Subjective and objective quality of sleep with radiant or convection cooling systems: a randomized, cross-over trial

Abstract

Sleep is influenced by environmental factors, and hot ambient temperature undermines sleep quality. Considering the global warming, it becomes increasingly important to keep bedroom temperature cool in the summer. Today, two electrical cooling systems are available: convection and radiant. Convection air cooling system supplies cooled air-flow. Radiant cooling system cools the room through radiation from the cooled surfaces, moisture condensation on which is prevented by reheated refrigeration cycle system. The present study was a randomized, cross-over trial setting room air temperature at 26 ℃ to compare subjective and objective quality of sleep between convection and radiant cooling in 7 healthy women. Subjective and objective quality of sleep were assessed by Oguri-Shirakawa-Azumi Sleep Inventory and EEG-based sleep parameters, respectively. Compared with convection cooling, radiant cooling provided lower relative humidity and air flow. Refreshness, one of the 4 indices of subjective quality of sleep, was higher with radiant cooling. Among the EEG-based sleep parameters, sleep efficiency was higher and sleep latency was shorter with radiant cooling. In spite of similar ambient room temperature, difference in cooling (radiant vs. convection) and/or dehumidifying (ordinary air conditioning vs. reheated refrigeration cycle) method provided difference in thermal comfort affecting subjective and objective quality of sleep.

Introduction

Subjectively and objectively measured poor sleep are associated with physical and mental health conditions such as cardiovascular disease, metabolic syndrome, and psychiatric disorders^1–6. In spite of the importance of the good sleep for good physical and mental health and a good quality of life, Japanese nationwide general population survey revealed high prevalence of sleep complaint².

Sleep is influenced by environmental factors such as acoustic, air quality, light and temperature. Indoor air quality affects sleep and next-day performance, which can be significantly improved by increasing the clean outdoor air supply rate in bedrooms⁷. Room acoustics also influence the sleep structures. Reduction of noise level and of reverberation leads to an increase in the amount of SWS and a reduction in nocturnal arousal⁸. Lighting and air conditioning system in our daily life have changed reflecting technological innovation and energy policy in our society. Incandescent and fluorescent light have been replaced with LED light for conservation of electricity, but LED light enriched with short wave length (blue light) raised concerns for its adverse effect on sleep^9,10.

Today, two electrical heating/cooling systems are available: convection and radiant. Convection air conditioning system supplies cooled or heated air-flow, which explains a possible discomfort due to direct hit by cool/warm air. On the other hand, radiant air conditioning system cools or heats ceiling or wall surfaces, which exchange heat with people in the room through radiation. Although the radiant air-conditioning systems also induces heat exchange through natural convection, this system doesn’t use air-blowing fan and includes less draft feeling. As a seminal social implementation trial in India, a new building was bult, half of which installed convection and the other half of which installed radiation cooling system to make a side-to-side comparison of energy cost over the year¹¹. Interestingly, survey on occupant comfort was better in a building with radiant cooling system. However, it was not cross-over experiment and the methods of survey for occupant comfort was not described.

Insomnia is characterized by difficulty in either initiating sleep, maintaining sleep continuity, or poor sleep quality assessed by subjective poor quality of sleep, according to the International Classification of Sleep Disorders (ICSD-3)¹². On the other hand, objective measure of sleep assessed by electroencephalogram (EEG) provides insight into sleep architecture of the subjects. Human sleep is consisted of four stages consist of non-REM sleep (N1, N2 and SWS) and REM sleep, which cyclically alternate¹³. As a fact worthy of note, individuals may report poor sleep but exhibit normal EEG sleep patterns, or vice versa¹⁴.

The present study was a randomized, cross-over trial investigating the subjective and objective quality of sleep in a bedroom with radiant and convection cooling systems. Subjective sleep quality was assessed using the Oguri-Shirakawa-Azumi Sleep Inventory Middle aged and Aged version (OSA-MA)¹⁵ which is commonly used in sleep studies of Japanese participants^16–19. EEG-based sleep parameters were assessed using a recently developed portable multichannel electroencephalography recorder at bedroom^14,20,21.

Methods

Participants

In this study, 7 healthy women were selected to participate based on inclusion criteria: age between thirties and sixties, BMI less than 30 (kg/m²) with self-reported poor sleep quality based on Athene insomnia scale (AIS ≧ 6), Epworth sleepiness scale (ESS ≦ 5). Exclusion criteria were recent shift work or transmeridian travel, smoking, excessive habitual alcohol intake (21.6 g of alcohol more than 3 times/week), ongoing medication for certain diseases, expecting vaccination during the study, and the use of medications affecting normal sleep. Physical characteristics of study participants are shown in Table 1. All methods were performed in accordance with the relevant guidelines and regulations. The study protocol was approved by the Chiyoda paramedical care clinic (approval number: 22111801). This study was registered in the University hospital Medical Information Network (UMIN) center, Japan on 19/01/2023, with study ID UMIN000050083. All participants provided written informed consent before study commencement.

Table 1.

Physical characteristics. BMI: body mass index, AIS: Athene insomnia scale, ESS: Epworth sleepiness scale.

Anthropometric variables	Mean ± SEM
Age (year)	50 ± 3
Weight (kg)	55 ± 3
BMI (kg/m2)	22 ± 1

Questionnaire variables	Mean ± SEM
AIS	7.3 ± 2.7
ESS	3.1 ± 1.2

Procedures

The present study was a randomized, cross-over trial investigating the effects of convective cooling system and radiant cooling system on subjective and objective quality of sleep of women. Participants maintained a regulatory sleep/wake schedule, following their habitual bed and wake times. During the experimental period, participants were instructed to abstain from consuming caffeine or alcohol. The study was conducted in 2 trials consisted with 4 consecutive nights, separated by 3 days washout period, as single-blind test (Fig. 1). Experiment was completed between July 24 and September 1, 2023.

Fig. 1.

Study protocol. This study was designed as a randomized, crossover intervention study to compare sleep in a bedroom with convective cooling system and radiant cooling system. The subjects were not informed which of the systems was in operation. In each session, participants slept 4 nights in an air-conditioned bedroom. Participants kept their ordinary daily life, including to go to work. Washout period between the 2 sessions were 3 days.

Sleep environment

Bed room (floor area 11.6 m² ceiling height 2.45 m) built as reinforced concrete structures was equipped with bed (W 127 cm, L 203 cm, H 35 cm), kitchen, bath and toilet. Two types of air conditioning system, convection cooling and radiant cooling, were installed in the room. For convection cooling, an ordinary air conditioning system for a space of 16.2 m² (CS-280DFL-W, Panasonic, Japan), which cools and dehumidifies room air, was used. Radiant cooling system consists with cooling and dehumidifying components as follows. A blue polypropylene tube was set on the ceiling, in which water cooled at 18 ℃ was circulated (Fig. 2). To prevent the moisture condensation around the polypropylene tube and maintain humidity of the room, reheated refrigeration cycle system (RAS-X J25M, Johnson Controls-Hitachi Air Conditioning, Tokyo) was used. Reheated refrigeration cycle system dehumidifies without affecting air temperature. With both air conditioning systems, air temperature near bed board was set at 26 ℃.

Fig. 2.

Air conditioning systems. For radiant cooling, blue polypropylene tube (bore 2.3 mm, outer diameter 3.4 mm, total length 905 m), in which water cooled at 18 ℃ was circulated, was set on the ceiling. Ordinary air conditioning system for convection cooling is surrounded by a red dotted square (left panel). Polypropylene tubing before attached to the ceiling was shown on right panel.

Measures

Sleep environment

During sleep assessment on 56 nights (4 nights x 2 conditions x 7 subjects) from 21:00 to 8:00, temperature and relative humidity were monitored with portable wireless logger set above the headboard of the bed (70 cm above the floor).

To further assess thermal environment of bedroom, temperature, relative humidity, radiant temperature, air flow, predicted mean vote (PMV) were assessed for 1 h under unoccupied condition, i.e., to maintain quiet environment for the sleep assessment by PMV monitor was performed only under an unoccupied condition. A portable PMV monitor (AM-101, Kyoto Electronics Manufacturing Co. Ltd.) put at 20 cm above the bed surface, i.e., 55 cm above the floor. Mean radiant temperature (MRT) was calculated according to the international standard of the International Organization for Standardization (ISO)²² as

1

where: t_g = globe temperature (°C), V_a = air velocity (ms⁻¹), t_a = air temperature (°C), D = globe diameter (mm) and є = emissivity (0.95 for a black globe). Thermal comfort was assessed as PMV which is an index that predicts the mean value of the votes of a large group of persons as 7-point thermal sensation scale (Table 2). PMV were calculated according to the international standard of the International Organization for Standardization (ISO)²³ as.

2

3

Table 2.

Seven-point thermal sensation scale.

PMV	Thermal sensation
+3	Hot
+2	Warm
+1	Slightly warm
0	Neutral
−1	Slightly cool
−2	Cool
−3	Cold

4

5

where: = metabolic rate (W/m²), = effective mechanical power (W/m²), = clothing insulation(), = clothing surface area factor, = air temperature (℃), = mean radiant temperature (℃), = relative air velocity (m/s), = water vapour partial pressure (Pa), = convective heat transfer coefficient (W/()), = clothing surface temperature (℃). The PMV evaluation in this study was conducted on the indoor environment only, rather than on the comfort assessment of each individual subject, and the metabolic rate = 40.74 (W/m² while sleeping) and clothing insulation = 0.2325 (m²·K/W, including bedding) were set as common fixed values. Thermal sensation scale for PMV is shown in Table 2.

Subjective quality of sleep

Subjective sleep quality was assessed using the Oguri-Shirakawa-Azumi Sleep Inventory Middle aged and Aged version (OSA-MA) the morning after each experimental night¹⁵. The OSA-MA is a standardized psychological scale to evaluate sleep introspection upon waking up comprising 16 items, which were categorized into the following five factors: Factor I, “sleepiness on rising”; Factor II, “initiation and maintenance of sleep”; Factor III, “frequent dreaming”; Factor IV, “refreshing”; and Factor V, “sleep length”. For all factors, a higher score indicates better sleep.

Objective quality of sleep

A portable multichannel electroencephalography recorder was used in the present study (InSomnograf K2; S’UIMIN Inc., Tokyo, Japan)¹⁴. The recording system consisted of five electroencephalogram derivations (Fp1–M2, Fp2–M1, Fp1–average M, Fp2–average M, and Fp1–Fp2). Sleep parameters were categorized at 30-s intervals as wakefulness, non-Rem sleep (N1, N2 and N3) and REM sleep, according to the standard criteria of the American Academy of Sleep Medicine²⁴.

Statistical analysis

The results are expressed as mean ± standard deviation (SD). A paired t-test was used to compare the mean values of temperature and relative humidity between the radiant cooling and convection cooling trials. The nonparametric Wilcoxon signed-rank tests were used to compare the mean values of subjective (the OSA-MA parameters) and objective (the sleep parameters) sleep between the mean value of the trials, and differences were considered significant when the error probability was less than 0.05. Data analysis was conducted using Prism 10 (GraphPad Software, San Diego, CA).

Results

Sleep environment

During sleep assessment on 56 nights from 21:00 to 8:00, temperature and relative humidity was continuously monitored with data logger set above the headboard of the bed (70 cm above the floor) (Fig. 3). Average of room temperature for radiant cooling (25.99 ± 0.33 ℃) was slightly higher than that of convection cooling (25.53 ± 0.42 ℃, p < 0.01). Stability of room temperature assessed as standard deviation of room temperature during each night was 0.14 ± 0.05 ℃ and 0.22 ± 0.12 ℃ for radiant cooling and convection cooling, respectively (p < 0.01). Average relative humidity for radiant cooling (60.1 ± 3.7%) was lower than that of convection cooling (79.9 ± 4.4%, p < 0.01). Standard deviation of relative humidity during each night was 1.9 ± 1.0% and 2.1 ± 0.7% for radiant cooling and convection cooling, respectively (p > 0.05).

Fig. 3.

Temperature and humidity throughout the night. Values are mean ± SD. Room temperature and relative humidity throughout the night are shown as hourly average in each solid line. The gray line represents the convection cooling condition, while the black line represents the radiant cooling condition.

To further assess vote environment of bedroom, temperature, relative humidity, radiant temperature, air flow and PMV were assessed for 1 h under unoccupied condition (Fig. 4). Although statistical analysis was not applied because of a single trial of measurement for each cooling condition, observation on temperature and relative humidity during sleep assessment was confirmed. Temperature was slightly lower and more stable with radiant cooling system compared with that with convection cooling system. Air flow with radiant cooling was lower than that of convection cooling. In a convection cooling condition, air flow increases and relative humidity decrease when room temperature is decreasing. Average radiant temperature with radiant cooling was slightly lower than that with convection cooling. PMV with radiant cooling was lower than that of convection cooling. Average values for 60 min measurement are presented in Table 3.

Fig. 4.

Sleep environment. Room temperature, radiant temperature, relative humidity, air flow and PMV were measured for 60 min under unoccupied condition using a portable PMV monitor put at 20 cm above the bed surface. The gray line represents the convection cooling condition, and the black line represents the radiant cooling condition. Gray shades indicate periods when room temperature is decreasing in a convection cooling condition.

Table 3.

Average room temperature, radiant temperature humidity, air flow and PMV. Values are mean ± SD of min values for 60 min. Statistical analysis on the difference of mean values was not applied because of a single trial of measurement for each cooling condition.

	Convection	Radiant
Room temperature (℃)	25.85 ± 0.17	25.98 ± 0.07
Radiant temperature (℃)	25.85 ± 0.257	25.35 ± 0.081
Relative humidity (%)	74.94 ± 2.24	47.64 ± 0.12
Air flow (m/sec)	0.105 ± 0.041	0.045 ± 0.021
PMV	0.278 ± 0.083	−0.032 ± 0.024

Subjective quality of sleep

Average over 4 experimental days for indexes of subjective quality of sleep were calculated, and refreshing (Factor IV) was significantly better in radiation cooling condition compared with those of convection cooling condition (Fig. 5).

Fig. 5.

Subjective quality of sleep. Values are mean ± SD. The asterisk represents a statistically significant difference between the convection and radiant trials by a Wilcoxon test (*p < 0.05).

Objective quality of sleep

Time course of sleep architecture (Fig. 6) was shown. Out of 56 measurements (4 nights for 2 trials of 7 subjects), recordings of 2 nights for convection cooling conditioning and 1 night for radiant cooling condition was incomplete, and excluded from the statistical analysis. In both trials, the time course of the sleep architecture showed characteristic changes; slow-wave sleep gradually decreased and was replaced by nonrapid eye movement (NREM) stage 2, REM sleep and wake. A closer look at the time course of sleep architecture revealed that decline of wake after bedtime is faster and its increase before awakening occurs later in trial with radiant cooling condition compared to those with convection cooling condition.

Fig. 6.

Time course of sleep architecture. Cumulative sleep architecture during the trial with convection cooling (upper panel) and that with radiant cooling condition (bottom panel). The percentage of subjects in stage W (wakefulness; black), stage N1 (gray), stage N2 (light blue), SWS (dark blue), and stage REM (red) changed with the sleep time.

Indexes of objective quality of sleep were calculated as averages of 4 or 3 nights of EEG recording. Among indexes, sleep efficiency (SE) and sleep latency SL) were significantly better in radiation cooling condition compared with those of convection cooling condition (Fig. 7).

Fig. 7.

Objective quality of sleep. Values are mean ± SD. The asterisk represents a statistically significant difference between the convection and radiant trials by a Wilcoxon test (*p < 0.05).

Discussion

Bedroom environment with convection and radiant cooling system

Four basic environmental factors, which define the human thermal environment and its sensation of thermal comfort, are ambient temperature, radiant temperature, humidity and air flow²⁵. Compared with convection cooling, radiant cooling provided slightly higher ambient room air temperature, but it may provide slightly cooler radiant temperature due to radiant heat transfer independent of room temperature. Convection cooling supplies cooled air-flow, while radiant cooling exchange heat with floor, walls and people in the room through radiation, which subsequently generates natural convection, i.e., radiant cooling provides lower air flow and lower radiant temperature. Radiant cooling doesn’t rely on air-blowing fan, and the action of radiant cooling on room air temperature was reflected as narrower range of room air temperature compared with that of convection cooling.

As a common practice to prevent moisture condensation on cooling tubing for radiant cooling, dehumidifier was adopted in the present study. For the convection cooling system, a single air conditioning system serves roles of cooling and dehumidification, on and off of which apparently synchronized; when convection cooling became off, dehumidification was also turned off. This intermittent interruption of dehumidification with convection cooling, at least partially, explains higher relative humidity with convection cooling condition in the present study.

Although room air temperature was similar between the two colling conditions, lower relative humidity and air flow with radiant cooling consequently led to a lower PMV; thermal environment more comfortable than that of convection cooling. In a recent study, better thermal comfort with radiant cooling was also observed in a wake state²⁶.

Sleep quality

The major finding of the present study is that differences in cooling system of the bedroom affected subjective and objective quality of sleep. With radiant cooling system, subjective refreshing in the morning was better than those with blower type air condition system. Although differences in other subjective quality of sleep did not reach statistical significance, all subjective quality of sleep with radiant cooling scored higher than those of convection cooling condition. It is worth mentioning that insomnia is characterized by difficulty in either initiating sleep, maintaining sleep continuity, or poor sleep quality, i.e., assessed by subjective poor quality of sleep, according to the International Classification of Sleep Disorders (ICSD-3)¹².

Discrepancy between subjective and objective sleep assessments, known as sleep misperception, warrants objective measure of sleep, which provides insight into sleep architecture of the subjects. A better profile of sleep was also shown as shorter sleep latency and higher sleep efficiency with radiant cooling system compared with those with convection type air condition system. Taken together, radiant cooling system provide environment for better sleep compared with convection type air condition system.

Bedroom environment and sleep quality

Thermoregulatory processes have long been implicated in the initiation of sleep²⁷. Transfer of heat from proximal to distal skin assessed as distal-proximal skin-temperature gradient is a good predictor for sleep onset latency²⁸. It is also known that preoptic area of the hypothalamus serves as an essential brain region to coordinate sleep and body temperature^29–31. To get insight into the mechanism how radiant cooling influences sleep quality, measurement of distal-proximal skin-temperature gradient and core body temperature are appropriate in the future study.

Range of ambient temperature between 17 and 28 ℃ and relative humidity between 40 and 60% are assumed for the optimal design for the bedroom environment³². Assessment of sleep in the present study was performed at the higher end of comfortable bedroom temperature at ~ 26 ℃. Lower relative humidity with radiant cooling may contribute to increase evaporative heat loss leading to shortened sleep latency. Convective heat transfer is the transfer of heat from one place to another due to the movement of air. On the other hand, radiant cooling doesn’t depend on the air flow in the room, which may explain more stable room temperature providing a better environment for sleep compared with blower type air conditioning. It is plausible that stable room temperature and low air flow provided better environment for sleep resulting better subjective quality of sleep and higher sleep efficiency.

Interpretation of the present results requires cautious interpretation. In the present experiment, reheated refrigeration cycle system, which dehumidifies the air without influencing air temperature, was additionally adopted for radiant cooling condition. As a result, relative humidity of the room air was not matched; radiant cooling system provided lower humidity compared with convection fooling system. Previous study has shown that high relative humidity (~ 70–80%) can negatively impact sleep quality and be uncomfortable for human. High relative humidity decreases skin temperature, affecting the thermoregulatory system. As a result, a decreased skin temperature has been found positively related to TST, sleep efficiency, and duration of REM sleep, and negatively correlated to WASO³³. It requires further study to attribute differences in objective and subjective quality of sleep observed in the present study to (1) essential difference in cooling system; convection vs. radiant heat transfer and/or (2) differences associated with two types of cooling system in the present study; humidity and stability of temperature. Experimental protocol to match relative humidity and its stability in two cooling conditions is warranted. For example, it is possible to install the reheated refrigeration cycle system with an ordinary air conditioning. Furthermore, the effects of the two cooling manners on cognitive performance in following daytime after sleep would provide valuable insights.

Limitation of the study

Protocol of the present study evaluated subjective and objective quality of habitual sleep of the subjects, and bedtime and wake up time was not regulated. Average of total sleep time was less than 6 h suggesting that subjects in the present study didn’t have enough sleep time, reflecting busy daily schedule of working age population. Total sleep time of the subjects is consistent with Japanese average sleep time, but shorter than that of many other countries³⁴. To generalize the present results, experiments with a larger sample and sufficient sleeping period remain to be performed. Secondly, heating system is used during the winter in many countries, and comparison of radiant heating and convection heating system is warranted.

Author contributions

T.K., M.M. and M.Y. conceptualized the study. R.M., H.Y., M.O., S.N. and C.T. performed the experiment.I.P., Y.C., F.K. and M.T. analyzed the data. I.P. and K.T. wrote the manuscript, and all authors reviewed the manuscript.

Data availability

All datasets presented in the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

F.K. received an honorarium from S’UIMIN Inc. T.K. and M.Y. are board members as well as stakeholders of S’UIMIN Inc. All the remaining authors declare no conflict of interest.