%, with an exact 95% confidence interval of 92.8% to 98.4%. Out of a total of 216 HT sessions, only 8 sessions were missed. Only six participants missed one or more sessions due to extenuating circumstances: one due to a snowstorm, another because of a minor car accident the previous evening, and a third missed two sessions during a COVID-19 isolation period. All other participants attended all 12 of their scheduled HT sessions. The likelihood of attending all 12 sessions was estimated at 66.7%, with a 95% confidence interval ranging from 41.0% to 86.7%. The Wald confidence intervals for these probabilities were similar, at 93.8% to 98.8% for attendance at any session and 44.9% to 88.4% for full attendance. Once a session was initiated, it was completed 100% of the time.

Throughout the study, four adverse events were reported, of which three were unrelated to the study protocol. The one related adverse event was mild, characterized by light-headedness and nausea during one of the HT sessions. Overall, feedback from a post-intervention survey was positive (Appendix Table 1), with participants expressing enjoyment and satisfaction with their involvement in the study. Most reported that participating was not burdensome on their schedules, highlighting the study’s minimal impact on their daily lives.

3.3. Secondary Outcomes

The secondary outcomes from the pilot study, summarized in Table 4, showed favorable changes in vascular measures, considering age and sex differences among participants. Specifically, mean arterial pressure (MAP) significantly decreased by an average of 6.6mmHg (Standard Deviation (SD) = 9.8, p=0.009) from pre-HT to post-HT, measured at rest before the participants entered the hot tub at both the first and last HT sessions, as shown in Figure 3A. Acute decreases in MAP were also observed during the HT session while inside the hot tub at both the initial and final visits, as depicted in Figure 3B. Similarly, diastolic blood pressure significantly decreased from pre-HT to post-HT by an average of 6.8mmHg (SD = 6.9, p<0.001).

Table 4.

Summary of Pre- and Post-Heat Therapy Intervention Measurements with Changes (Δ) and Corresponding P-Values

	CONDITION	OVERALL	P-VALUE
MAP (mmhg)	pre	94 (5.13)	0.009
	post	87 (8.59)
	∆	−6.62 (9.83)
GluAUC (mg/dL·h)	pre	26200 (6070)	0.359
	post	27100 (7320)
	∆	927 (3780)
BMI (kg/m²)	pre	29.1 (5.44)	0.064
	post	28.9 (5.49)
	∆	−0.270 (0.556)
BW (kg)	pre	82.7 (19.9)	0.089
	post	82.0 (19.9)
	∆	−0.672 (1.47)
HbA1c (mmol/mol)	pre	48.5 (7.2)	0.229
	post	46.1 (9.5)
	∆	−1.8 (5.7)
Systolic BP (mmhg)	pre	131 (8.34)	0.125
	post	125 (15.4)
	∆	−6.28 (17.3)
Diastolic BP (mmhg)	pre	76 (5.96)	0.0008
	post	69 (6.48)
	∆	−6.78 (6.91)
PSQI	pre	7.39 (3.96)	0.106
	post	6.39 (4.13)
	∆	−1.00 (2.63)
INSULIN AUC	pre	2057.5 (4854)	0.335
	post	2536.6 (4840)
	∆	210.4 (1998)
HOMA-IR	pre	1.30 (9.7)	0.307
	post	1.16 (12.2)
	∆	−0.24 (1.9)

This table presents the pre- and post-intervention measurements along with the changes (∆) observed for various parameters. The values represent mean (standard deviation) for each condition. The p-values indicate the statistical significance of the changes observed between pre- and post-intervention measurements adjusted for age and sex. Negative values for mean change indicate a decrease post-intervention, while positive values indicate an increase. MAP (mean arterial pressure), systolic BP (systolic blood pressure), diastolic BP (diastolic blood pressure) were measured at first heat treatment session and last heat treatment session at rest before entering the hot tub. BMI (body mass index), BW (body weight), GluAUC (glucose area under the curve), HbA1c (glycated hemoglobin), systolic BP, and PSQI (Pittsburg sleep quality index).

Figure 3. Effects of Heat Therapy on Mean Arterial Pressure.

A) Individual participants at the resting (outside of the hot tub, seated rest prior to entering the hot tub) mean arterial pressure (MAP) values at the initial and final heat therapy session. Resting MAP significantly reduced (Final-Initial) following the heat therapy intervention. B) Acute changes of MAP while inside the hot tub (showing the effect of heat therapy on blood pressure acutely during the session). There is a significance difference of MAP with time spent in the hot tub.

Post-intervention, after adjusting for age and sex, additional trends were noted. BMI presented a borderline significant association with an average reduction of 0.27kg/m² (p = 0.06). Body weight (BW) also trended towards an association, with a decrease of 0.67kg, though this trend approached but did not achieve statistical significance (p = 0.09), again adjusted for age and sex.

3.4. Exploratory Outcomes

In terms of exploratory outcomes, whole gray matter cerebral blood flow significantly increased from pre-HT to post-HT, averaging 3.7 mL * 100g tissue⁻¹ * min⁻¹ increase (SD = 7.0, p=0.032), as detailed in Figure 4.

Figure 4. Effects of Heat Therapy on Cerebral Blood Flow.

Pre to post change in whole gray matter cerebral blood flow measured using arterial spin labeling. Bar represents mean +/− SEM with individual data points plotted. n = 16 individuals.

4.0. Discussion

Understanding the metabolic implications of HT is critical given the rising prevalence of AD and the limited success of conventional treatments. This pilot study explored the feasibility and adherence of HT in older adults at metabolic risk, addressing the gap in alternative interventions for this vulnerable cohort. The current cohort, equally split between male and female participants, racially and ethnically diverse, and 44% being ApoE4 carriers, enhances the generalizability of our findings across a diverse population at risk for AD. The HT intervention – three weekly HWI sessions over four weeks – mirrors potential real-world application, addressing both metabolic health and potential barriers like exercise adherence. Key findings demonstrated a high probability of participants attending HT sessions (96.3%), with only 8 out of 216 sessions missed. Notably, participants facing extenuating circumstances, such as a snowstorm or COVID-19 isolation, missed only a few sessions. This is consistent with prior studies showing excellent adherence to HT (45, 46), emphasizing its practicality and acceptance. The broad and inclusive nature of this cohort, coupled with overwhelmingly positive feedback from participants, further supports that HT is not only feasible but also well-tolerated and enjoyable, crucial factors for long-term adherence.

In the broader context of HT literature, the current study demonstrates the feasibility, adherence, and potential cardiometabolic benefits – particularly in terms of blood pressure – within a high-risk population. Despite the brief four-week intervention, significant changes in MAP and diastolic blood pressure, as well as slight changes in body composition, were observed. These findings reinforce the value of HT as an intervention and suggest that it can impart meaningful benefits even within a condensed timeframe.

The reductions observed in MAP and diastolic blood pressure following four weeks of HT are consistent with improvements in endothelial function reported in previous studies (47–49), suggesting that HT may contribute to enhanced vascular relaxation, reduced arterial stiffness, and protection against endothelial dysfunction. Individuals with high metabolic risk often exhibit impaired endothelial function, commonly due to hypertension or inflammation (50). Endothelial cells, which line blood vessels, play a crucial role in regulating vascular tone and prolonged inflammation can impair endothelial function (51). Prior studies indicate enhanced endothelial function following HT (52–55), which could be impactful for older adults with compromised metabolic function (24, 56–59). Our research group has made a significant contribution to the literature demonstrating the benefits of HT in preclinical models (60–65), and potential mechanisms such as decreased inflammation, improved insulin signaling, and enhanced mitochondrial function as a result of Heat Shock Protein (HSP) activation need to be examined in humans (66).

An additional benefit of HT assessed in the current study as an exploratory outcome was improved cerebral blood flow. Prior work from our research group demonstrated that in a population of both APOE4 carriers and non-carriers, an acute bout of aerobic exercise increased cerebral blood flow (36). In addition, in older adults with poor cerebrovascular health, hippocampal blood flow rapidly increased following a bout of moderate-intensity exercise (67). The brain benefits of chronic aerobic exercise likely occur, at least in part, through increases in cerebral blood flow (19). The observed increase in cerebral blood flow in the current study suggests a potential for enhanced cerebral oxygenation and metabolic support following HT, which could be particularly beneficial for older adults who may have compromised cerebrovascular health. This finding is consistent with previous research demonstrating modification of cerebrovascular dynamics in response to thermal exposure with and without exercise (68–70). However, comparisons are difficult as prior studies examined the cerebral blood flow response to acute bouts of HWI and employed measures of internal and external carotid artery conductance and/or shear stress using ultrasonography. Understanding how HT influences cerebrovascular function, and its potential long-term benefits could provide valuable insights into therapeutic strategies for enhancing brain health in aging populations. A larger clinical trial in this same population is ongoing in hopes of addressing these important outcomes (Clinicaltrials.gov NCT06023407).

While baseline measurements of AD-related blood biomarkers (Aβ42/40, pTau181, NfL, and GFAP) were reported to assess participants’ cognitive function, their role in understanding the overall impact of HT on metabolic health remains an important area for future research. More detailed analyses of these biomarkers before and after HT are essential to elucidate potential benefits on cognitive health. Future studies should investigate the impact of heat therapy on ATN biomarkers to understand underlying neurobiological changes and potential risks of neurodegenerative diseases.

Limitations of the present study include a small sample size, the absence of a control group, and the short intervention of four weeks. These factors may affect the generalizability of the results and the observation of long-term effects, however, this study cohort exhibited considerable diversity. Due to a relatively small sample size and the pilot nature of the trial, the ability to explore additional factors such as sex was limited. Future more well-powered studies should investigate the impact of sex on metabolic outcomes. The potential influence of hydrostatic pressure on cardiovascular function, as well as non-specific effects related to familiarity with repeated interventions and/or benefits of social interactions require consideration and future studies with control groups (e.g., thermoneutral water tub) to isolate passive heating effects from other influences.

Additionally, no significant effects of HT on glucose tolerance were observed in the current study. This outcome is consistent with Ely et al., where significant improvements in glucose tolerance in individuals with PCOS were not observed until after the midpoint (five weeks) of their study (24). In addition, Qiu et al showed no change in fasting glucose or improvements in glucose tolerance after four weeks of HWI in patients with varying degrees of stenosis (71). While several studies showed improvements in fasting glucose with only 2–4 weeks of HWI (23, 25, 33, 72), it is difficult to compare across these different study populations and additional research examining HWI in large, diverse participant groups is needed.

5.0. Conclusion

The current pilot study demonstrates the potential of HT as a potential therapeutic approach for older adults at risk of AD due to metabolic factors. Specifically, HT via HWI, was not only feasible but also enjoyable, achieving high adherence rates among participants. These outcomes were especially noteworthy given the limited exercise engagement typically observed in this demographic. The feasibility and high adherence for HT support its use as an alternative or supplement to exercise in managing health risks associated with metabolic dysfunction, a known contributing factor to cognitive decline. Considering the intertwined nature of metabolic and cognitive health, particularly in the context of aging populations, further research is essential. Future studies should aim to address current limitations and explore the broader implications of HT on cognitive functions, including its potential role in preventing or mitigating neurodegenerative diseases such as AD.

Appendix

Appendix Table 1.

Post-intervention survey

Survey question	Rating scale	Mean score
The study staff explained the study purpose and procedures with clarity	1–5	4.7 (0.6)
Receiving heat treatment 3 times a week was not a burden on my schedule	1–5	4.4 (1.0)
Receiving heat treatment for an additional 6 weeks would not be too long	1–5	3.9 (1.0)
My heat treatment was challenging but not painful	1–5	3.7 (1.4)
The ingestible temperature sensor was no harder to swallow than a multivitamin capsule	1–5	4.8 (0.4)
Overall I enjoyed participating in the FIGHT-AD study	1–5	4.8 (0.5)

Following the HT intervention, participants were asked to give ratings using a likert scale with the following values: Strongly disagree (1), Disagree (2), Neutral (3), Agree (4), Strongly agree (5)