Introduction
This report assesses the effects of indoor cold temperatures on health to support the development of the World Health Organization’s (WHO) Housing and health guidelines. The aim of this systematic review is to provide the best available evidence from existing research to contribute to the deliberations of the Guideline Development Group (GDG). During the preparation of the systematic review for this topic, an existing up-to-date review published in October 2014 was identified, Public Health England’s (PHE) “Minimum home temperature thresholds in winter – A systematic review” (Wookey 2014). A search in 2015 found only one eligible study published after the PHE review search date.
Therefore, this report provides a critical appraisal of the PHE review, with additional evidence from the single eligible study published after February 2014, when the searches for the PHE review were done, and additional evidence from a 2018 update search.
After a short background section, this report provides information on population, exposure, comparator, outcomes (PECO), and eligibility criteria for the systematic review, which were agreed with the WHO. This is followed by summary information on the search strategy and methods that were used to identify and summarize potentially relevant studies published after February 2014; and a discussion of the findings of the PHE review and subsequent eligible studies.
This report should be read in conjunction with the PHE review. PHE review results have been screened for relevance to PECO outcomes, and are provided as evidence profiles and a summary of findings table.
Background
The adverse effects of cold temperature on human physiology have been well-documented (Mercer 2003; Nahya 2002; Wilson 2001). In 1987, the WHO guidelines on indoor temperatures recommended indoor temperatures be maintained at 18°C, or 20–21°C in rooms used by the elderly (WHO 1987), but those guidelines were based on older recommendations, the evidential basis for which can no longer be traced.
This review topic was set with the aim of determining whether the 18°C threshold is supported by current research evidence. Additional information clarified that 18°C was not a predetermined threshold, but is intended to provide a starting point for determining what ideal indoor temperatures thresholds might be, for the general population worldwide, and across different vulnerable groups.
The challenge in weighing evidence for indoor temperature thresholds lies in selecting appropriate exposure measures. Multiple studies have been published showing the relationship between outdoor temperatures and health outcomes, but there are few studies of the effects of indoor temperatures. There are multiple other environmental exposures correlated with indoor and outdoor temperature, such as wind, precipitation, relative humidity, and levels of clothing; as well as possible differences in behaviour and physical activity levels while indoors or outdoors. Hence, the effects of outdoor temperature exposure may not adequately represent the effects of indoor temperature.
Similarly, the effects of specific temperature exposures in laboratory conditions may not well reflect the effects of indoor temperature exposure when engaged in day-to-day living. Consideration of an indoor cold temperature threshold should therefore weigh evidence according to the setting in which the evidence has been collected.
Eligibility criteria and PECO
The finalized research question for this review is:
Do residents living in housing where indoor temperatures are below 18°C have worse health outcomes than those living in housing with indoor temperatures above 18°C?
shows the inclusion and exclusion criteria that were used.
Table 1Inclusion and exclusion criteria for the review
View in own window
| Inclusion criteria | Exclusion criteria |
|---|
| Context | Domestic houses or flats in the community setting | |
|---|
| Participants | People of all age groups, with a particular emphasis on vulnerable sub-populations:
Infants <12 months Children <5 years Adults over 65 years
| |
|---|
| Intervention | Indoor temperature below 18°C | |
|---|
| Comparison | Indoor temperature above 18°C | |
|---|
| Outcomes | Health related outcomes (as ranked by the GDG):
Respiratory morbidity and mortality All cause-mortality in infants Hospital admissions Cardiovascular morbidity and mortality Depression
| |
|---|
Search strategies and checking of articles
The initial search for this review was conducted using PubMed with the following search strategy:
(Cold OR Cool OR Chill OR Low OR Minimum OR Severe OR Hibernal OR Hiemal) AND (Indoor OR Home OR Domestic OR Dwelling OR House OR Inside OR Household OR Outdoor OR Ambient OR Outside OR Exterior OR Threshold) AND (temperature OR climate OR thermal OR degrees OR weather) AND (“infant mortality” OR cardiovascular OR respiratory OR hospitalization OR admissions OR depression OR mood), limited to humans.
The final date on which this search was conducted was 1 January 2015, yielding 1884 results. 102 records were selected for abstract screening. After screening 44 of these abstracts, nine had been identified for full paper review. At this point, the PHE review was identified. The PHE review had considered all nine of those papers, and also resolved a set of questions about how best to address and integrate different possible additional search terms. Therefore, of the remaining 58 abstracts yet to be screened, only the 14 published in 2014 were screened. Of these, two papers met inclusion criteria and were not already included in the PHE review (Saeki 2014a; Saeki 2014b). These two papers had the same authors and published results from different aspects of the same study.
A second search was performed using the PHE search terms on 31 March 2015, including only papers published since 1 February 2014. That search identified no further eligible papers.
In order to bring the systematic review up-to-date, new searches for eligible studies were done on 28 March 2018 to identify articles published since 1 January 2015. We used the original search strategies to re-run the search in PubMed. The retrieved records were checked by two authors (RL and MC) and the full text was sought for all studies judged to be potentially eligible. When obtained, the full text of each of these articles was checked by two authors (RL and MC). shows the flow of articles through the original search and updating process.
Extraction of information, preparation of narrative summaries, evidence profiles and summary of findings tables
Evidence presented in the PHE review (Wookey 2014) was screened for relevance to the outcomes identified for this PECO. Of the 20 papers covered in the PHE review, seven are included below.
Study information extracted from the PHE review included:
Location and date of study
Type and number of participants
Details of the intervention and any comparator
Design, including the methods used for any comparison
Results for all relevant outcomes reported.
PHE review assessments of study limitations were extracted to ascertain:
The same information was extracted by two authors (RL and MC) from the four papers identified in the update searches for this review, as described in the previous section. The accuracy of the extracted data and the risk of bias assessments were confirmed through discussion.
This information was used to complete an evidence profile for each study. These are shown in Appendix 1. Summary of Findings tables were also prepared, which describe the evidence in narrative terms, reflecting the types of study that were identified (Appendix 2), and the characteristics of the included studies are presented in Appendix 3. The results of the risk of bias assessment of studies included during the update are shown in Appendix 4.
Findings
The findings of the relevant papers identified in the PHE review and papers identified through subsequent searches are summarized below.
Respiratory morbidity and mortality
Of the four studies investigating the effects of indoor cold on respiratory health, three found that colder indoor temperatures increased respiratory morbidity. One cross-sectional study in adults with COPD found better health status with more hours of indoor temperature at and above 21°C. A dose-response trend was observed for number of days with bedroom temperatures of 18°C and above for at least nine hours. The greatest effects were observed in adults who smoked compared with non-smokers (Osman 2008). Similarly, modelling based on the results of a randomized trial involving children with asthma found that every 1°C increase in room temperature below the threshold of 9°C, was associated with a small but significant increase in lung function. Bedroom exposure was shown to have stronger association with asthmatic children’s lung function than living room exposure (Pierse 2013).
In addition, one cohort study from China, including adults with COPD, reported reduced respiratory problems with an indoor temperature at 18.2°C regardless of whether indoor humidity was low, moderate or high (Mu 2017). In contrast, a case-control study in children with and without upper respiratory tract infections showed no consistent associations with indoor temperature (Ross 1990).
The certainty of the evidence that warming a cold house (perhaps to a minimum indoor temperature of 18°C) would reduce the risk of respiratory mortality and morbidity was assessed as moderate.
Cardiovascular morbidity and mortality: blood pressure
Of the six included studies that assessed the association between indoor temperature and blood pressure, all showed that lower temperatures were associated with higher blood pressure, including two randomized trials in Japan that found higher blood pressure in people living in colder homes (Saeki 2013, Saeki 2015).
A cohort study in Japan of adults over 60 years of age found that decreases of 1°C in indoor temperatures were significantly associated with increased blood pressure levels at different times of the day, even after controlling for potential confounders (Saeki 2014a, Saeki 2014b). There was a stronger association of indoor temperature than outdoor temperature with ambulatory blood pressure, which suggested that excess winter cardiovascular mortality could be prevented by improving the housing thermal environment (Saeki 2014b). Two cohort studies from Scotland found people in housing heated to less than 18°C had a greater risk of high blood pressure (Shiue 2014, Shiue 2016). This risk increased if temperatures were below 16°C (OR 4.92) (Shiue 2014). Similarly, a cohort study in the United Kingdom found a decrease in systolic and diastolic blood pressure of 0.5 mmHg per 1°C increase in room temperature (Bruce 1991).
The review also identified six studies of temperature and blood pressure that were done under laboratory conditions (Collins 1985, Inoue 1992, Leppäluoto 2001, Mercer 1999, Neild 1994, Wagner 1987). The studies showed a relationship between warming and lower blood pressure, but because evidence for the relationship between blood pressure and housing indoor temperature was indirect, the studies were not included in the evidence summary.
The certainty of the evidence that warming a cold house (to a minimum indoor temperature of 18°C) would reduce the risk of cardiovascular mortality and morbidity was assessed as moderate.
Discussion
The PHE review concluded that heating homes to at least 18°C in winter poses minimal risk to the health of a sedentary person, wearing suitable clothing. It found this threshold to be particularly important for people over 65 years of age, or with pre-existing medical conditions, but allowed that healthy people aged 1 to 64 years might wish to heat their homes to slightly less than 18°C if they were wearing appropriate clothing and were active (during the day); or with sufficient bedding, clothing and thermal blankets or heating aids as appropriate at night.
The PHE review was strong in its methodology and comprehensiveness. However, it had two main limitations as a source of evidence for the PECO of the present review. The first of these limitations was that it was prepared for an audience from the United Kingdom and as such, was conducted with a much narrower target audience in mind than the present review. Further, the PHE review was conducted with the aim of defining indoor temperature thresholds which would “[protect] health whilst reducing carbon emissions and avoiding unnecessary expenditure on fuel” (Wookey 2014, p. 4), whereas the present review was only interested in direct health effects. These differences might have influenced the assessment of the strength of the evidence base and the conclusions drawn from it.
The second potential limitation lay in the mismatch between the limited PECO outcomes and the broader scope of the PHE review. The PHE review included studies of the relationship between indoor temperature and body mass index or obesity; perceptions of thermal comfort; and also indirect studies of the effect of temperature on biomarkers under laboratory conditions. While the evidence for a relationship between higher indoor temperature and higher body mass index was mixed and unlikely to have affected the recommendations made by the PHE review, inclusion of broader and indirect measures of health provided greater weight to the threshold recommendation.
In terms of health outcomes of interest to this review, findings of four (Saeki 2015, Shiue 2014a, Shiue 2014b, Shiue 2016) of the five studies published subsequent to the PHE review were strongly consistent with the previous PHE review findings on the relationship between blood pressure and indoor temperature; they strengthened the evidence, but did not add to it. The other study (Mu 2017) provided evidence for a relationship between low indoor temperature and COPD risk, with the risk increased by high humidity. Therefore, the evidence shows the benefits for respiratory health and blood pressure associated with the warming of cold houses, but no clear temperature threshold for these benefits, and no studies of other health outcomes originally prioritized by the Guideline Development Group such as all-cause mortality, hospital admissions and depression.
Contributors
Lead: Lucy Telfar Barnard (University of Otago, Wellington and He Kainga Oranga/Housing and Health Research Programme, New Zealand).
Team: Philippa Howden-Chapman (University of Otago, Wellington and He Kainga Oranga/Housing and Health Research Programme, New Zealand), Mike Clarke (Queen’s University of Belfast, Northern Ireland and Evidence Aid, United Kingdom), Ramona Ludolph (Department of Public Health, Environmental and Social Determinants of Health, World Health Organization, Switzerland).
References
Bruce
N, Elford
J, Wannamethee
G, Shaper
AG. The contribution of environmental temperature and humidity to geographic variations in blood pressure. Journal of Hypertension. 1991;9(9):851–8. [
PubMed: 1663987]
Collins
KJ, Easton
JC, Belfield-Smith
H, Exton-Smith
AN, Pluck
RA. Effects of age on body temperature and blood pressure in cold environments. Clinical Science (London). 1985;69(4):465–70. [
PubMed: 4042547]
Inoue
Y, Nakao
M, Araki
T, Ueda
H. Thermoregulatory responses of young and older men to cold exposure. European Journal of Applied Physiology and Occupational Physiology. 1992;65(6):492–8. [
PubMed: 1483436]
Mercer
JB, Osterud
B, Tveita
T. The effect of short-term cold exposure on risk factors for cardiovascular disease. Thrombosis Research. 1999;95(2):93–104. [
PubMed: 10418798]
Mercer
JB. Cold - an underrated risk factor for health. Environmental Research
2003;92:8–13 [
PubMed: 12706750]
Mu
Z, Chen
P-L, Geng
F-H, Ren
L, Gu
W-C, Ma
J-Y, et al. Synergistic effects of temperature and humidity on the symptoms of COPD patients. International Journal of Biometeorology. 2017;61(11):1919–25. [
PubMed: 28567499]
Nahya
S. Cold and the risk of Cardiovascular Diseases: A review. International Journal of Circumpolar Health
2002;61(4):373–80. [
PubMed: 12546195]
Neild, PJ, Keatinge
WR, Donaldson
GC, Mattock
M, Caunce
M. Cold-induced increases in erythrocyte count, plasma cholesterol and plasma fibrinogen of elderly people without a comparable rise in protein C or factor X. Clinical Science
1994;86(1):43–8. [
PubMed: 8306550]
Osman
LM, Ayres
JG, Garden
C, Reglitz
K, Lyon
J, Douglas
JG. Home warmth and health status of COPD patients. European Journal of Public Health. 2008;18(4):399–405. [
PubMed: 18367496]
Pierse
N, Arnold
R, Keall
M, Howden-Chapman
P, Crane
J, Cunningham
M, et al. Modelling the effects of low indoor temperatures on the lung function of children with asthma. Journal of Epidemiology and Community Health. 2013;67(11):918–25. [
PubMed: 23940250]
Ross
A, Collins
M, Sanders
C. Upper respiratory tract infection in children, domestic temperatures, and humidity. Journal of Epidemiology and Community Health
1990;44(2): 142–6. [
PMC free article: PMC1060623] [
PubMed: 2370503]
Saeki
K, Obayashi
K, Iwamoto
J, Tanaka
Y, Tanaka
N, Takata
S, et al. Influence of room heating on ambulatory blood pressure in winter: a randomized controlled study. Journal of Epidemiology and Community Health. 2013:jech-2012-201883. [
PubMed: 23447647]
Saeki
K, Obayashi
K, Iwamoto
J, Tone
N, Okamoto
N, Tomioka
K, et al. Stronger association of indoor temperature than outdoor temperature with blood pressure in colder months. Journal of Hypertension. 2014;32(8):1582–9. [
PubMed: 24937638]
Saeki
K, Obayashi
K, Iwamoto
J, Tone
N, Okamoto
N, Tomioka
K, et al. The relationship between indoor, outdoor and ambient temperatures and morning BP surges from inter- seasonally repeated measurements. Journal of Human Hypertension. 2014;28(8):482–8. [
PubMed: 24553634]
Saeki
K, Obayashi
K, Kurumatani
N. Short-term effects of instruction in home heating on indoor temperature and blood pressure in elderly people: a randomized controlled trial. Journal of Hypertension. 2015;33(11):2338–43. [
PubMed: 26372318]
Shiue
I, Shiue
M. Indoor temperature below 18 degrees C accounts for 9% population attributable risk for high blood pressure in Scotland. Int J Cardiol
2014;171(1):e1–2. [
PubMed: 24315341]
Shiue
I. Cold homes are associated with poor biomarkers and less blood pressure check-up: English Longitudinal Study of Ageing, 2012–2013. Environmental Science and Pollution Research International. 2016;23(7):7055–9. [
PMC free article: PMC4820485] [
PubMed: 26873825]
Wagner
JA, Horvath
SM, Kitagawa
K, Bolduan
NW. Comparisons of blood and urinary responses to cold exposures in young and older men and women. Journal of Gerontology. 1987;42(2):173–9. [
PubMed: 3819343]
Wilson
JS, Smith
AF, Goodwin
J, Hawker
J, Taylor
R. A Systematic Review of the Biological and Health Effects of Exposure to Cold Temperature in Healthy and Diseased Adults Including the Elderly. In: Health
Do, ed.: Birmingham Health Technology Assessment Group, University of Birmingham, 2001:140.
Wookey
R, Bone
A, Carmichael
C, Crossley
A. Minimum home temperature thresholds for health in winter – A systematic literature review. London, UK: Public Health England, 2014.
World Health Organization. Health impact of low indoor temperatures. Copenhagen: World Health Organization; Regional Office for Europe, 1987.
Appendices
Appendix 3. Characteristics of included studies
Download PDF (328K)
Appendix 4. Risk of bias assessment of studies included during the update in 2018
View in own window
| Study | Did the study address a clearly focused issue? | Was the cohort recruited in an acceptable way? | Was the exposure accurately measured to minimize bias? | Was the outcome accurately measured to minimize bias? | Have the authors identified all important confounding factors? | Were confounders taken into account in the analysis? | Was the follow up of subjects complete enough? | Was the follow up of subjects long enough? | Are the results reliable? | Can the results be applied to the local population? | Was there a description of how study size was arrived at? | Was there an adequate description of the statistical analysis? | Is there an adequate description of the study participants? |
|---|
| Mu 2017 | + | + | unclear | + | − | + | + | + | unclear | unclear | + | + | + |
| Saeki 2015 | + | − | + | + | + | + | + | + | + | unclear | + | + | + |
| Shiue 2016 | + | + | + | + | − | − | N/A | N/A | unclear | unclear | + | + | − |
