2022 Volume 29 Issue 12 Pages 1791-1807
Aim: Issuance of the WHO Housing and health guidelines has paralleled growing interest in the housing environment. Despite accumulating evidence of an association between outdoor temperature and serum cholesterol, indoor temperature has not been well investigated. This study examined the association between indoor temperature and serum cholesterol.
Methods: We collected valid health checkup data of 2004 participants (1333 households), measured the indoor temperature for 2 weeks in winter, and divided participants according to whether they lived in a warm (average bedroom temperature ≥ 18℃), slightly cold (12–18℃) or cold house (＜12˚C). The relationship between bedroom temperature and serum cholesterol was analyzed using multivariate logistic regression models, adjusting for demographics, lifestyle habits and the season in which the health checkup was conducted, with a random effect of climate areas in Japan.
Results: The sample sizes for warm, slightly cold, and cold houses were 206, 940, and 858, respectively. Compared to those in warm houses, the odds ratio of total cholesterol exceeding 220 mg/dL was 1.83 (95%CI: 1.23–2.71, p=0.003) for participants in slightly cold houses and 1.87 (95%CI: 1.25–2.80, p=0.002) in cold houses. Similarly, the odds ratio of LDL/non-HDL cholesterol exceeding the standard range was 1.49 (p=0.056)/1.67 (p=0.035) for those in slightly cold houses and 1.64 (p=0.020)/1.77 (p=0.021) in cold houses. HDL cholesterol and triglycerides were not significantly associated with bedroom temperature.
Conclusion: Besides lifestyle modification, improving indoor thermal environment through strategies such as installing high thermal insulation and appropriate use of heating devices may contribute to better serum cholesterol condition.
See editorial vol. 29: 1704-1705
A list of all the Smart Wellness Housing survey group members is shown in Supplemental Table 1.
Clinical trial registration number: UMIN000030601.
Cardiovascular disease (CVD) continues to be the leading cause of death worldwide. In addition to hypertension (raised blood pressure), hyperlipidemia (raised serum cholesterol) is also considered a key intermediate risk factor for CVD1, 2). The Framingham Heart Study identified risk factors of coronary heart disease (CHD)3, 4) and developed a risk score that includes blood pressure and serum cholesterol to estimate the 10-year risk of CHD5). In Japan, the Suita Study developed a risk score in a similar manner to the Framingham Heart Study for the Japanese population6, 7). The Evidence for Cardiovascular Prevention from Observational Cohorts in Japan (EPOCH-JAPAN) also examined lifetime risk of CHD using blood pressure and serum cholesterol8). Stroke is also an important outcome, with the Asia Pacific Cohort Studies Collaboration, which combines data from 29 cohorts in Asia and Australia, demonstrating an association between serum cholesterol and the risk of ischemic stroke9). Therefore, reducing cholesterol level and blood pressure is essential for preventing CVD. While serum cholesterol is affected by lifestyle-related factors such as diet10-13), exercise14, 15), and smoking16-18), effects of improving lifestyle are limited because they depend on individual effort. A potentially complementary approach of improving one’s life environment could enhance efforts to prevent CVD.
Consistent with the above context and the fact that modern humans spend between 60% and 70% of their time at home19-21), the World Health Organization (WHO) has issued Housing and health guidelines22). The guideline focuses on “low indoor temperature and CVD” and summarizes studies on the relationship between indoor temperature and blood pressure but not serum cholesterol. A previous review23) reported that one possible cause of increased serum cholesterol is air temperature, and evidence on the association between outdoor temperature and blood lipids has recently increased24-27). However, whether or not indoor temperature is associated with serum cholesterol remains largely unclear. Given that indoor temperature, unlike outdoor temperature, is a controllable factor, evidence of an association between indoor temperature and serum cholesterol would be beneficial for preventing CVD. Such a finding may be particularly pertinent in Japan, where 30% of about 50 million existing houses are not insulated and only 11% of houses were sufficiently insulated such that they meet the 1999 standards (the highest thermal insulation standards in Japan) as of 2018 28), indicating substantial room for improvement in managing the indoor thermal environment. In fact, a nationwide survey on indoor temperature revealed that the average living room temperature in Japan was 16.8°C 29), which is below the WHO recommendation of 18°C 22). Such poor indoor thermal environments could have adverse health effects.
We conducted a nationwide non-randomized controlled trial on housing and health in Japan, named the Smart Wellness Housing (SWH) survey. In this paper, health checkup data obtained through the baseline survey were used to investigate the relationship between indoor temperature at home and serum cholesterol level according to several indices, namely total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), triglycerides (TG), and non-HDL-C.
The study was conducted according to the principles of the Declaration of Helsinki. The study protocol and informed consent procedure were approved by the ethics committee of the Hattori Clinic Institutional Review Board (Approval No. S1410-J03). The study protocol was registered at the University Hospital Medical Information Network Clinical Trials Registry (UMIN000030601). All of the participants provided written informed consent to participate and to have their data published as a group.
Study DesignThe aims and study design of the SWH survey are reported elsewhere30). Briefly, this survey was conducted as a non-randomized controlled trial with an insulation retrofitting group and non-insulation retrofitting (control) group to examine the cardiovascular health benefits of living in insulation retrofitted houses. Participants were recruited by construction companies throughout all 47 prefectures of Japan. Inclusion criteria were: (1) intention to conduct insulation retrofitting, (2) age over 20 years, and (3) pre-renovation house which did not meet S (Supreme) standards of the “Act on the Promotion of Dissemination of Long-Lasting Quality Housing” in Japan. The sample size of each prefecture is shown in Supplemental Fig.1 In this paper, we performed a cross-sectional analysis of data from the baseline (before insulation retrofitting) survey conducted in fiscal years 2014 to 2017. We focused on data obtained before insulation retrofitting to reflect the actual condition of houses in Japan, most of which have low insulation performance.
Sample size of each prefecture in Japan
Participants were asked to submit the results of their health checkup. Among the items examined in the health checkup, we used TC, LDL-C, HDL-C, and TG. In Japanese health checkups, blood samples are collected after a fasting time of at least 10 hours in accordance with the Program for General Health Checkup and Guidance issued by the Ministry of Health, Labour and Welfare (MHLW). After excluding TG data ≥ 400 mg/dL, TC level was estimated using the Friedewald equation as follows31): TC (mg/dL)=LDL-C (mg/dL)＋HDL-C (mg/dL)＋TG (mg/dL)/5. Non-HDL-C, which is included in the current Japanese guideline32) and closely associated with arteriosclerosis and CVD33-37), was calculated using the following equation: Non-HDL-C (mg/dL)=TC (mg/dL)−HDL-C (mg/dL).
Indoor and Outdoor Temperature MeasurementsIndoor temperature and relative humidity at 1.0 m above the floor were measured in the living room and bedroom at 10-min intervals (TR-72wf; T&D Corp., Nagano, Japan) in the winter season (November−March). The temperature and humidity logger was installed such that it was out of direct sunlight and far away from heating equipment or heat-generating devices like refrigerators and televisions to avoid extreme outliers. Outdoor temperature data were obtained from the closest local meteorological observatory to each participant’s house.
Questionnaire and Diary SurveysThe questionnaire survey included housing conditions; demographics such as age, sex, height, weight, household income; lifestyle indicators such as dietary habits, exercise, smoking, alcohol consumption; and health conditions related to CVD. Participants indicated their household income by choosing from multiple options which were subsequently classified as low (＜2 million Japanese yen (JPY)), middle (2−6 million JPY) or high (≥ 6 million JPY) in accordance with the National Health and Nutrition Survey. The salt check sheet score and frequency of vegetable intake were used as a measure of dietary habits. The salt check sheet score was classified into 4 groups (low (0−8 points), medium (9−13 points), high (14−19 points), or very high (≥ 20 points))38) while responses related to vegetable intake were categorized according to whether or not participants ate vegetables regularly. Although the salt check sheet was not developed to assess serum cholesterol, it was used to indirectly evaluate whether or not participants’ dietary habits were healthy. We confirmed in advance that some cholesterol indices gradually increased (HDL-C decreased) with higher salt check sheet score, as shown in Supplemental Table 2. Responses related to exercise, smoking and alcohol consumption were treated as two-valued variables: whether or not participants did regular moderate exercise, whether or not they were a current smoker, and whether or not they were a current drinker, respectively. A diary survey was also conducted, in which participants provided details of their waking time, bedtime, and time spent at home on a daily basis.
| Variable | Low (0−8 points) | Medium (9−13 points) | High (14−19 points) | Very high (≥ 20 points) | p for trend |
|---|---|---|---|---|---|
| Ave (SD) | Ave (SD) | Ave (SD) | Ave (SD) | ||
| Blood lipids | |||||
| TC | 209 (38) | 209 (38) | 209 (35) | 210 (37) | 0.990 |
| LDL-C | 122 (32) | 123 (32) | 125 (31) | 125 (31) | 0.218 |
| HDL-C | 68 (17) | 66 (16) | 62 (16) | 61 (17) | ＜0.001 |
| TG | 97 (52) | 103 (59) | 111 (65) | 117 (65) | 0.002 |
| Non-HDL-C | 141 (36) | 144 (36) | 147 (36) | 149 (38) | 0.028 |
TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride.
First, cold, slightly cold and warm houses were defined. We extracted bedroom temperature during sleep based on a diary survey because participants, especially those of working age, are more likely to spend long periods of time in the bedroom when they are at home. Subsequently, we divided participants into three groups according to the average bedroom temperature: ≥ 18°C (warm houses), 12−18°C (slightly cold houses), and ＜12°C (cold houses), in accordance with the temperature recommended by the WHO Housing and health guidelines (18°C)22) and a UK report on the temperature at which CVD risk begins to increase (12°C)39).
To examine the association between serum cholesterol indices and bedroom temperature, the Jonckheere-Terpstra trend test and the Cochran-Armitage trend test were performed on continuous and two-valued variables, respectively. Univariate and multivariate logistic regression analyses were also conducted with serum cholesterol as the objective variable (whether or not each item was within the standard range of the current32) or previous40) Japan Atherosclerosis Society Guidelines: TC ≥ 220 mg/dL, LDL-C ≥ 140 mg/dL, HDL-C ＜40 mg/dL, TG ≥ 150 mg/dL, and non-HDL-C ≥ 170 mg/dL). Bedroom temperature (warm vs slightly cold vs cold houses) was inputted as the explanatory variable. The analysis was adjusted for the participants’ basic characteristics, namely age, sex, body mass index (BMI, calculated as weight[kg]/height[m]2), household income, salt check sheet score, vegetable intake, exercise, smoking, alcohol consumption and antihypertensive drug use. The season in which the health checkup was conducted (in winter or not) was also inputted into the model to account for seasonal variations in serum cholesterol23, 41, 42). Additionally, we included climate areas in Japan (Supplemental Fig.2) as a random effect to account for regional differences. Sub-group analyses of participants without hyperlipidemia were also conducted to exclude the effects of medical treatment for cholesterol and confirm the robustness of the results.
HDD18-18 means the heating degree days value (difference between indoor temperature of 18°C and daily mean outdoor temperature, where the daily mean outdoor temperature is less than 18°C
All P values were two sided, and a two-sided P value less than 0.05 was considered statistically significant. All analyses were performed using SPSS Ver. 26 (SPSS Inc., Chicago, Illinois, USA).
Fig.1 shows the selection of valid samples. Of 3775 participants in the SWH survey, 2230 residents submitted health checkup data and 2004 residents (1333 households) were included as valid samples. Distribution of the bedroom temperature during sleep is shown in Fig.2. The average bedroom temperature was 12.7℃. Participants were categorized as sleeping in warm, slightly cold or cold houses based on the results shown in Fig.2. The sample sizes of the three groups were 206, 940 and 858, respectively.
LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride.
Distribution of average bedroom temperature
Table 1 shows the housing conditions. Participants in Area 2 (which mostly consists of Hokkaido) lived in warmer houses than those in other areas. Older type heating devices such as oil fan heaters and gas stoves were more frequently used in colder houses. About 40% of participants reported feeling cold in the bedroom, and this percentage gradually increased with colder houses. Participants’ average duration of residence in their current house was 25.8 years, with a difference of 10 years between warm and cold houses. The average outdoor temperature was 5.8℃ during the measurement period, and there was no clear association between bedroom temperature (warm vs slightly cold vs cold) and outdoor temperature.
| Variable | Overall (n = 2004) | Warm (n = 206) | Slightly cold (n = 940) | Cold (n = 858) |
|---|---|---|---|---|
| n (%) | n (%) | n (%) | n (%) | |
| Climate area | ||||
| Area 2 | 94 (5) | 40 (19) | 44 (5) | 10 (1) |
| Area 3 | 59 (3) | 6 (3) | 17 (2) | 36 (4) |
| Area 4 | 187 (9) | 3 (1) | 59 (6) | 125 (15) |
| Area 5 | 533 (27) | 32 (16) | 229 (24) | 272 (32) |
| Area 6 | 1025 (51) | 114 (55) | 527 (56) | 384 (45) |
| Area 7 | 106 (5) | 11 (5) | 64 (7) | 31 (4) |
| Heating devices in bedroom | ||||
| Oil fan heater, gas stove, etc. | 422 (21) | 30 (15) | 204 (22) | 188 (22) |
| Floor heating, heated floor mat, etc. | 156 (8) | 20 (10) | 85 (9) | 51 (6) |
| Self-rated coldness | ||||
| Feeling cold in bedroom | 765 (39) | 62 (31) | 354 (38) | 349 (41) |
| Variable | Overall (n = 2004) | Warm (n = 206) | Slightly cold (n = 940) | ColdCold (n = 858) |
| Ave (SD) | Ave (SD) | Ave (SD) | Ave (SD) | |
| Duration of residence in house, years | 25.8 (15.3) | 19.7 (13.4) | 23.5 (14.7) | 29.7 (15.5) |
| Outdoor temperature, ℃ | 5.8 (3.5) | 5.8 (4.3) | 6.9 (3.6) | 4.5 (2.7) |
Table 2 shows the baseline characteristics of the residents overall and by group. The average age was 58 years and approximately 30% were 65 years and older. About half were men, the average BMI was 22.8 kg/m2 and approximately 20% were classified as overweight. Approximately 40% of participants belonged to high income households, which is greater than the proportion reported by the National Health and Nutrition Survey in Japan (27.4%). While the number of patients with stroke, angina/myocardial infarction and diabetes was small, those with hyperlipidemia and hypertension comprised more than 10%. The number of participants with systolic blood pressure exceeding 140 mmHg or diastolic blood pressure exceeding 90 mmHg in the health checkup was smaller than that with self-reported hypertension, indicating that some participants received antihypertensive drugs and had well controlled BP. Basic characteristics were adjusted in the subsequent multivariate logistic regression model to account for differences between the three groups.
| Variable | Overall (n = 2004) | Warm (n = 206) | Slightly cold (n = 940) | Cold (n = 858) |
|---|---|---|---|---|
| n (%) | n (%) | n (%) | n (%) | |
| Demographics | ||||
| Age (≥ 65 years) | 611 (30) | 49 (24) | 288 (31) | 274 (32) |
| Men | 1008 (50) | 98 (48) | 475 (51) | 435 (51) |
| Body mass index (≥ 25 kg/m2) | 412 (21) | 38 (18) | 196 (21) | 178 (21) |
| Household income | ||||
| Low (＜2 million JPY) | 169 (9) | 10 (5) | 80 (9) | 79 (10) |
| Middle (2−6 million JPY) | 914 (50) | 87 (47) | 407 (47) | 420 (53) |
| High (≥ 6 million JPY) | 762 (41) | 90 (48) | 375 (44) | 297 (37) |
| Lifestyle | ||||
| Salt check sheet | ||||
| Low (0−8 points) | 254 (13) | 28 (14) | 112 (13) | 114 (14) |
| Medium (9−13 points) | 775 (41) | 80 (41) | 370 (41) | 325 (40) |
| High (14−19 points) | 739 (39) | 73 (37) | 348 (39) | 318 (39) |
| Very high (≥ 20 points) | 138 (7) | 14 (7) | 62 (7) | 62 (8) |
| Eat vegetables regularly | 1549 (78) | 163 (80) | 710 (76) | 676 (79) |
| Regular exercise | 646 (32) | 61 (30) | 281 (30) | 304 (36) |
| Current smoker | 270 (15) | 24 (12) | 128 (15) | 118 (15) |
| Current drinker | 1095 (55) | 113 (55) | 523 (56) | 459 (54) |
| Antihypertensive drug use | 481 (25) | 37 (19) | 225 (25) | 219 (27) |
| Health condition | ||||
| Stroke | 29 (1) | 4 (2) | 12 (1) | 13 (2) |
| Angina/Myocardial infarction | 62 (3) | 4 (2) | 25 (3) | 33 (4) |
| Diabetes | 140 (7) | 17 (9) | 62 (7) | 61 (7) |
| Hyperlipidemia | 377 (19) | 25 (13) | 183 (20) | 169 (21) |
| Hypertension | 463 (24) | 34 (17) | 222 (24) | 207 (25) |
| Health checkup | ||||
| Systolic blood pressure ≥ 140 mmHg | 295 (16) | 29 (15) | 145 (16) | 121 (15) |
| Diastolic blood pressure ≥ 90 mmHg | 168 (9) | 14 (7) | 82 (9) | 72 (9) |
Table 3 shows the average and standard deviation of serum cholesterol indices and Table 4 shows the number of participants with serum cholesterol that exceeded the standard range in the baseline survey. According to data from both tables, TC significantly increased with colder houses. Although not significant (p＜0.10), there was an increasing trend in LDL-C and non-HDL-C with colder houses.
| Variable | Overall (n = 2004) | Warm (n = 206) | Slightly cold (n = 940) | Cold (n = 858) | p for trend |
|---|---|---|---|---|---|
| Ave (SD) | Ave (SD) | Ave (SD) | Ave (SD) | ||
| Blood lipids | |||||
| TC | 209 (37) | 202 (37) | 209 (36) | 211 (38) | 0.007 |
| LDL-C | 124 (32) | 119 (31) | 123 (31) | 125 (33) | 0.061 |
| HDL-C | 64 (17) | 63 (16) | 64 (17) | 65 (17) | 0.130 |
| TG | 106 (61) | 100 (57) | 106 (59) | 107 (63) | 0.593 |
| Non-HDL-C | 145 (36) | 140 (36) | 145 (35) | 146 (37) | 0.083 |
TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride.
| Variable | Overall (n = 2004) | Warm (n = 206) | Slightly cold (n = 940) | Cold (n = 858) | p for trend |
|---|---|---|---|---|---|
| n (%) | n (%) | n (%) | n (%) | ||
| Blood lipids | |||||
| TC ≥ 220 mg/dL | 725 (36) | 57 (28) | 345 (37) | 323 (38) | 0.032 |
| LDL-C ≥ 140 mg/dL | 590 (29) | 52 (25) | 270 (29) | 268 (31) | 0.071 |
| HDL-C ＜40 mg/dL | 98 (5) | 16 (8) | 56 (6) | 26 (3) | ＜0.001 |
| TG ≥ 150 mg/dL | 369 (18) | 31 (15) | 182 (19) | 156 (18) | 0.662 |
| Non-HDL-C ≥ 170 mg/dL | 458 (23) | 34 (17) | 218 (23) | 206 (24) | 0.061 |
TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride.
The results of logistic regression analysis are shown in Table 5. Compared to participants living in warm houses (≥ 18℃), the adjusted odds ratio of exceeding the standard range for TC was 1.83 (95%CI: 1.23–2.71, p=0.003) for those living in slightly cold houses (12−18℃) and 1.87 (95%CI: 1.25–2.80, p=0.002) for those in cold houses (＜12℃). Similarly, the adjusted odds ratio of exceeding the standard range for LDL-C/non-HDL-C was 1.49 (95%CI: 0.99–2.25, p=0.056)/1.67 (95%CI: 1.04–2.68, p=0.035) for those living in slightly cold houses and 1.64 (95%CI: 1.08–2.49, p=0.020)/1.77 (95%CI: 1.09–2.87, p=0.021) for those living in cold houses. HDL-C and TG were not significantly associated with bedroom temperature.
| Objective variable | Explanatory variable | Unadjusted | Adjusted＊ | ||||
|---|---|---|---|---|---|---|---|
| Odds ratio | (95%CI) | p value | Odds ratio | (95%CI) | p value | ||
| TC ≥ 220 mg/dL | Warm (≥ 18℃) | Ref. | Ref. | ||||
| Slightly cold (12−18℃) | 1.77 | (1.20, 2.61) | 0.004 | 1.83 | (1.23, 2.71) | 0.003 | |
| Cold (＜12℃) | 1.80 | (1.22, 2.66) | 0.003 | 1.87 | (1.25, 2.80) | 0.002 | |
| LDL-C ≥ 140 mg/dL | Warm (≥ 18℃) | Ref. | Ref. | ||||
| Slightly cold (12−18℃) | 1.50 | (1.00, 2.24) | 0.051 | 1.49 | (0.99, 2.25) | 0.056 | |
| Cold (＜12℃) | 1.60 | (1.06, 2.40) | 0.025 | 1.64 | (1.08, 2.49) | 0.020 | |
| HDL-C ＜40 mg/dL | Warm (≥ 18℃) | Ref. | Ref. | ||||
| Slightly cold (12−18℃) | 0.93 | (0.48, 1.83) | 0.840 | 0.87 | (0.45, 1.69) | 0.678 | |
| Cold (＜12℃) | 0.62 | (0.30, 1.27) | 0.189 | 0.59 | (0.29, 1.19) | 0.138 | |
| TG ≥ 150 mg/dL | Warm (≥ 18℃) | Ref. | Ref. | ||||
| Slightly cold (12−18℃) | 1.51 | (0.94, 2.44) | 0.090 | 1.43 | (0.87, 2.36) | 0.163 | |
| Cold (＜12℃) | 1.29 | (0.79, 2.10) | 0.309 | 1.29 | (0.77, 2.15) | 0.327 | |
| Non-HDL-C ≥ 170 mg/dL | Warm (≥ 18℃) | Ref. | Ref. | ||||
| Slightly cold (12−18℃) | 1.71 | (1.08, 2.73) | 0.024 | 1.67 | (1.04, 2.68) | 0.035 | |
| Cold (＜12℃) | 1.75 | (1.09, 2.81) | 0.021 | 1.77 | (1.09, 2.87) | 0.021 | |
＊Adjusted for age, sex, body mass index, household income, salt check sheet score, vegetable consumption, exercise, smoking, alcohol consumption, antihypertensive drug use and season in which health checkup was conducted.
TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride.
The results of sub-group analyses of participants without hyperlipidemia are shown in Table 6. As with the results shown in Table 5, the adjusted odds ratios of exceeding the standard range for TC (1.94 for slightly cold houses and 1.93 for cold houses), LDL-C (1.63 for slightly cold houses and 1.67 for cold houses), and non-HDL-C (1.73 for cold houses) were significant, indicating the presence of associations of cold indoor environments with TC, LDL-C, and non-HDL-C even after excluding the effects of medical treatment for cholesterol.
| Objective variable | Explanatory variable | Unadjusted | Adjusted＊ | ||||
|---|---|---|---|---|---|---|---|
| Odds ratio | (95%CI) | p value | Odds ratio | (95%CI) | p value | ||
| TC ≥ 220 mg/dL | Warm (≥ 18℃) | Ref. | Ref. | ||||
| Slightly cold (12−18℃) | 1.83 | (1.20, 2.81) | 0.005 | 1.94 | (1.25, 3.00) | 0.003 | |
| Cold (＜12℃) | 1.84 | (1.19, 2.85) | 0.006 | 1.93 | (1.24, 3.01) | 0.004 | |
| LDL-C ≥ 140 mg/dL | Warm (≥ 18℃) | Ref. | Ref. | ||||
| Slightly cold (12−18℃) | 1.60 | (1.02, 2.50) | 0.041 | 1.63 | (1.03, 2.58) | 0.036 | |
| Cold (＜12℃) | 1.62 | (1.03, 2.54) | 0.039 | 1.67 | (1.05, 2.66) | 0.030 | |
| HDL-C ＜40 mg/dL | Warm (≥ 18℃) | Ref. | Ref. | ||||
| Slightly cold (12−18℃) | 0.87 | (0.41, 1.85) | 0.711 | 0.81 | (0.38, 1.71) | 0.580 | |
| Cold (＜12℃) | 0.63 | (0.28, 1.39) | 0.250 | 0.60 | (0.27, 1.32) | 0.201 | |
| TG ≥ 150 mg/dL | Warm (≥ 18℃) | Ref. | Ref. | ||||
| Slightly cold (12−18℃) | 1.34 | (0.78, 2.30) | 0.293 | 1.28 | (0.73, 2.24) | 0.390 | |
| Cold (＜12℃) | 1.15 | (0.66, 1.99) | 0.627 | 1.15 | (0.65, 2.03) | 0.638 | |
| Non-HDL-C ≥ 170 mg/dL | Warm (≥ 18℃) | Ref. | Ref. | ||||
| Slightly cold (12−18℃) | 1.68 | (1.00, 2.82) | 0.051 | 1.70 | (1.00, 2.89) | 0.052 | |
| Cold (＜12℃) | 1.69 | (1.00, 2.86) | 0.052 | 1.73 | (1.01, 2.98) | 0.048 | |
＊Adjusted for age, sex, body mass index, household income, salt check sheet score, vegetable consumption, exercise, smoking, alcohol consumption, antihypertensive drug use and season in which the health checkup was conducted.
TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride.
This cross-sectional analysis of 2004 participants in 1333 households showed that 1) the average bedroom temperature was 12.7℃, which is below the WHO recommendation of 18℃; 2) TC, LDL-C and non-HDL-C gradually increased with colder houses; 3) compared to warm houses, the adjusted odds ratio for TC exceeding 220 mg/dL was 1.83 and 1.87, that for LDL-C exceeding 140 mg/dL was 1.49 and 1.64, and that for non-HDL-C exceeding 170 mg/dL was 1.67 and 1.77, for those living in slightly cold houses and cold houses, respectively; and 4) participants without hyperlipidemia showed the same trends as those in the main analysis, indicating the robustness of the results. A potential mechanism that could explain our results is that cold stress activates the hypothalamus-pituitary-adrenal (HPA) axis, leading to secretion of cortisol, which is generated from cholesterol. Thus, persistent cold exposure and excess cortisol might result in high serum cholesterol levels. Previous studies have reported the mechanism of this process43-45), the association between HPA activity/cortisol and serum cholesterol43, 46), and identified cold as an important physical stressor43, 44).
Several studies have investigated the association between serum cholesterol levels and outdoor temperature. Hong et al.25) showed that LDL-C increased and HDL-C decreased with decreasing outdoor temperature. Sartini et al.27) also demonstrated an increase in TC and LDL-C with decreasing outdoor temperature. In contrast, Halonen et al.24) and Basu et al.26) showed the opposite association between blood lipids and outdoor temperature. One possible cause of these discrepant results is the indoor temperature, suggesting the need to examine its association with blood lipids. One study that examined the relationship between indoor temperature at home and blood lipids was conducted by Shiue et al.47) The researchers analyzed the relationship between indoor temperature and biomarkers during nurses’ interviews among 7997 participants and showed that those living in houses with indoor temperatures below 18℃ had high TC and LDL-C but not HDL-C and TG, which is consistent with the present results. However, because this relationship was examined at just one point in time – during the nurses’ interviews – it is impossible to extrapolate the relationship between cholesterol and indoor temperature, to which the participants are exposed on a daily basis. The findings in the present paper suggest that the relationship between serum cholesterol level and indoor thermal environment at home may not be temporary.
We propose that improving the home thermal environment is an effective way to reduce serum cholesterol levels and future CVD risk. Many clinical trials have examined the relationship between serum cholesterol levels and CVD48-51), and a guideline has been issued on how to reduce cholesterol levels, with improvements to lifestyle habits being one recommended measure52, 53). However, because improving lifestyle habits depends on individual effort, interventions on lifestyle habits have not been sufficiently effective in long-term studies54) or community studies55), thereby highlighting the inherent challenge of promoting lifestyle changes. It may be more effective to improve both lifestyle habits and living environment. However, at present, the guideline mentioned above does not address improvements to the living environment, such as protection from the cold in indoor environments like the home. We expect that evidence on the relationship between living environment and cholesterol levels will continue to accumulate, which we hope will prompt the promotion of interventions related to environmental factors.
Progression of arteriosclerosis, which is a known risk factor of CVD, occurs as vascular endothelial cells become injured due to high blood pressure, followed by the accumulation of cholesterol on these injuries. When cholesterol accumulates and blood vessels become constricted, a vicious cycle arises leading to further increases in blood pressure56, 57). The present analysis identified a significant relationship between high cholesterol levels and residents living in houses with low room temperatures. Similarly, low room temperature environments increase blood pressure30). As such, compared to residents living in warm houses, those living in cold houses likely experience a vicious cycle of increasing blood pressure, accumulating cholesterol, and arteriosclerosis progression.
A major strength of the present study was that we used objective data from blood samples and 2-week indoor temperature measurements in our analyses, which might have reduced subjective biases. Nevertheless, the study also had several limitations. First, a degree of selection bias was likely present, given that we recruited residents who intended to conduct insulation retrofitting. To examine the possibility of this bias, we compared serum cholesterol indices in the present survey with those in the National Health and Nutrition Survey by sex and age in Supplemental Fig.3. The two surveys showed the same patterns (e.g., TC and LDL-C in women were lower than those in men below the age of 50 years old, but rose rapidly and exceeded those in men after age 50 years), indicating that participants in the present study had comparable characteristics to those of the general Japanese population. However, we conducted the survey only on the Japanese population. Thus, future studies should examine the external validity and applicability of our findings to non-Japanese populations. Second, the timing of the health checkup varied from person to person. However, we adjusted for the season in which the health checkup was conducted in the logistic model to account for seasonal variations in CVD biomarkers. Third, we could not check whether or not blood samples were actually collected after at least a 10-hour fast, which could affect the TC level estimated from the Friedewald equation. However, as explained in the Methods section, health checkups in Japan are generally conducted after a fasting time of at least 10 hours. Additionally, we excluded TG data ≥ 400 mg/dL. Based on these two reasons, we believe the serum cholesterol data were unbiased. Finally, we cannot identify a clear biological mechanism to explain the discrepancy in several serum cholesterol indices. Thus, experimental studies examining more detailed physiological parameters are required to confirm our findings. The ultimate goal of this research is to clarify the relationship between arteriosclerosis progression and housing by comparing a cold housing group with a warm housing group in long-term follow-up studies.
Non-HDL-C was calculated by Non-HDL-C (mg/dL)=TC (mg/dL) − HDL-C (mg/dL).
Men and women in 20s were excluded because the sample size of the SWH survey was too small (men: n=8, women; n=13).
The present study showed that residents in cold houses had high serum cholesterol levels. A clinical implication of our study is that a warmer indoor thermal environment may contribute to maintaining lower serum cholesterol levels. Thus, besides lifestyle modification, improving the indoor environment through strategies such as installing high thermal insulation and appropriate use of heating devices may lead to better cardiovascular health.
We gratefully acknowledge the numerous construction companies, study investigators, and research committee members throughout all 47 prefectures in Japan who participated in the SWH survey. Members of the research committee for promotion of SWH who participated in this study are listed in Supplemental Table 1.
| (a) Members of the Research Committee for the Promotion of Smart Wellness Housing | (b) Members of the Subcommittee for Analysis of the Smart Wellness Housing Survey | ||
|---|---|---|---|
| Chairperson | Chairperson | ||
| Shuzo MURAKAMI＊ | Institute for Built Environment and Carbon Neutral for SDGs | Toshiharu IKAGA＊ | Keio University |
| Vice-chairperson | Vice-chairperson | ||
| Takesumi YOSHIMURA＊ | University of Occupational and Environmental Health | Yoshihisa FUJINO＊ | University of Occupational and Environmental Health |
| Hiroshi YOSHINO＊ | Tohoku University | Organizer | |
| Kazuomi KARIO＊ | Jichi Medical University | Shintaro ANDO＊ | University of Kitakyushu |
| Organizer | Tatsuhiko KUBO | Hiroshima University | |
| Toshiharu IKAGA＊ | Keio University | Committee member | |
| Committee member in medicine | Wataru UMISHIO | Tokyo Institute of Technology | |
| Suminori AKIBA | Kagoshima University | Yuko OGUMA | Keio University |
| Mikio ARITA | Sumiya Rehabilitation Hospital | Naoki KAGI | Tokyo Institute of Technology |
| Michiya IGASE | Ehime University | Hiroshi KANEGAE | Genki Plaza Medical Center for Health Care |
| Masayoshi ICHIBA | Saga University | Shun KAWAKUBO | Hosei University |
| Nami IMAI | Mie University | Yoshinobu SAITO | Kanagawa University of Human Services |
| Masaki UEMURA | At Home, LLC | Keigo SAEKI | Nara Medical University |
| Hiroyuki UEHARA | National Assembly Promoting Healthy and Energy Conserving Housing | Masaru SUZUKI | Tokyo Dental College Ichikawa General Hospital |
| Haruo UGUISU | Tokushima Bunri University | Tsuyoshi SEIKE＊ | Tokyo University |
| Kensuke ESATO | Yamaguchi University | Takayuki TAJIMA | Tokyo Metropolitan University |
| Akira EBOSHIDA | Hiroshima University | Experts committee member | |
| Yuko OGUMA | Keio University | Maki ITO | Japan Federation of Housing Organizations |
| Toshiyuki OJIMA | Hamamatsu University School of Medicine | Hiroshi KOJIMA | Keio University |
| Shimato ONO | Marugame Ono Clinic | Natsue DOIHARA | Keio University |
| Yoshio OMATA | Hoju, Co., Ltd. | Adviser | |
| Takahiko KATOH | Kumamoto University | Takesumi YOSHIMURA＊ | University of Occupational and Environmental Health |
| Masahiko KATO | Tottori University | Kazuomi KARIO＊ | Jichi Medical University |
| Shinya KUNO | University of Tsukuba | Tanji HOSHI＊ | Tokyo Metropolitan University |
| Kiyokage KUBO | Kubo Clinic | ＊: members of the Research Planning Committee for the Promotion of Smart Wellness Housing | |
| Yoshiki KURODA | University of Miyazaki | ||
| Yasuaki SAIJO | Asahikawa Medical University | ||
| Kazuhiro SATO | University of Fukui | ||
| Eiji SHIBATA | Yokkaichi Nursing and Medical Care University | ||
| Kuninori SHIWAKU | Shimane University | ||
| Narufumi SUGANUMA | Kochi University | ||
| Tomotaka SOBUE | Osaka University | ||
| Toshiro TAKEZAKI | Kagoshima University | ||
| Masatoshi TANAKA | Fukushima Medical University | ||
| Tsuyoshi TANABE | Yamaguchi University | ||
| Susumu TSUKAMOTO | Saitama Jikei Hospital | ||
| Hiroyuki DOI | Okayama University | ||
| Kunio DOBASHI | Jobu Hospital for Respiratory Diseases | ||
| Chisato NAGATA | Gifu University | ||
| Hiroyuki NAKAMURA | Kanazawa University | ||
| Kunio NAKAYAMA | Former Osaka University | ||
| Norihiro NOGATA | Saiseikai Karatsu Hospital | ||
| Takashi HANATO | Eigenji Clinic | ||
| Yoshihisa FUJINO＊ | University of Occupational and Environmental Health | ||
| Tanji HOSHI＊ | Tokyo Metropolitan University | ||
| Satoshi HOSHIDE | Jichi Medical University | ||
| Takahiro MAEDA | Nagasaki University | ||
| Muneo MINOSHIMA | Minoshima Clinic | ||
| Takashi MURAWAKA | Yumemokuba, SNPC | ||
| Hidekazu YAMADA | Kindai University Nara Hospital | ||
| Misako YOSHINAGA | Kusunoki Hospital | ||
| Committee member in architecture | |||
| Akihiko IWASA | Hosei University | ||
| Atsushi IWAMAE＊ | Kindai University | ||
| Akihito OZAKI | Kyushu University | ||
| Satoru KUNO | Nagoya University | ||
| Minoru KUMANO | Miyazaki University | ||
| Shoichi KOJIMA | Saga University | ||
| Yasuyuki SHIRAISHI | University of Kitakyushu | ||
| Hirotaka SUZUKI | Hokkaido Research Organization | ||
| Tsuyoshi SEIKE＊ | Tokyo University | ||
| Naoki TAKAGI | Shinshu University | ||
| Masaki TAJIMA | Kochi University of Technology | ||
| Yoshito TANAKA | Nagasaki Institute of Applied Science | ||
| Takayuki TAMAI | National Institute of Technology, Yonago College | ||
| Mitsutaka TSUJI | Gifu Academy of Forest Science and Culture | ||
| Reiji TOMIKU | Oita University | ||
| Hisaya NAGAI | Mie University | ||
| Daisaku NISHINA | Hiroshima University | ||
| Hideyo NIMIYA | Kagoshima University | ||
| Kenichi HASEGAWA | Akita Prefectural University | ||
| Hirofumi HAYAMA＊ | Hokkaido University | ||
| Akira FUKUSHIMA | Former Hokkaido University of Science | ||
| Yuji HORI | University of Toyama | ||
| Takeo MATSUOKA | Asia University | ||
| Teruaki MITAMURA | Maebashi Institute of Technology | ||
| Shinji YOSHIDA | Nara Women’s University | ||
| ＊: members of the Research Planning Committee for the Promotion of Smart Wellness Housing | |||
This study was partly supported by the Ministry of Land, Infrastructure, Transport and Tourism as part of the Model Project for Promotion of SWH and a JSPS KAKENHI [Grant Numbers JP17H06151: Principal Investigator: Prof. Toshiharu Ikaga]. Funding organizations had no role in deciding the study design and conducting the study; collection, management, analysis, and interpretation of the data; preparation of the article; or the decision to submit the article for publication.
T Ikaga has received research grants from Tokyo Gas Co., Ltd., Osaka Gas Co., Ltd., Panasonic Homes Co. Ltd., Fuyo Home Co. Ltd., Asahi Kasei Homes Corp., LIXIL Corp., Azbil Corp., Kajima Corp., Shimizu Corp., Nice Corp., Japan Gas Association and Japan Sustainable Building Consortium. The above grants/funds/honorarium have been received outside the submitted work.
