- Email: [email protected]
Cold—an underrated risk factor for health
Environmental Research 92 (2003) 8–13
Cold—an underrated risk factor for health James B. Mercer Department of Medical Physiology, Faculty of Medicine, University of Troms^, Troms^ N–9037, Norway Received 4 February 2002
Abstract Cardiovascular diseases (CVD) are responsible for around 20% of all deaths worldwide (approximately 14 million) and are the principal cause of death in all developed countries, accounting for 50% of all deaths. Variations in the annual per capita death rates in different countries are well documented. Less well known are seasonal variations in death rates, with the highest levels occurring during the colder winter months, which have been described in many countries. This phenomenon is referred to as excess winter mortality. CVD-related deaths account for the majority of excess winter deaths (up to 70% in some countries), while about half of the remaining are due to increases in respiratory diseases. Paradoxically, CVD mortality increases to a greater extent with a given fall in temperature in regions with warm winters. While much of the indirect evidence points to the notion that cold is somehow involved in explaining excess winter deaths, the mechanism by which seemingly mild exposure to cold ambient conditions can increase the risk of death remains unclear. The strong indirect epidemiological evidence coupling cold climate to mortality may be related to indoor rather than outdoor climatic conditions (e.g., cold/damp houses versus warm/dry houses) coupled with a plethora of factors including health status, ageing-related deterioration in physiological and behavioral thermoregulation, toxicology, and socioeconomic factors. r 2002 Elsevier Science (USA). All rights reserved. Keywords: Cold climate; Winter mortality; CVD; Risk factor; Air temperature
1. Introduction One of the many environmental factors which affect human health is climate and the health risks associated with exposure to extreme environmental conditions are well documented. However, while living and working in extreme climatic conditions constitutes an obvious potential risk to health it is not always immediately clear that people living and working under milder climatic situations are also at risk. The purpose behind this paper is to briefly review some of the evidence, both direct and indirect, indicating that cold may be a more important risk factor for morbidity and mortality from disease than most people realize, even in countries with mild winter climates. Since cardiovascular diseases (CVD) are responsible for around 20% of all deaths worldwide (approximately 14 million) and are the principal cause of death in all developed countries, accounting for 50% of all deaths (42% in the United States), this view will mainly be concerned with these diseases of which the most common disease categories
Fax: +47-776-45440. E-mail address: [email protected].
are hypertension, ischemic heart disease, and cerebrovascular disease. In examining the problem of cold as a risk factor, epidemiological evidence showing seasonal variations in morbidity and mortality will be described. Epidemiological evidence showing that winter mortality is more pronounced in countries having milder winter climates will also be presented. Since the majority of registered deaths from CVD occur in the elderly, at least in the developed countries, a brief examination will be made of the question whether people in this age group during the course of their normal daily lives really do experience exposure to cold to a degree which may be detrimental for health.
2. Excess winter mortality— epidemiological studies In the so-called developed countries mortality due to CVD is the most common cause of death. Mortality from CVD is also emerging as a prominent public health problem in developing countries, ranking third (ca. 16% of all deaths), with particularly high impact on low
0013-9351/03/$ - see front matter r 2002 Elsevier Science (USA). All rights reserved. doi:10.1016/S0013-9351(02)00009-9
J.B. Mercer / Environmental Research 92 (2003) 8–13
socioeconomic groups, and CVD-related deaths have already become the first cause of mortality in developing countries such as Argentina, Chile, Cuba, Republic of Korea, Mauritius, Singapore, Sri Lanka, Trinidad and Tobago, and Uruguay (WHOSIS, 2001). While variations in the per capita death rates in different countries are well documented and familiar to most of us, seasonal variations in mortality and morbidity are generally less well known. In fact, seasonal variations in mortality with the highest levels occurring during the colder winter months have been described throughout Europe (Kunst et al., 1993; Eng and Mercer, 1998; Rose, 1966; Mackenbach et al., 1992; Keatinge and Donaldson, 1995), in the Unites States (Kloner et al., 1999; Lanska and Hoffmann, 1999), and in other parts of the world, including China (Cheng, 1993) and the southern hemisphere. Even in Japan, a country considered to have one of the world’s lowest per capita mortality rates from CVD, there is a very clear increase in winter mortality (Ornato et al., 1990). In addition to variations between countries, regional variations in mortality within countries have also been described with higher mortality rates in colder or more northern parts of a country (Momiyama, 1968; Smith, 1984; Gyllerup et al., 1991). The phenomenon of increased winter mortality is often referred to as excess winter mortality. When discussing excess winter mortality, one is mainly concerned with the older section of the population, at least in the developed countries, since the majority of deaths from CVD occur in the elderly. For example, in Norway, where CVD, including ischemic heart disease and cerebrovascular disease, accounts for about 46% of all causes of death, about 90% of all CVD-related deaths occur in people who are 65 years or older (Eng and Mercer, 1998). While there are associations between high CVD mortality in the elderly during the winter and high CVD mortality in the coldest parts of Europe and other countries, the actual reason for the excess number of CVD-related deaths in winter is not fully known. Much of the indirect evidence points to the notion that cold is somehow involved. However, the mechanism by which seemingly mild exposure to cold ambient conditions can increase the risk of death remains unclear. Not all excess winter deaths can be attributed to CVD and, while mortality from CVD seems to account for the majority of excess winter deaths (up to 70% in some countries), about half of the remaining may be accounted for by increases in respiratory disease (Eurowinter Group, 1997; Mackenbach et al., 1992). While this article focuses mainly on CVD-related deaths, the importance of cold-related respiratory mortality should not be underestimated. Indeed, there is some evidence that respiratory deaths resulting from influenza epidemics in winter may account for as much
9
as two-thirds of excess winter deaths (Tillett et al., 1983) It has also been reported that the temperature of the air breathed greatly affects respiratory infections (Tyrrell et al., 1989) and that cold housing affects respiratory mortality in winter (Keatinge and Donaldson, 1995). Furthermore, in a fashion similar to that of CVD mortality (see below), cold spells are followed, albeit with some delay, by an increase in respiratory mortality (Donaldson and Keatinge, 1997). One of the many plausible explanations for increased winter mortality lies in the fact that annual changes in temperature and acclimatization to a given set of environmental conditions may influence physiological responses to toxic chemicals (Gordon, 2001).
3. Winter mortality is greater in countries with milder winters—a paradox? In those countries in which seasonal fluctuations in mortality from CVD have been documented, the amplitude of the fluctuations show quite large variations, being much greater in some countries than in others. While at first this might seem surprising, levels of excess winter mortality are relatively lower in Scandinavian countries than in Great Britain (Curwen, 1990; Laake and Sverre, 1996). Likewise, even though the yearly per capita mortality rate is relatively low in Mediterranean countries, the seasonal changes in mortality are greater in many Mediterranean countries than in Scandinavia (Curwen, 1990). The data presented in Fig. 1, which compares seasonal changes in mortality from CVD averaged over a 10-year period in Norway, a country with a cold winter climate, and Ireland, a country with a mild winter climate, are used to emphasize this apparently paradoxical point. In Norway and Ireland, the mortality rate is, respectively, 28% and 45% higher in winter (December) than in summer (August) (Fig. 1, top). In some years this winter– summer difference has been as high as 60% (Eng and Mercer, 1998). The amplitude of the seasonal changes in mortality from CVD are similar irrespective of whether ones looks at the effect on all causes of CVD or at more specific causes of death. For example, in both Norway and Ireland, the amplitude of the seasonal changes in mortality for stroke and acute myocardial infarction are similar to the respective seasonal changes in all CVD-related deaths (Eng and Mercer, 2000).
4. Seasonal variations in incidences of acute myocardial infarction There are rather few seasonality-related morbidity studies based, for example, on hospital admissions.
10
J.B. Mercer / Environmental Research 92 (2003) 8–13
(AMI), at 1474 US Hospitals, clear seasonal variations were seen, with January being the month with the most reported cases (13,025 cases/30-day month) and July with the least (5978 cases/30-day month).
5. Mortality and air temperature Since mortality rates from CVD increases in the winter it can be expected that mortality will be inversely related to air temperature, and in nearly all countries so far examined a strong negative correlation between the outside air temperature and the mortality from CVD has been shown (Rose, 1966; Bull, 1973; Bull and Morton, 1978; McKee; 1990; Douglas et al., 1995; Laake and Sverre, 1996; Eng and Mercer, 1998, 2000). This point is demonstrated in Fig. 1 with data from Norway and Ireland. Fig. 1 (bottom), in which CVD mortality is related to air temperature, shows that mortality rate increases sharply when the outside air temperature falls below about 01C in Norway. However, in Ireland, which has a milder winter climate than Norway (Fig. 1, middle), mortality rate begins to rise sharply at higher outside air temperatures (ca. 10–151C) (Fig. 1, bottom). It has been suggested that this difference reflects, among other things, poorer housing standards (insulation, heating, etc.) and/or different thermoregulatory behavior patterns in the elderly in Ireland as compared to those in Norway (Eng and Mercer, 1998). However, there is some debate with regard to the relative importance of indoor cold stress versus outdoor cold stress in causing winter mortality. While, the present article stresses the importance of the former (see below) there is evidence that the latter also has a contributory role (Keatinge, 1986; Eurowinter Group, 1997). Fig. 1. (Top) Mean monthly CVD mortality ratio (ICD 9:390–459) in Norway and Ireland in the age group 60 years and older. (Middle) Mean monthly air temperatures in Norway (10 weather stations) and Ireland (five weather stations). (Bottom) Relation between mean monthly cardiovascular disease mortality ratio (ICE 9: 390–459) and mean monthly air temperature in Norway and Ireland in the age group 60 years and older. All data are based on monthly air temperature in Norway and Ireland in the age group 60 years and older. All data are based on monthly averages for the 10-year period 1985–1994; Norway, closed symbols; Ireland open symbols. Mortality ratio is defined as the average number of deaths per day in a given month divided by the average number of deaths per day for the whole year.
However, the few available studies clearly demonstrate clear seasonal variations (e.g., Spencer et al., 1998; Danet et al., 1999). In the study of Spencer et al., (1998), which was based on data from the National Registry of Myocardial Infarction in the United States using a cross-sectional observational database of patients admitted to hospitals with acute myocardial infraction
6. The lag effect In the cases of both myocardial infarction and stroke there appears to be a significant interval between a temperature change and the onset of the episode leading to death. In other words, a sudden large decrease in air temperature is not associated with an immediate change in mortality rate. Increases in mortality first become apparent several days or even a week later (Boyd, 1960; Donaldson and Keatinge, 1997; Kunst et al., 1993). This fact is potentially important in that if one could identify a link in the chain of events leading from temperature change to death one might be able to break it and avert the fatal outcome (Bull and Morton, 1975). However, studies to date have not established a clear chain of events leading from a change in external temperature to death. Nevertheless, it remains very likely that changes
J.B. Mercer / Environmental Research 92 (2003) 8–13
in ambient temperature cause changes in death rates, especially in the elderly.
7. Changes in risk factors with season and with cold exposure There are several factors considered indicators of risk for CVD which have been shown to have clear seasonal variations, including winter increases in plasma cholesterol, plasma fibrinogen, blood pressure, and red and white blood cell counts (Stout and Crawford, 1991; Woodhouse et al., 1994). In addition, there are several changes in blood composition which are known to occur during acute cold exposure. These include increased red cell count, increased plasma cholesterol, and increased plasma fibrinogen (Keatinge et al., 1984; Kannel et al., 1987; Kawahara et al., 1989; Qizibash et al., 1991; Stout and Crawford, 1991). In another cold exposure study, the density distribution of blood platelet subpopulations was shifted to an increase in less dense platelets that were more sensitive toward aggregation-inducing agents (Opper et al., 1995). In a cold exposure study in young adults, mean blood pressure, b-thromboglobulin, platelet factor 4, and plasma noradrenaline were increased. In addition, cold exposure-induced increases in plasma fibrinogen levels and thromboxane B2 have also been reported (Mercer et al., 1999). From the above findings it can be concluded that short-term exposure to cold initiates a mild inflammatory reaction and a tendency for an increased state of hypercoagulability. However, it is important to recognize that most of these data have been obtained in experiments involving the exposure of healthy young subjects to levels of cold stress which most of us, especially the elderly, will rarely experience during the course of our normal daily lives. This raises the issue of the actual levels of cold exposure experienced by people on a day-to-day basis. For example, it has been suggested that short-term exposure to even rather mild cold conditions may be a risk for the elderly (Houdas et al., 1992; Keatinge and Donaldson, 1995), while others (see below) feel that the effect of cold may be more subtle and may be more related to long-term exposure to mild levels of cold.
8. Indoor climatic conditions and excess winter mortality—inappropriately insulated houses and/or thermoregulatory behavior The question concerning how much cold exposure people are actually subjected to during the normal course of their daily lives was examined in a recent European Economic Community-supported study (Eurowinter Group, 1997) investigating cold exposure and winter mortality from ischemic heart disease,
11
cerebrovascular disease, respiratory disease, and all causes in warm and cold regions of Europe. The main findings of this study were that mortality increased to a greater extent with a given fall in temperature in regions with warm winters, in populations with cooler homes, and among people who wore fewer clothes and were less active outdoors. That people may become cold-exposed under thermal conditions which most of us consider mild is actually very plausible when one considers that humans, from a physiological point of view, are tropical animals (a naked adult human requires an air temperature of ca. 301C to feel thermally comfortable—the socalled lower critical temperature). When we consider that about 90% of all cardiovascular-related deaths occur in persons who are 65 years and older (Eng and Mercer, 1998) and who often spend a considerable portion of their elderly lives living indoors (in some cases as much as 100% of their time is spent indoors), it is clear that we cannot ignore the possibility that adverse climatic conditions within people’s homes may be involved in explaining excess winter mortality. However, while there is strong indirect evidence that housing standards and indoor climate may be important factors in explaining excess winter mortality (Eurowinter Group, 1997), the question of whether this is really true arises, especially in relatively wealthy countries such as Norway in which the general standards of housing are recognized as being very high. To answer these sorts of questions we need to have information on many different factors related to cold exposure during the course of our normal daily lives, including information on indoor climatic conditions and thermoregulatory behavior patterns, especially in the elderly section of the population. While people living indoors may be potentially at risk for cold exposure, particularly during the winter months, they need not necessarily be exposed to cold if they restrict heat loss by behavioral means, for example by wearing appropriate clothing and adequately heating their homes. However, it has to be recognized also that, while the quality of housing construction is important for maintaining a warm indoor climate, it is equally possible to maintain a warm indoor climate in a badly insulated house if one uses enough fuel just as it possible to have a cold indoor climate in the most modern and well-insulated house if one restricts the use of home heating. In this respect, knowledge of behavior patterns, especially those related to thermoregulatory behavior and home heating habits, is very important. Unfortunately, while many countries have clear regulations governing indoor climatic conditions in public buildings such as schools, hospitals, and the like, there is little or no regulation, simply recommendations, for private homes, and published data on indoor climatic conditions and thermoregulatory behavior patterns in private homes are scarce. Preliminary studies from Ireland and Norway confirm
12
J.B. Mercer / Environmental Research 92 (2003) 8–13
that the houses in Ireland are generally cooler and damper (higher relative humidity) (Middleton et al., 2000) than the considerably warmer and dryer Norwegian homes (B^kenes and Mercer, 2000). To form a more complete picture concerning how adverse indoor climatic conditions may affect people’s health, information concerning the medical status of the people involved is also needed. For example, if a normal healthy person feels cold, he or she puts on more clothes, heats the room, has a warm drink, or exercises to maintain normal body temperature. However, the situation is often different in the elderly and both the ability of people to detect that their body temperature is falling and the ability to prevent such a decrease deteriorates with age (Wagner and Horvath, 1985; Jennings et al., 1993; Young and Lee, 1995). The body’s ability to regulate body temperature may also be affected by a variety of diseases and by the use of commonly used pharmaceutical agents, such as those used for sleeping and depressive disorders and betablockers used in the treatment of cardiovascular disorders. In connection with normal ageing-related alterations in the way that people respond to their thermal environments, some preliminary results from a study of elderly subjects living in private homes in northern Norway indicate that even in the well-insulated Norwegian homes the elderly have peripheral (foot skin) temperatures which are lower during the winter months than the summer months, indicating that they are more cold-stressed during this time of the year (J.B. Mercer and B^kenes, unpublished observation).
9. Conclusions Clearly the reasons behind excess winter mortality are complex and involve many different factors, which span a variety of medical and nonmedical disciplines including gerontology, physiology, cardiology, pharmacology, epidemiology, climatology, toxicology, housing quality, energy usage, and socioeconomic factors. Very often studies have been restricted to one or another of these disciplines and there has been rather little effort to try to connect them together, for example indoor climate and heart disease. Despite the enormity of the problem on a worldwide scale, for example, in Europe alone, there are about one-quarter of a million excess winter deaths each year, there is surprisingly little research being conducted in this field.
References B^kenes, L., Mercer, J.B., 2000. Indoor climate and thermoregulatory patterns in aged living in Troms^, Norway. 11th International Congress on Circumpolar Health, Harstad, Norway.
Boyd, J.T., 1960. Climate, air pollution and mortality. Br. J. Preventive Social Med. 14, 123–135. Bull, G.M., 1973. Meterological correlates with myocardial and cerebral infarction and respiratory disease. Br. J. Preventive Social Med. 27, 108–113. Bull, G.M., Morton, J., 1975. Seasonal and short-term relationship of temperature with deaths from myocardial and cerebral infarction. Age Ageing 4, 19–31. Bull, G.M., Morton, J., 1978. Environment, temperature & death rates. Age Ageing 7, 210–224. Cheng, G., 1993. Investigation on the correlation between the mortality of cerebrovaxcular diseases and the meteorological factors in Zhanjiang City. Chung. Liu. Hsing. Ping. Hseuh. Tsa. Chih. 14, 234–236. Curwen, M., 1990. Excess winter mortality: a British phenomenon? Health Trends 4, 169–175. Danet, S., Richard, F., Montaye, M., Beauchant, S., Lemaire, B., Graux, C., Cottel, D., Mare´caux, N., Amouyel, P., 1999. Unhealthy effects of atmospheric temperature and pressure on the occurrance of myocardial infarction and coronary deaths. Circulation 100, E1–E7. Donaldson, G.C., Keatinge, W.R., 1997. Early increase in ischaemic heart disease mortality dissociated from and later changes associated with respiratory mortality after cold weather in south east England. J. Epidemiol. Community Health 51, 643–648. Douglas, A.A., Dunningan, M.G., Allan, T.M., Rawles, J.M., 1995. Seasonal variation in coronary heart disease in Scotland. J. Epidemiol. Community Health 49, 575–582. Eng, H., Mercer, J.B., 1998. Seasonal variations in mortality caused by cardiovascular disease in Norway and Ireland. J. Cardiovasc. Risk 5, 89–95. Eng, H., Mercer, J.B., 2000. The relationship between mortality caused by cardiovascular diseases and two climatic factors in densely populated areas in Norway and Ireland. J. Cardiovasc. Risk 7, 369–375. Eurowinter Group, Keatinge, W.R., Donaldson, G.C. et al., 1997. Cold exposure and winter mortality from ischaemic heart disease, cerebrovascular disease, respiratory disease, and all causes in warm and cold regions of Europe. Lancet 349, 1341–1346. Gordon, C.J., 2001. Role of environmental stress in the physiological response to chemical toxicants. Environ Res. in press. Gyllerup, S., Lanke, J., Lindholm, L.H., Schersten, B., 1991. High coronary mortality in cold regions of sweden. J. Intern. Med. 230, 479–485. Houdas, Y., Deklunder, G., Lecroart, J.L., 1992. Cold exposure and ischemic heart disease. Int. J. Sports Med. 13 (Suppl. 1), S179–S181. Jennings, J.R., Reynolds, C.F., Houck, P.R., et al., 1993. Age and sleep modify finger temperature responses to facial cooling. J. Gerontol. 48, M108–M116. Kannel, W.B., Wolf, P.A., Castelli, .W.P., D’Agostino, R.B., 1987. Fibrinogen and risk of cardiovascular disease. J. Am. Med. Assoc. 258, 1183–1186. Kawahara, J., Sano, H., Fukuzaki, H., saito, K., Hirouchi, H., 1989. Acute effects of exposure to cold on blood pressure, platelet function and symapthetic nervous activity in humans. Am. J. Hypertens. 2, 724–726. Keatinge, W.R., 1986. Seasonal mortality among elderly people with unrestricted home heating. Br. Med. J. 293, 732–733. Keatinge, W.R., Donaldson, G.C., 1995. Cardiovacular mortality in winter. Arctic Med. Res. 54 (Suppl. 2), 16–18. Keatinge, W.R., Coleshaw, S.R.K., Corter, F., Mattock, K., Murphy, M., Chelliah, R., 1984. Increases in platelets and red cell counts, blood viscosity, and arterial pressure during mild surface cooling: factors in mortality from coronary and cerebral thrombsis in winter. Br. Med. J. 289, 1405–1408.
J.B. Mercer / Environmental Research 92 (2003) 8–13 Kloner, R.A., Poole, W.K., Perritt, R.L., 1999. When throughout the year is coronary death most likely to occur? A 12-year populationbased analysis of over 220,000 cases. Circulation 100, 1630–1634. Kunst, A.E., Looman, C.W.N., Mackenbach, J.P., 1993. Outdoor air temperature and mortality in the Netherlands: a time-series analysis. Am. J. Epidemiol. 137, 331–341. Laake, K., Sverre, J.M., 1996. Winter excess mortality: a comparison between Norway and England plus Wales. Age Ageing 25, 343–348. Lanska, D.J., Hoffmann, R.G., 1999. Seasonal variation in stroke mortality rates. Neurology 52, 984–990. Mackenbach, J.P., Kunst, A.E., Looman, C.W.N., 1992. Seasonal variation in mortality in The Netherlands. J. Epidemiol. Community Health 46, 261–265. McKee, C.M., 1990. Deaths in winter in Northern Ireland: the role of low temperature. Ulster Med. J. 59, 17–22. Mercer, J.B., Østerud, B., Tveita, T., 1999. The effect of short term mild exposure on risk factors for cardiovascular disease. Thrombosis Res. 95, 93–104. Middleton, M., Andrews, J.F., Mercer, J.B., 2000. Thermal and relative humidity conditions in the houses of the aged living in Ireland, and their risk of effective acute cold stress. J. Circumpolar Health 59, 249–254. Momiyama, M., 1968. Biometeorological study of the seasonal variation of mortality in Japan and other countries on the seasonal disease calendar. Int. J. Biometeorol. 12, 377–393. Opper, C., Hennig, J., Clement, C., et al., 1995. Lowering of body temperature affects human platelet functions and norepinephrine release. Pharmacol. Biochem. Behav. 51, 217–221. Ornato, J.P., Siegel, L., Caren, E.J., Nelson, N., 1990. Increased incidence of cardiac death attributed to acute myocardial infarction during winter. Coronary Artery Dis. 1, 199–204.
13
Qizibash, N., Jones, L., Warlow, C., Mann, J., 1991. Fibrinogen and lipid concentration as risk factors for transient ischaemic attacks and minor ischaemic strokes. Br. Med. J. 303, 605–609. Rose, G., 1966. Cold weather and ischaemic heart disease. Br. J. Preventive Social Med. 20, 97–100. Smith, W.C., 1984. European regional variation in cardiovascular mortality. Br. Med. Bull. 40, 374–379. Spencer, F.A., Goldberg, R.J., Becker, R.C., Gore, J.M., 1998. Seasonal distribution of acute myocardial infarction in the second National Registry of Myocardial Infarction. J. Am. Coll. Cardiol. 31, 1226–1233. Stout, R.W., Crawfold, V., 1991. Seasonal variations in fibrinogen concentration among elderly people. Lancet 338, 9–13. Tillett, H.E., Smith, J.W., Gooch, C.C., 1983. Excess deaths attributable to influenza in England and Wales: age of death certified cause. Int. J. Epidemiol. 12, 344–352. Tyrrell, D., Barrow, I., Arthur, J., 1989. Seasonal mortality among elderly people with unrestricted home heating. Br. Med. J. 298, 1280–1283. Wagner, J.A., Horvath, S.M., 1985. Influence of age and gender on human thermoregulatory response to cold exposure. J. Appl. Physiol. 58, 180–186. WHOSIS, 2001. World Health Organization Statistical Information System, http://www.who.int/whosis/. Woodhouse, P.R., Khaw, K.T., Plummer, M., Foley, A., Meade, T.W., 1994. Seasonal variations of plasma fibrinogen and factor VII activity in the elderly: winter infections and death from cardiovascular disease. Lancet 343, 345–439. Young, A.J., Lee, D.T., 1995. Ageing and human cold tolerance. Exp. Ageing Res. 23, 45–67.
Cold—an underrated risk factor for health James B. Mercer Department of Medical Physiology, Faculty of Medicine, University of Troms^, Troms^ N–9037, Norway Received 4 February 2002
Abstract Cardiovascular diseases (CVD) are responsible for around 20% of all deaths worldwide (approximately 14 million) and are the principal cause of death in all developed countries, accounting for 50% of all deaths. Variations in the annual per capita death rates in different countries are well documented. Less well known are seasonal variations in death rates, with the highest levels occurring during the colder winter months, which have been described in many countries. This phenomenon is referred to as excess winter mortality. CVD-related deaths account for the majority of excess winter deaths (up to 70% in some countries), while about half of the remaining are due to increases in respiratory diseases. Paradoxically, CVD mortality increases to a greater extent with a given fall in temperature in regions with warm winters. While much of the indirect evidence points to the notion that cold is somehow involved in explaining excess winter deaths, the mechanism by which seemingly mild exposure to cold ambient conditions can increase the risk of death remains unclear. The strong indirect epidemiological evidence coupling cold climate to mortality may be related to indoor rather than outdoor climatic conditions (e.g., cold/damp houses versus warm/dry houses) coupled with a plethora of factors including health status, ageing-related deterioration in physiological and behavioral thermoregulation, toxicology, and socioeconomic factors. r 2002 Elsevier Science (USA). All rights reserved. Keywords: Cold climate; Winter mortality; CVD; Risk factor; Air temperature
1. Introduction One of the many environmental factors which affect human health is climate and the health risks associated with exposure to extreme environmental conditions are well documented. However, while living and working in extreme climatic conditions constitutes an obvious potential risk to health it is not always immediately clear that people living and working under milder climatic situations are also at risk. The purpose behind this paper is to briefly review some of the evidence, both direct and indirect, indicating that cold may be a more important risk factor for morbidity and mortality from disease than most people realize, even in countries with mild winter climates. Since cardiovascular diseases (CVD) are responsible for around 20% of all deaths worldwide (approximately 14 million) and are the principal cause of death in all developed countries, accounting for 50% of all deaths (42% in the United States), this view will mainly be concerned with these diseases of which the most common disease categories
Fax: +47-776-45440. E-mail address: [email protected].
are hypertension, ischemic heart disease, and cerebrovascular disease. In examining the problem of cold as a risk factor, epidemiological evidence showing seasonal variations in morbidity and mortality will be described. Epidemiological evidence showing that winter mortality is more pronounced in countries having milder winter climates will also be presented. Since the majority of registered deaths from CVD occur in the elderly, at least in the developed countries, a brief examination will be made of the question whether people in this age group during the course of their normal daily lives really do experience exposure to cold to a degree which may be detrimental for health.
2. Excess winter mortality— epidemiological studies In the so-called developed countries mortality due to CVD is the most common cause of death. Mortality from CVD is also emerging as a prominent public health problem in developing countries, ranking third (ca. 16% of all deaths), with particularly high impact on low
0013-9351/03/$ - see front matter r 2002 Elsevier Science (USA). All rights reserved. doi:10.1016/S0013-9351(02)00009-9
J.B. Mercer / Environmental Research 92 (2003) 8–13
socioeconomic groups, and CVD-related deaths have already become the first cause of mortality in developing countries such as Argentina, Chile, Cuba, Republic of Korea, Mauritius, Singapore, Sri Lanka, Trinidad and Tobago, and Uruguay (WHOSIS, 2001). While variations in the per capita death rates in different countries are well documented and familiar to most of us, seasonal variations in mortality and morbidity are generally less well known. In fact, seasonal variations in mortality with the highest levels occurring during the colder winter months have been described throughout Europe (Kunst et al., 1993; Eng and Mercer, 1998; Rose, 1966; Mackenbach et al., 1992; Keatinge and Donaldson, 1995), in the Unites States (Kloner et al., 1999; Lanska and Hoffmann, 1999), and in other parts of the world, including China (Cheng, 1993) and the southern hemisphere. Even in Japan, a country considered to have one of the world’s lowest per capita mortality rates from CVD, there is a very clear increase in winter mortality (Ornato et al., 1990). In addition to variations between countries, regional variations in mortality within countries have also been described with higher mortality rates in colder or more northern parts of a country (Momiyama, 1968; Smith, 1984; Gyllerup et al., 1991). The phenomenon of increased winter mortality is often referred to as excess winter mortality. When discussing excess winter mortality, one is mainly concerned with the older section of the population, at least in the developed countries, since the majority of deaths from CVD occur in the elderly. For example, in Norway, where CVD, including ischemic heart disease and cerebrovascular disease, accounts for about 46% of all causes of death, about 90% of all CVD-related deaths occur in people who are 65 years or older (Eng and Mercer, 1998). While there are associations between high CVD mortality in the elderly during the winter and high CVD mortality in the coldest parts of Europe and other countries, the actual reason for the excess number of CVD-related deaths in winter is not fully known. Much of the indirect evidence points to the notion that cold is somehow involved. However, the mechanism by which seemingly mild exposure to cold ambient conditions can increase the risk of death remains unclear. Not all excess winter deaths can be attributed to CVD and, while mortality from CVD seems to account for the majority of excess winter deaths (up to 70% in some countries), about half of the remaining may be accounted for by increases in respiratory disease (Eurowinter Group, 1997; Mackenbach et al., 1992). While this article focuses mainly on CVD-related deaths, the importance of cold-related respiratory mortality should not be underestimated. Indeed, there is some evidence that respiratory deaths resulting from influenza epidemics in winter may account for as much
9
as two-thirds of excess winter deaths (Tillett et al., 1983) It has also been reported that the temperature of the air breathed greatly affects respiratory infections (Tyrrell et al., 1989) and that cold housing affects respiratory mortality in winter (Keatinge and Donaldson, 1995). Furthermore, in a fashion similar to that of CVD mortality (see below), cold spells are followed, albeit with some delay, by an increase in respiratory mortality (Donaldson and Keatinge, 1997). One of the many plausible explanations for increased winter mortality lies in the fact that annual changes in temperature and acclimatization to a given set of environmental conditions may influence physiological responses to toxic chemicals (Gordon, 2001).
3. Winter mortality is greater in countries with milder winters—a paradox? In those countries in which seasonal fluctuations in mortality from CVD have been documented, the amplitude of the fluctuations show quite large variations, being much greater in some countries than in others. While at first this might seem surprising, levels of excess winter mortality are relatively lower in Scandinavian countries than in Great Britain (Curwen, 1990; Laake and Sverre, 1996). Likewise, even though the yearly per capita mortality rate is relatively low in Mediterranean countries, the seasonal changes in mortality are greater in many Mediterranean countries than in Scandinavia (Curwen, 1990). The data presented in Fig. 1, which compares seasonal changes in mortality from CVD averaged over a 10-year period in Norway, a country with a cold winter climate, and Ireland, a country with a mild winter climate, are used to emphasize this apparently paradoxical point. In Norway and Ireland, the mortality rate is, respectively, 28% and 45% higher in winter (December) than in summer (August) (Fig. 1, top). In some years this winter– summer difference has been as high as 60% (Eng and Mercer, 1998). The amplitude of the seasonal changes in mortality from CVD are similar irrespective of whether ones looks at the effect on all causes of CVD or at more specific causes of death. For example, in both Norway and Ireland, the amplitude of the seasonal changes in mortality for stroke and acute myocardial infarction are similar to the respective seasonal changes in all CVD-related deaths (Eng and Mercer, 2000).
4. Seasonal variations in incidences of acute myocardial infarction There are rather few seasonality-related morbidity studies based, for example, on hospital admissions.
10
J.B. Mercer / Environmental Research 92 (2003) 8–13
(AMI), at 1474 US Hospitals, clear seasonal variations were seen, with January being the month with the most reported cases (13,025 cases/30-day month) and July with the least (5978 cases/30-day month).
5. Mortality and air temperature Since mortality rates from CVD increases in the winter it can be expected that mortality will be inversely related to air temperature, and in nearly all countries so far examined a strong negative correlation between the outside air temperature and the mortality from CVD has been shown (Rose, 1966; Bull, 1973; Bull and Morton, 1978; McKee; 1990; Douglas et al., 1995; Laake and Sverre, 1996; Eng and Mercer, 1998, 2000). This point is demonstrated in Fig. 1 with data from Norway and Ireland. Fig. 1 (bottom), in which CVD mortality is related to air temperature, shows that mortality rate increases sharply when the outside air temperature falls below about 01C in Norway. However, in Ireland, which has a milder winter climate than Norway (Fig. 1, middle), mortality rate begins to rise sharply at higher outside air temperatures (ca. 10–151C) (Fig. 1, bottom). It has been suggested that this difference reflects, among other things, poorer housing standards (insulation, heating, etc.) and/or different thermoregulatory behavior patterns in the elderly in Ireland as compared to those in Norway (Eng and Mercer, 1998). However, there is some debate with regard to the relative importance of indoor cold stress versus outdoor cold stress in causing winter mortality. While, the present article stresses the importance of the former (see below) there is evidence that the latter also has a contributory role (Keatinge, 1986; Eurowinter Group, 1997). Fig. 1. (Top) Mean monthly CVD mortality ratio (ICD 9:390–459) in Norway and Ireland in the age group 60 years and older. (Middle) Mean monthly air temperatures in Norway (10 weather stations) and Ireland (five weather stations). (Bottom) Relation between mean monthly cardiovascular disease mortality ratio (ICE 9: 390–459) and mean monthly air temperature in Norway and Ireland in the age group 60 years and older. All data are based on monthly air temperature in Norway and Ireland in the age group 60 years and older. All data are based on monthly averages for the 10-year period 1985–1994; Norway, closed symbols; Ireland open symbols. Mortality ratio is defined as the average number of deaths per day in a given month divided by the average number of deaths per day for the whole year.
However, the few available studies clearly demonstrate clear seasonal variations (e.g., Spencer et al., 1998; Danet et al., 1999). In the study of Spencer et al., (1998), which was based on data from the National Registry of Myocardial Infarction in the United States using a cross-sectional observational database of patients admitted to hospitals with acute myocardial infraction
6. The lag effect In the cases of both myocardial infarction and stroke there appears to be a significant interval between a temperature change and the onset of the episode leading to death. In other words, a sudden large decrease in air temperature is not associated with an immediate change in mortality rate. Increases in mortality first become apparent several days or even a week later (Boyd, 1960; Donaldson and Keatinge, 1997; Kunst et al., 1993). This fact is potentially important in that if one could identify a link in the chain of events leading from temperature change to death one might be able to break it and avert the fatal outcome (Bull and Morton, 1975). However, studies to date have not established a clear chain of events leading from a change in external temperature to death. Nevertheless, it remains very likely that changes
J.B. Mercer / Environmental Research 92 (2003) 8–13
in ambient temperature cause changes in death rates, especially in the elderly.
7. Changes in risk factors with season and with cold exposure There are several factors considered indicators of risk for CVD which have been shown to have clear seasonal variations, including winter increases in plasma cholesterol, plasma fibrinogen, blood pressure, and red and white blood cell counts (Stout and Crawford, 1991; Woodhouse et al., 1994). In addition, there are several changes in blood composition which are known to occur during acute cold exposure. These include increased red cell count, increased plasma cholesterol, and increased plasma fibrinogen (Keatinge et al., 1984; Kannel et al., 1987; Kawahara et al., 1989; Qizibash et al., 1991; Stout and Crawford, 1991). In another cold exposure study, the density distribution of blood platelet subpopulations was shifted to an increase in less dense platelets that were more sensitive toward aggregation-inducing agents (Opper et al., 1995). In a cold exposure study in young adults, mean blood pressure, b-thromboglobulin, platelet factor 4, and plasma noradrenaline were increased. In addition, cold exposure-induced increases in plasma fibrinogen levels and thromboxane B2 have also been reported (Mercer et al., 1999). From the above findings it can be concluded that short-term exposure to cold initiates a mild inflammatory reaction and a tendency for an increased state of hypercoagulability. However, it is important to recognize that most of these data have been obtained in experiments involving the exposure of healthy young subjects to levels of cold stress which most of us, especially the elderly, will rarely experience during the course of our normal daily lives. This raises the issue of the actual levels of cold exposure experienced by people on a day-to-day basis. For example, it has been suggested that short-term exposure to even rather mild cold conditions may be a risk for the elderly (Houdas et al., 1992; Keatinge and Donaldson, 1995), while others (see below) feel that the effect of cold may be more subtle and may be more related to long-term exposure to mild levels of cold.
8. Indoor climatic conditions and excess winter mortality—inappropriately insulated houses and/or thermoregulatory behavior The question concerning how much cold exposure people are actually subjected to during the normal course of their daily lives was examined in a recent European Economic Community-supported study (Eurowinter Group, 1997) investigating cold exposure and winter mortality from ischemic heart disease,
11
cerebrovascular disease, respiratory disease, and all causes in warm and cold regions of Europe. The main findings of this study were that mortality increased to a greater extent with a given fall in temperature in regions with warm winters, in populations with cooler homes, and among people who wore fewer clothes and were less active outdoors. That people may become cold-exposed under thermal conditions which most of us consider mild is actually very plausible when one considers that humans, from a physiological point of view, are tropical animals (a naked adult human requires an air temperature of ca. 301C to feel thermally comfortable—the socalled lower critical temperature). When we consider that about 90% of all cardiovascular-related deaths occur in persons who are 65 years and older (Eng and Mercer, 1998) and who often spend a considerable portion of their elderly lives living indoors (in some cases as much as 100% of their time is spent indoors), it is clear that we cannot ignore the possibility that adverse climatic conditions within people’s homes may be involved in explaining excess winter mortality. However, while there is strong indirect evidence that housing standards and indoor climate may be important factors in explaining excess winter mortality (Eurowinter Group, 1997), the question of whether this is really true arises, especially in relatively wealthy countries such as Norway in which the general standards of housing are recognized as being very high. To answer these sorts of questions we need to have information on many different factors related to cold exposure during the course of our normal daily lives, including information on indoor climatic conditions and thermoregulatory behavior patterns, especially in the elderly section of the population. While people living indoors may be potentially at risk for cold exposure, particularly during the winter months, they need not necessarily be exposed to cold if they restrict heat loss by behavioral means, for example by wearing appropriate clothing and adequately heating their homes. However, it has to be recognized also that, while the quality of housing construction is important for maintaining a warm indoor climate, it is equally possible to maintain a warm indoor climate in a badly insulated house if one uses enough fuel just as it possible to have a cold indoor climate in the most modern and well-insulated house if one restricts the use of home heating. In this respect, knowledge of behavior patterns, especially those related to thermoregulatory behavior and home heating habits, is very important. Unfortunately, while many countries have clear regulations governing indoor climatic conditions in public buildings such as schools, hospitals, and the like, there is little or no regulation, simply recommendations, for private homes, and published data on indoor climatic conditions and thermoregulatory behavior patterns in private homes are scarce. Preliminary studies from Ireland and Norway confirm
12
J.B. Mercer / Environmental Research 92 (2003) 8–13
that the houses in Ireland are generally cooler and damper (higher relative humidity) (Middleton et al., 2000) than the considerably warmer and dryer Norwegian homes (B^kenes and Mercer, 2000). To form a more complete picture concerning how adverse indoor climatic conditions may affect people’s health, information concerning the medical status of the people involved is also needed. For example, if a normal healthy person feels cold, he or she puts on more clothes, heats the room, has a warm drink, or exercises to maintain normal body temperature. However, the situation is often different in the elderly and both the ability of people to detect that their body temperature is falling and the ability to prevent such a decrease deteriorates with age (Wagner and Horvath, 1985; Jennings et al., 1993; Young and Lee, 1995). The body’s ability to regulate body temperature may also be affected by a variety of diseases and by the use of commonly used pharmaceutical agents, such as those used for sleeping and depressive disorders and betablockers used in the treatment of cardiovascular disorders. In connection with normal ageing-related alterations in the way that people respond to their thermal environments, some preliminary results from a study of elderly subjects living in private homes in northern Norway indicate that even in the well-insulated Norwegian homes the elderly have peripheral (foot skin) temperatures which are lower during the winter months than the summer months, indicating that they are more cold-stressed during this time of the year (J.B. Mercer and B^kenes, unpublished observation).
9. Conclusions Clearly the reasons behind excess winter mortality are complex and involve many different factors, which span a variety of medical and nonmedical disciplines including gerontology, physiology, cardiology, pharmacology, epidemiology, climatology, toxicology, housing quality, energy usage, and socioeconomic factors. Very often studies have been restricted to one or another of these disciplines and there has been rather little effort to try to connect them together, for example indoor climate and heart disease. Despite the enormity of the problem on a worldwide scale, for example, in Europe alone, there are about one-quarter of a million excess winter deaths each year, there is surprisingly little research being conducted in this field.
References B^kenes, L., Mercer, J.B., 2000. Indoor climate and thermoregulatory patterns in aged living in Troms^, Norway. 11th International Congress on Circumpolar Health, Harstad, Norway.
Boyd, J.T., 1960. Climate, air pollution and mortality. Br. J. Preventive Social Med. 14, 123–135. Bull, G.M., 1973. Meterological correlates with myocardial and cerebral infarction and respiratory disease. Br. J. Preventive Social Med. 27, 108–113. Bull, G.M., Morton, J., 1975. Seasonal and short-term relationship of temperature with deaths from myocardial and cerebral infarction. Age Ageing 4, 19–31. Bull, G.M., Morton, J., 1978. Environment, temperature & death rates. Age Ageing 7, 210–224. Cheng, G., 1993. Investigation on the correlation between the mortality of cerebrovaxcular diseases and the meteorological factors in Zhanjiang City. Chung. Liu. Hsing. Ping. Hseuh. Tsa. Chih. 14, 234–236. Curwen, M., 1990. Excess winter mortality: a British phenomenon? Health Trends 4, 169–175. Danet, S., Richard, F., Montaye, M., Beauchant, S., Lemaire, B., Graux, C., Cottel, D., Mare´caux, N., Amouyel, P., 1999. Unhealthy effects of atmospheric temperature and pressure on the occurrance of myocardial infarction and coronary deaths. Circulation 100, E1–E7. Donaldson, G.C., Keatinge, W.R., 1997. Early increase in ischaemic heart disease mortality dissociated from and later changes associated with respiratory mortality after cold weather in south east England. J. Epidemiol. Community Health 51, 643–648. Douglas, A.A., Dunningan, M.G., Allan, T.M., Rawles, J.M., 1995. Seasonal variation in coronary heart disease in Scotland. J. Epidemiol. Community Health 49, 575–582. Eng, H., Mercer, J.B., 1998. Seasonal variations in mortality caused by cardiovascular disease in Norway and Ireland. J. Cardiovasc. Risk 5, 89–95. Eng, H., Mercer, J.B., 2000. The relationship between mortality caused by cardiovascular diseases and two climatic factors in densely populated areas in Norway and Ireland. J. Cardiovasc. Risk 7, 369–375. Eurowinter Group, Keatinge, W.R., Donaldson, G.C. et al., 1997. Cold exposure and winter mortality from ischaemic heart disease, cerebrovascular disease, respiratory disease, and all causes in warm and cold regions of Europe. Lancet 349, 1341–1346. Gordon, C.J., 2001. Role of environmental stress in the physiological response to chemical toxicants. Environ Res. in press. Gyllerup, S., Lanke, J., Lindholm, L.H., Schersten, B., 1991. High coronary mortality in cold regions of sweden. J. Intern. Med. 230, 479–485. Houdas, Y., Deklunder, G., Lecroart, J.L., 1992. Cold exposure and ischemic heart disease. Int. J. Sports Med. 13 (Suppl. 1), S179–S181. Jennings, J.R., Reynolds, C.F., Houck, P.R., et al., 1993. Age and sleep modify finger temperature responses to facial cooling. J. Gerontol. 48, M108–M116. Kannel, W.B., Wolf, P.A., Castelli, .W.P., D’Agostino, R.B., 1987. Fibrinogen and risk of cardiovascular disease. J. Am. Med. Assoc. 258, 1183–1186. Kawahara, J., Sano, H., Fukuzaki, H., saito, K., Hirouchi, H., 1989. Acute effects of exposure to cold on blood pressure, platelet function and symapthetic nervous activity in humans. Am. J. Hypertens. 2, 724–726. Keatinge, W.R., 1986. Seasonal mortality among elderly people with unrestricted home heating. Br. Med. J. 293, 732–733. Keatinge, W.R., Donaldson, G.C., 1995. Cardiovacular mortality in winter. Arctic Med. Res. 54 (Suppl. 2), 16–18. Keatinge, W.R., Coleshaw, S.R.K., Corter, F., Mattock, K., Murphy, M., Chelliah, R., 1984. Increases in platelets and red cell counts, blood viscosity, and arterial pressure during mild surface cooling: factors in mortality from coronary and cerebral thrombsis in winter. Br. Med. J. 289, 1405–1408.
J.B. Mercer / Environmental Research 92 (2003) 8–13 Kloner, R.A., Poole, W.K., Perritt, R.L., 1999. When throughout the year is coronary death most likely to occur? A 12-year populationbased analysis of over 220,000 cases. Circulation 100, 1630–1634. Kunst, A.E., Looman, C.W.N., Mackenbach, J.P., 1993. Outdoor air temperature and mortality in the Netherlands: a time-series analysis. Am. J. Epidemiol. 137, 331–341. Laake, K., Sverre, J.M., 1996. Winter excess mortality: a comparison between Norway and England plus Wales. Age Ageing 25, 343–348. Lanska, D.J., Hoffmann, R.G., 1999. Seasonal variation in stroke mortality rates. Neurology 52, 984–990. Mackenbach, J.P., Kunst, A.E., Looman, C.W.N., 1992. Seasonal variation in mortality in The Netherlands. J. Epidemiol. Community Health 46, 261–265. McKee, C.M., 1990. Deaths in winter in Northern Ireland: the role of low temperature. Ulster Med. J. 59, 17–22. Mercer, J.B., Østerud, B., Tveita, T., 1999. The effect of short term mild exposure on risk factors for cardiovascular disease. Thrombosis Res. 95, 93–104. Middleton, M., Andrews, J.F., Mercer, J.B., 2000. Thermal and relative humidity conditions in the houses of the aged living in Ireland, and their risk of effective acute cold stress. J. Circumpolar Health 59, 249–254. Momiyama, M., 1968. Biometeorological study of the seasonal variation of mortality in Japan and other countries on the seasonal disease calendar. Int. J. Biometeorol. 12, 377–393. Opper, C., Hennig, J., Clement, C., et al., 1995. Lowering of body temperature affects human platelet functions and norepinephrine release. Pharmacol. Biochem. Behav. 51, 217–221. Ornato, J.P., Siegel, L., Caren, E.J., Nelson, N., 1990. Increased incidence of cardiac death attributed to acute myocardial infarction during winter. Coronary Artery Dis. 1, 199–204.
13
Qizibash, N., Jones, L., Warlow, C., Mann, J., 1991. Fibrinogen and lipid concentration as risk factors for transient ischaemic attacks and minor ischaemic strokes. Br. Med. J. 303, 605–609. Rose, G., 1966. Cold weather and ischaemic heart disease. Br. J. Preventive Social Med. 20, 97–100. Smith, W.C., 1984. European regional variation in cardiovascular mortality. Br. Med. Bull. 40, 374–379. Spencer, F.A., Goldberg, R.J., Becker, R.C., Gore, J.M., 1998. Seasonal distribution of acute myocardial infarction in the second National Registry of Myocardial Infarction. J. Am. Coll. Cardiol. 31, 1226–1233. Stout, R.W., Crawfold, V., 1991. Seasonal variations in fibrinogen concentration among elderly people. Lancet 338, 9–13. Tillett, H.E., Smith, J.W., Gooch, C.C., 1983. Excess deaths attributable to influenza in England and Wales: age of death certified cause. Int. J. Epidemiol. 12, 344–352. Tyrrell, D., Barrow, I., Arthur, J., 1989. Seasonal mortality among elderly people with unrestricted home heating. Br. Med. J. 298, 1280–1283. Wagner, J.A., Horvath, S.M., 1985. Influence of age and gender on human thermoregulatory response to cold exposure. J. Appl. Physiol. 58, 180–186. WHOSIS, 2001. World Health Organization Statistical Information System, http://www.who.int/whosis/. Woodhouse, P.R., Khaw, K.T., Plummer, M., Foley, A., Meade, T.W., 1994. Seasonal variations of plasma fibrinogen and factor VII activity in the elderly: winter infections and death from cardiovascular disease. Lancet 343, 345–439. Young, A.J., Lee, D.T., 1995. Ageing and human cold tolerance. Exp. Ageing Res. 23, 45–67.