Reply from Vienna E. Brunt, Matthew J. Howard, Michael A. Francisco, Brett R. Ely and Christopher T. Minson

We would like to thank Chiesa and colleagues for their letter (Chiesa et al. 2017) regarding our recent publication in The Journal of Physiology (Brunt et al. 2016). As these authors mention, our paper has received much positive attention, and we are excited that this therapeutic modality is gaining such wide interest. We believe passive heat therapy has considerable potential to reduce cardiovascular risk and prevent disease and mortality, and hope that healthy discussion such as this paves the way for eventual translation of heat therapy into clinical practice.

The topic of Chiesa et al.’s letter concerned our measurements of carotid intima media thickness (cIMT). In our paper, we reported a significant and relatively large reduction in carotid wall thickness (0.43 ± 0.01 to 0.37 ± 0.01 mm) following 8 weeks of passive heat therapy in young, sedentary subjects, whereas we observed no change in wall thickness of the superficial femoral artery (SFA). Chiesa et al. raised some excellent points regarding the limitations of the technique. Despite these limitations, we believe our results do reflect a physiological effect of heat therapy on conduit vessel remodelling.

We would first like to highlight the point made by these authors regarding the resolution of traditional ultrasound machines. Traditional ultrasound transducers operate at a maximal frequency of ∼10–12 MHz, which is not great enough to distinguish between the intimal and medial layers of blood vessels. We absolutely agree with Chiesa et al. on this point, and wish more investigators would recognize this important distinction. For this reason, we refer to our measurement in this response and in our paper (which was performed using a 10.0 MHz linear array ultrasound transducer) as ‘wall thickness’ rather than IMT as we believe this term more accurately reflects what typical ultrasounds are capable of measuring. We therefore acknowledge that the changes in wall thickness we observed following heat therapy could have been due to changes in the medial layer rather than (or in addition to) the intima.

There are a few plausible mechanisms by which heat therapy could have reduced wall thickness. First, heat shock proteins (HSPs), which are upregulated in response to acute heat exposure and chronically following heat acclimation, have numerous anti‐inflammatory effects that could reduce wall thickness. For example, Hsp70 has been shown to inhibit vascular smooth muscle cell hypertrophy (Zheng et al. 2006), which would reduce medial thickness. Hyperthermia, presumably through the action of HSPs, has also been shown to reduce expression of endothelial adhesion molecules (Kohn et al. 2002; Nakabe et al. 2015), which signal white blood cells to adhere to the intimal layer of the vessel walls. As cell adhesion is constantly in flux, a reduction in how many cells are adhered to the vessel walls is something that could be impacted fairly quickly and thus could lead to a physiological reduction in IMT within the time frame of an 8‐week heat therapy intervention. Along these lines, significant increases in carotid and SFA wall thickness of a similar magnitude to the changes we observed have been observed over an 8‐week bed rest study, which were totally (carotid) or partly (SFA) prevented by resistive vibration exercise (Van Duijnhoven et al. 2010).

Second, as Chiesa et al. point out, medial thickness is related to changes in blood pressure. Elevations in carotid distending pressure have been associated with carotid wall hypertrophy (Tanaka et al. 2001). We observed a significant reduction in mean arterial pressure following heat therapy (Brunt et al. 2016) and so it is possible this resulted in a concurrent reduction in carotid medial thickness. Furthermore, local distending pressures seem to play a larger role in arterial wall remodelling in the carotid artery, due to greater pulsatile distension, compared to peripheral arteries such as the SFA (Boutouyrie et al. 1999), which may help explain why we observed reductions in carotid wall thickness and not in SFA.

This point transitions to another comment made by Chiesa et al. We agree it is intriguing that we observed a reduction in wall thickness in the carotid artery, but not in superficial femoral artery, whereas, we observed greater effects of heat therapy on other variables, such as beta‐stiffness index, in the lower limbs. In addition to differential distending pressures being a potential explanation, there is some evidence that wall thickness adapts independently of local shear stresses and more so in response to systemic (perhaps circulating) factors (Thijssen et al. 2011). This phenomenon is highlighted by a study by Rowley et al. in which the authors compared wall thickness and conduit vessel diameters in wheelchair athletes, able‐bodied athletes who perform either predominantly upper body (canoeists) or lower body (runners and cyclists) exercise, and in sedentary able‐bodied and wheelchair user controls (Rowley et al. 2012). As expected, athletes who performed primarily upper body exercise (canoeists and wheelchair athletes) had larger brachial artery diameters than other groups. Athletes who performed predominantly lower body exercise (able‐bodied runners and cyclists) had the largest SFA diameters, while both active and sedentary wheelchair users had considerably smaller SFA diameters than all able‐bodied individuals. Common carotid diameter was similar across all groups. To the contrary, wall thickness in the brachial, superficial femoral and carotid arteries were comparable across all athletes and significantly lower than controls, indicating adaptations in wall thickness are independent of local shear stress. Rowley et al.’s results may explain why we observed a reduction in carotid artery wall thickness, in the absence of changes in carotid artery stiffness and in the presence of considerably lower shear rates during hot water immersion compared to other conduit arteries. Perhaps changes due to this systemic mechanism were not great enough for us to detect a significant reduction in SFA wall thickness, particularly in these young, healthy able‐bodied subjects.

We are not the first to report changes in carotid wall thickness but not SFA. Using resistive exercise or resistive exercise with vibration, Weber and colleagues (Weber et al. 2013) reported no statistically significant changes in SFA, but observed a statistically significant 4.2% reduction in carotid artery wall thickness after 6 weeks. As discussed in that paper, Thijssen et al. (2012) previously concluded that high exercise intensities or high exercise volumes (or comparable cardiovascular haemodynamics) are required to affect carotid wall thickness. The stimulus applied in our study satisfies both of these criteria, as subjects were heated to or above 38.5°C for 60 min, 4–5 times per week. This is a very significant thermal stress in terms of both intensity and volume of the stimuli.

Regarding concerns about methodology, we acknowledge that measuring wall thickness by caliper adds variability. The use of automated software would have been preferable had it been available to us at the time. However, we would like to point out that it is common to determine wall thickness with software that involves manual selection of wall layers, and several studies which have established wall thickness as a predictor of cardiovascular morbidity and mortality have done so using this approach (e.g. Bots et al. 1997; Anderson et al. 2011). All our measurements were made in triplicate along a section of vessel from three different angles, and measurements were video‐recorded to verify initial measurements, and a portion were analysed offline by a blinded investigator. We also kept the settings identical throughout testing from pre‐ to post‐heat therapy, as ultrasound settings can impact thickness measurements (Potter et al. 2008).

To summarize, the reduction in carotid wall thickness we observed following heat therapy may have been caused by (systemic) HSP‐mediated effects on the medial and/or intimal layers of the carotid artery, and may have been secondary to the reduction in blood pressure we observed. This latter mechanism, combined with a lower baseline thickness of the superficial femoral artery and thus a potential baseline effect, could explain why we observed a reduction in carotid wall thickness but not in SFA wall thickness.

We do, however, recognize that there are limitations to this technique. Therefore, it will be important for future investigations to confirm our results. If confirmed, we believe that the magnitude of reduction in carotid wall thickness we observed relative to reductions that have been reported with pharmacological interventions, such as statin treatment (Masoura et al. 2011), speaks to the robustness of heat therapy. We further opine that integrative, dynamic stressors such as exercise and heat therapy are likely to be much more robust stimuli, in terms of degree of remodelling and timing, for altering vascular health than pharmacological treatments. That is, heat therapy affects numerous physiological processes and thus should be expected to have a more profound effect than drug therapies with specific targets.

Additional information

Competing interests

None declared.

Author contributions

All authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.