Most countries suffer from 5 to 30% excess winter mortality (EWM), the majority of additional winter deaths being caused by cerebrovascular diseases, which may be related, among other factors, to the pressor effect of cold weather. The possibility to control the blood pressure (BP) changes associated with low environmental temperature is thus a relevant issue, for both physicians and public health officers. In this regard, previous investigations, as well as recommendations given to citizens, have been developing into two different directions, underlying the importance of indoor or outdoor temperature, respectively, although the health impact of cold strain seems to be a more complex issue.
The diffusion of effective systems for heating homes has always been considered as an important element in the promotion of public health, especially in world regions characterized by cold winters. To reduce the environmental impact of ambient indoor heating and at the same time to optimize its effects, there is now increasing attention towards technologies that can improve heating systems efficiency, by reducing heat loss from homes walls, doors and windows.
These efforts seem to have been rewarding because, using a 5-year moving average, which smoothens out short-term fluctuations, excess winter deaths have been declining steadily since 1960–1961 up to current times. However, the importance of thermal efficiency standards in relation to housing is not equally felt as an important issue in the different European countries. This may be partly due to the observation that, in Europe, although mortality does increase as weather gets colder, differences in outdoor temperature only explain a small amount of the variance in winter mortality, and high levels of EWM can occur during relatively mild winters . More precisely, Healey  showed that EWM varied widely within Europe, and that countries with very low outdoor winter temperatures in Scandinavia and Northern Europe, such as Finland and Germany, somehow unexpectedly had very low rates of EWM, whereas countries with very mild winter temperatures in Southern Europe, such as Portugal and Spain, displayed very high rates of EWM.
In fact, European countries with milder winters also tend to have homes with poorer thermal efficiency (e.g. fewer homes have cavity wall insulation and double glazing), which makes it harder to keep homes constantly warm during winter . Available data on cross-country thermal efficiency standards in housing indicate that those countries with the poorest thermal housing efficiency (Portugal, Greece, Ireland, the UK) do indeed demonstrate the highest excess winter mortality .
The same epidemiological data can, however, be read in a different perspective. The behavioural capability to cope with cold weather was also shown to display quite wide variations within Europe. The Eurowinter group  reported that, compared with people living in countries with cold winters, individuals from warmer countries were less likely to wear warm protective clothing in cold weather. The same international survey showed an independent association of outdoor, as well as indoor, low ambient temperatures with excess mortality in such countries during cold weather .
People in the retirement age are particularly vulnerable to winter temperatures, and show higher mortality rates with cold weather. Indeed, in Great Britain, blue collars at working age (50–59 years) had lower cold weather-related mortality as compared either with their wives of similar age or with men of the same social class after retirement age (65–74 years). These observations suggest that body internal heat production from manual work protected men of working age against daytime cold-related stress, as well as against the associated risk of higher mortality.
Moreover, elderly people living in sheltered houses that were fully heated, but who often went outdoors, had as much winter mortality as the general elderly population including those living in less well heated houses . This is not surprising, since the majority of additional winter deaths are caused by cerebrovascular diseases, and, for example, cold temperature-related stress affecting people waiting at a bus stop, exposed to cold winds in the open air, can exceed the effect of any other kind of stress experienced indoors. The superficial impression that high winter mortality rates, and winter peaks in healthcare management problems, might be prevented simply by warmer housing policies, might thus be highly misleading.
The issue is not only of theoretical interest, but also of practical relevance, because it may influence the kind of public advice given to the general population. Again, two different lines of public advice have been implemented in relation to cold weather exposure, with some charities giving advice on avoidance of cold temperature-related stress outdoors, whereas other recommendations have focused on the importance of preventing indoor low temperatures. According to the former approach, apart from personal measures such as warm clothing and exercise performance when outdoor in cold weather, heating of waiting areas for public transport, and construction of windproof shelters on bus routes subjected to unscheduled delays, have been identified as obvious protective measures that might help in reducing low temperature-related health problems. According to the latter approach, on the contrary, excess of winter deaths is the deplorable result of too cold homes, the only sustainable solution to this problem being represented by public investment to increase the energy efficiency of housing, with the declared goal of making cold homes only a memory of past times.
The study by Saeki et al., published in the current issue of the Journal of Hypertension, was aimed at exploring the relationship between indoor temperature and ambulatory BP levels during winter, and to assess whether indoor temperature indeed shows stronger association than outdoor temperature with ambulatory BP indices, after correcting for physical activity. More specifically, the authors compared the associations of ambient temperatures, by separately considering indoor temperatures when individuals were at home, and outdoor temperatures when individuals were out of their houses, with ambulatory BP indices, including daytime BP, night-time BP, nocturnal BP fall (%), and morning BP surge. Eight hundred and eighty home-dwelling men and women older than 60 years were recruited from September 2010 to March 2013 for the Housing Environments and Health Investigation among Japanese Older People in Nara, Kansai Region (HEIJO-KYO) prospective community-based cohort study. Ambulatory BP was measured using a validated device on the nondominant arm at 30-min intervals. Indoor temperature was measured at 10-min intervals in the living room and bedroom 60 cm above the floor. Outdoor temperatures were also measured at 10-min intervals and were provided by the local meteorological office in Nara. Indoor and outdoor temperatures were related to each other, but their correlation became weaker with decreasing outdoor temperature. Whereas outdoor temperature was not significantly associated with ambulatory BP, a 1°C reduction of indoor temperature was significantly associated with a 0.22 mmHg increase in daytime SBP, a 0.18% increase in nocturnal BP fall, and with 0.34 mmHg increase in sleep-trough morning BP surge. These relationships survived after accounting for potential confounders, such as physical activity. Night-time SBP did not show any significant correlation with indoor and outdoor temperatures, but it was correlated with bed temperature. The authors’ conclusion  is that, as it might have been expected, indoor temperature is more closely associated than outdoor temperature with ambulatory BP in colder months. This study is potentially relevant because ambulatory BP is known to be a better predictor of cardiovascular disease mortality compared with conventional BP levels in the clinic [6,7]. The second important strength of the study by Saeki et al. is represented by the attention paid to consider the confounding role of physical activity. Notwithstanding these premises, however, this study does not add other important new elements to clarify the respective importance of indoor and outdoor temperatures in relation to BP changes, because it does not provide us with an estimation of the time spent outdoor by recruited participants. Without this important piece of information, the relative weight of exposure to outdoor temperature in relation to health-related outcomes cannot be precisely estimated. Secondly, the study by Saeki et al. does not help in solving another practical issue, which has relevance for both events prevention and advice to be provided to the public, that is, which indoor temperature should be considered. More precisely, it is not yet clear whether ambient temperatures measured always in the same standard place of the house are sufficient (as suggested by Saeki et al.) to characterize an individual's reactivity to weather changes or whether, on the contrary, we would need to measure the ambient temperature at the personal level (PET), as suggested by others . Focus on PET might indeed reconcile the different views of those emphasizing the importance of outdoor or of indoor temperatures, respectively . Recent technical progress has made low-cost wearable thermistors available, and this calls for their routine use in assessing weather-related health problems. Saeki et al. suggest that standard instruments placed always in the same place of the house do record different ambient temperature values for individuals living indoor as compared to PET devices. This observation, however, does not offer information precise enough to help in deciding which indoor temperature measuring device should be used, because measurements are taken in different locations of the house. We acknowledge that collinearity may exist between PET and conventional indoor temperature measurements. Inability to account for this collinearity might lead to erroneously attribution of the health impact of one temperature measure approach to the other. The issue of which indoor ambient temperature best reflect the impact of weather on health is thus yet unresolved, given these methodological problems, and additional studies implementing a more rigorous methodology are needed . Indeed, Fig. 1, in which we have put together some of the data provided by Saeki et al., shows that even in their study, the closer the measured temperature to the individual, the greater is its relationship with SBP, which strengthens the importance of having environmental temperature measured at the personal level.
In conclusion, assessment of indoor temperatures seems to better reflect the risk of weather-related health problems than assessment of outdoor temperatures. However, important questions remain to be answered, when translating these results into daily practice. Indeed, to place a thermistor in the bedroom is quite simple, but the value of this approach depends on how much time a given individual spends in his/her bedroom, and, in addition, on how many blankets are placed on the bed. Measuring ambient temperatures can be easy. Assessing the health impact of low ambient temperatures can be a much more difficult task.
Conflicts of interest
There are no conflicts of interest.
1. Brown G, Fearn V, Wells C. Exploratory analysis of seasonal mortality in England and Wales, 1998 to 2007. Health Stat Q
2. Healey JD. Excess winter mortality in Europe: a cross country analysis identifying key risk factors. J Epidemiol Commun Health
3. The Eurowinter GroupCold exposure and inter mortality from ischaemic heart disease, cerebrovascular disease, respiratory disease, and all causes in warm and cold regions of Europe. Lancet
4. Keatinge WR. Seasonal mortality among people with unrestricted home heating. BMJ
5. Saeki K, Obayashi K, Iwamoto J, Tone N, Okamoto N, Tomioka K, Kurumatani N. Stronger association of indoor temperature than outdoor temperature with blood pressure in colder months. J Hypertens
6. Staessen JA, Thijs L, Fagard R, O’Brien ET, Clement D, Leeuw PD, et al. Predicting cardiovascular risk using conventional vs. ambulatory blood pressure in older patients with systolic hypertension. JAMA
7. Hansen TW, Kikuya M, Thijs L, Bjorklund-Bodegard K, Kuznetsova T, Ohkubo T, et al. Prognostic superiority of daytime ambulatory over conventional blood pressure in four populations: a meta-analysis of 7030 individuals. J Hypertens
8. Brook RD, Weder AB, Rajagopalan S. Environmental hypertensionology the effects of environmental factors on blood pressure in clinical practice and research. J Clin Hypertens
9. Modesti PA. Season, temperature and blood pressure: a complex interaction. Eur J Intern Med
10. Modesti PA, Morabito M, Massetti L, Rapi S, Orlandini G, Mancia G, et al. Seasonal blood pressure changes: an independent relationship with temperature and daylight hours. Hypertension