Short‐term fluctuations in death by cause,temperature, and income in the United States, 1930–1985

Abstract

Disclosures that this decade has had the five hottest years ever recorded globally raise concern that extreme temperatures might be associated with higher mortality. An analysis of fluctuations in annual cause‐specific deaths, seasonal temperatures, and annual income per capita in Massachusetts, Michigan, Washington, Utah, North Carolina, and Mississippi, 1930 to 1985, suggests that, on the contrary, a temperature increase throughout the year was associated with fewer deaths from all causes combined, including deaths from infectious diseases, heart diseases, cerebrovascular diseases, pneumonia, and influenza. An average temperature increase of one degree Fahrenheit was associated with a more than 2 per cent decline in deaths from pneumonia and influenza. The only category of deaths showing no significant association was death from malignant neoplasms. Compared to spring, summer, and fall temperature fluctuations, unusually cold winter temperatures had the strongest fatal effects, but only in North Carolina and Mississippi. The greatest cumulative temperature effects on mortality were found in the same two states. Controlling for annual fluctuations in income per capita did not influence the relationship between temperature and mortality. There was evidence suggesting that the level of wealth ameliorated the fatal effects of extreme temperatures. In conclusion, unusually warm weather was followed by fewer deaths; unusually cold weather, by more deaths.     

The association between temperature and mortality in tropical middle income Thailand from 1999 to 2008

We have investigated the association between tropical weather condition and age-sex adjusted death rates (ADR) in Thailand over a 10-year period from 1999 to 2008. Population, mortality, weather and air pollution data were obtained from four national databases. Alternating multivariable fractional polynomial (MFP) regression and stepwise multivariable linear regression analysis were used to sequentially build models of the associations between temperature variable and deaths, adjusted for the effects and interactions of age, sex, weather (6 variables), and air pollution (10 variables). The associations are explored and compared among three seasons (cold, hot and wet months) and four weather zones of Thailand (the North, Northeast, Central, and South regions). We found statistically significant associations between temperature and mortality in Thailand. The maximum temperature is the most important variable in predicting mortality. Overall, the association is nonlinear U-shape and 31 °C is the minimum-mortality temperature in Thailand. The death rates increase when maximum temperature increase with the highest rates in the North and Central during hot months. The final equation used in this study allowed estimation of the impact of a 4 °C increase in temperature as projected for Thailand by 2100; this analysis revealed that the heat-related deaths will increase more than the cold-related deaths avoided in the hot and wet months, and overall the net increase in expected mortality by region ranges from 5 to 13 % unless preventive measures were adopted. Overall, these results are useful for health impact assessment for the present situation and future public health implication of global climate change for tropical Thailand.     

A simple heat alert system for Melbourne,Australia

A simple heat alert system, based solely on predicted maximum and minimum daily temperatures, has been developed for the city of Melbourne in southeast Australia. The system is based upon a demonstration that, when mean daily temperature exceeds a threshold of 30°C (mean of today’s maximum temperature and tonight’s minimum temperature), the average daily mortality of people aged 65 years or more is about 15–17% greater than usual. Similar numbers of excess deaths also occur when daily minimum temperatures exceed 24°C (increases of 19–21% over expected death rate), so a heat alert system based solely on this widely available weather forecast variable is also feasible. No strong signal of excess heat-related deaths appears when the data are stratified using daily maximum temperatures. This may be because in Melbourne some days with very high maximum temperatures will be affected by the passage of cool changes and cold fronts in the afternoon, leading to a rapid drop in temperature (i.e., some days with high maximum temperatures will not continue to be hot throughout the day and into the evening). A single day with temperatures exceeding the thresholds noted above is sufficient to cause this increase in mortality, rather than requiring an extended heat wave. The increased daily mortality does not appear to represent a short-term advancement of mortality.     

Seasonality and coronary heart disease deaths in United States firefighters

United States firefighters have a high on-duty fatality rate, and coronary heart disease is the leading cause. Seasonality affects the incidence of cardiovascular events in the general population, but its effects on firefighters are unknown. This study statistically examined the seasonal and annual variation of all on-duty coronary heart disease deaths among US firefighters between 1994 and 2004 using the chi-square distribution and Poisson regression model of the monthly fatality counts. It also examined the effect of ambient temperature (apparent as well as wind chill temperature) on coronary heart disease fatalities during the study span using a time-stratified, case-crossover study design. When grouped by season, we observed the distribution of the 449 coronary heart disease fatalities to show a relative peak in winter (32%) and relative nadir in spring (21%). This pattern was significantly different (p=0.005) from the expected distribution under the null hypothesis of season having no effect. The pattern persisted in additional analyses, stratifying the deaths by the type of duty in which the firefighters were engaged at the time of their deaths. In the Poisson regression model of the monthly fatality counts, the overall goodness-of-fit between the actual and predicted case counts was excellent (χ₄²=16.63; p=0.002). Two distinct peaks were detected: one in January-February and the other in August-September. Overall temperature was not associated with increased risk of on-duty death. After allowing for different effects of temperature in mild/hot versus cold periods, a 1°C increase was not protective in cold weather; nor did it increase the risk of death in warmer weather. The findings of this study reveal statistical evidence for excess coronary heart disease deaths among firefighters during winter; however, the temporal pattern of coronary heart disease deaths was not linked to temperature variation. The seasonal pattern was also found to be independent of duty-related risks.     

Seasonality and Coronary Heart Disease Deaths in United States Firefighters

United States firefighters have a high on‐duty fatality rate, and coronary heart disease is the leading cause. Seasonality affects the incidence of cardiovascular events in the general population, but its effects on firefighters are unknown. This study statistically examined the seasonal and annual variation of all on‐duty coronary heart disease deaths among US firefighters between 1994 and 2004 using the chi‐square distribution and Poisson regression model of the monthly fatality counts. It also examined the effect of ambient temperature (apparent as well as wind chill temperature) on coronary heart disease fatalities during the study span using a time‐stratified, case‐crossover study design. When grouped by season, we observed the distribution of the 449 coronary heart disease fatalities to show a relative peak in winter (32%) and relative nadir in spring (21%). This pattern was significantly different (p=0.005) from the expected distribution under the null hypothesis of season having no effect. The pattern persisted in additional analyses, stratifying the deaths by the type of duty in which the firefighters were engaged at the time of their deaths. In the Poisson regression model of the monthly fatality counts, the overall goodness‐of‐fit between the actual and predicted case counts was excellent (χ₄²=16.63; p=0.002). Two distinct peaks were detected: one in January–February and the other in August–September. Overall temperature was not associated with increased risk of on‐duty death. After allowing for different effects of temperature in mild/hot versus cold periods, a 1°C increase was not protective in cold weather; nor did it increase the risk of death in warmer weather. The findings of this study reveal statistical evidence for excess coronary heart disease deaths among firefighters during winter; however, the temporal pattern of coronary heart disease deaths was not linked to temperature variation. The seasonal pattern was also found to be independent of duty‐related risks.     

Too hot to handle? Balancing increased trapability with capture mortality in hot weather pitfall trapping

Trapping at air temperatures close to, or exceeding, critical thermal maxima is important for comprehensive sampling of vertebrate assemblages and collection of sufficient data for impact assessment. However, pitfall trapping on hot days also potentially exposes trapped animals to stress or death through overheating or desiccation. We investigate causes of mortality from 14 305 captures over a 22‐year pitfall trap study in arid South Australia and compared mortality rates with maximum temperatures, solar radiation and rainfall. Overall mortality rate was 3.2% with chewing by rodents and handling accidents the most influential cause of death recorded. The highest mortality rates were experienced by the tiny skink, Lerista labialis, which was difficult to detect in traps each day and hence problematic to assess the effect of weather variables on capture mortality. For all other abundant species, high maximum temperature was only a significant explanatory variable for increased death rates of the house mouse Mus domesticus, and increased solar radiation was positively related to capture mortality for the house mouse, the frog Neobatrachus sudelli and the small skink Ctenotus schomburgkii. However, capture rates for these taxa and eight other common species would have been significantly lower if trapping did not occur on days of 40 °C or more. We conclude that trapping in hot weather is both desirable and justifiable and suggest techniques for further reducing mortality rates in pitfall studies.     

Weather and deaths in Pittsburgh,Pennsylvania: A comparison with Birmingham,Alabama

The relationship between weather and daily mortality was examined over a 4-year period in the temperate climate of Pittsburgh, Pennsylvania. Eight weather parameters were correlated with daily mortality using multiple, simple, and partial correlation techniques. Results from this study were then compared with results obtained from a previous investigation involving an identical analysis of the effects of weather on death in the subtropical climate of Birmingham, Alabama. Although the relationship between weather and total mortality is statistically significant in both areas, weather in the temperate region accounts for a greater portion of the daily variation in number of deaths. In both cities the effect of weather increases with age and is more intense among the white than the nonwhite population but does not appear to vary with sex. In both places weather significantly influences death due to respiratory diseases and circulatory diseases in general, but affects little, mortality from cancer or behaviorally related causes. The cities differ, however, in that Pittsburgh weather is significantly associated with deaths from ischemic heart disease but not with cerebrovascular mortality, while the reverse is observed in Birmingham. The cities also differ in specific meteorological factors and in the seasonal distribution of the intensity of the weather-mortality relationship.     

The potential effects of climate change on winter mortality in England and Wales

In Britain death rates from several important causes, particularly circulatory and respiratory diseases, rise markedly during the colder winter months. This close association between temperature and mortality suggests that climate change as a result of global warming may lead to a future reduction in excess winter deaths. This paper gives a brief introductory review of the literature on the links between cold conditions and health, and statistical models are subsequently developed of the associations between temperature and monthly mortality rates for the years 1968 to 1988 for England and Wales. Other factors, particularly the occurrence of influenza epidemics, are also taken into account. Highly significant negative associations were found between temperature and death rates from all causes and from chronic bronchitis, pneumonia, ischaemic heart disease and cerebrovascular disease. The statistical models developed from this analysis were used to compare death rates for current conditions with those that might be expected to occur in a future warmer climate. The results indicate that the higher temperatures predicted for 2050 might result in nearly 9000 fewer winter deaths each year with the largest contribution being from mortality from ischaemic heart disease. However, these preliminary estimates might change when further research is able to make into account a number of additional factors affecting the relationship between mortality and climate.     

The effects of monthly temperature fluctuations on mortality in the United States from 1921 to 1985

The impact of short-term temperature fluctuations on mortality has been studied mainly on historical populations, thus providing a limited ability to generalize to contemporary conditions, which would be more useful in determining public health policies aimed at reducing mortality. Therefore, this study examined the effects of monthly temperature fluctuations on mortality in the United States from 1921 to 1985. Monthly data about mortality from the Vital Statistics and temperature from the National Oceanic and Atmospheric Administration and the US Department of Agriculture Weather Bureau were used. Six states were selected to be studied (Massachusetts, Michigan, Washington, Utah, North Carolina, and Mississippi). The analysis was carried out using distributed lag models. The analysis-showed that warmer than usual temperatures in July and August, and unusually cold temperatures from January to June are linked to higher mortality. From September to December unusually low temperatures are associated with higher mortality in most states, while temperature has no significant effect on mortality in June and September. In January and February mortality is especially affected by unusually cold weather in the southern states of Mississippi and North Carolina. For example, a one degreee drop in the mean temperature in 1921 is associated with a more than 3.5% increase in the February crude death rate in Mississippi and North Carolina and a less than 1% increase in the four other states examined. Finally, in the months from January to March the relationship between monthly fluctuations in the crude death rate and temperature declined over time and became relatively weak by 1985.     

Hot weather warning might help to reduce elderly mortality in Hong Kong

While there was evidence on the relationship between extreme hot weather and the increase in mortality, particularly from ischaemic heart disease (IHD) and cerebrovascular disease (stroke), some researchers suggested that early warning systems might reduce mortality. In this study, the relationship between Very Hot Weather Warning (VHWW) and mortality was examined in the context of Hong Kong, which has a sub-tropical climate. An observational study was conducted on the daily number of deaths due to IHD and stroke in the Hong Kong elderly population (aged 65 or above) during summer (May–September) in 1997–2005. Totals of 4,281 deaths from IHD and 4,764 deaths from stroke occurred on days with maximum temperature reaching/exceeding 30.4°C. Multiple linear regression models were used to study the association between VHWW and the daily mortality rates from IHD and from stroke, respectively. Results showed that absence of VHWW was associated with an increase of about 1.23 (95% CI: 0.32, 2.14) deaths from IHD and 0.97 (95% CI: 0.02, 1.92) deaths from stroke among the elderly per day. Public education is required to inform the elderly to take appropriate preventive measures and to remind the public to pay more care and attention to the elderly on days which are not considered to be stressful to the general public. Warning systems tailored for the elderly could also be considered.     

Daily average temperature and mortality among the elderly: a meta-analysis and systematic review of epidemiological evidence

The impact of climate change on the health of vulnerable groups such as the elderly has been of increasing concern. However, to date there has been no meta-analysis of current literature relating to the effects of temperature fluctuations upon mortality amongst the elderly. We synthesised risk estimates of the overall impact of daily mean temperature on elderly mortality across different continents. A comprehensive literature search was conducted using MEDLINE and PubMed to identify papers published up to December 2010. Selection criteria including suitable temperature indicators, endpoints, study-designs and identification of threshold were used. A two-stage Bayesian hierarchical model was performed to summarise the percent increase in mortality with a 1°C temperature increase (or decrease) with 95% confidence intervals in hot (or cold) days, with lagged effects also measured. Fifteen studies met the eligibility criteria and almost 13 million elderly deaths were included in this meta-analysis. In total, there was a 2-5% increase for a 1°C increment during hot temperature intervals, and a 1-2 % increase in all-cause mortality for a 1°C decrease during cold temperature intervals. Lags of up to 9 days in exposure to cold temperature intervals were substantially associated with all-cause mortality, but no substantial lagged effects were observed for hot intervals. Thus, both hot and cold temperatures substantially increased mortality among the elderly, but the magnitude of heat-related effects seemed to be larger than that of cold effects within a global context.     

Environmental temperature and mortality from acute myocardial infarction

Mortality from acute myocardial infarction (MI) over the 5 year period 1982–1987 in Brown County, Wisconsin, was analyzed to assess the relationship with environmental temperature. Deaths occurrring on the day of and the day following a significant snowfall as well as deaths occuring in health care facilities were eliminated from consideration because the focus was upon temperature, not snowfall or events within a hospital. These criteria resulted in the inclusion of 1,802 days and 926 cases of acute MI. The mean temperature on the day of death was obtained from climatological data and were grouped into six categories covering a range of temperatures from<–17.8°C (0°F) to 16.1°C (61°F). The number of deaths in each category was tabulated. The effect of temperature, sex, and age were analyzed by regression analysis. The results indicated a linear increase in mortality as mean daily temperature decreased over the temperature range. The inverse temperature effect was most pronounced in males over the age of 60. These results indicate that cold temperatures appear to be associated with an increased mortality from myocardial infarction.     

Drowning and the Influence of Hot Weather

Background

Drowning deaths are devastating and preventable. Public perception does not regard hot weather as a common scenario for drowning deaths. The objective of our study was to test the association between hot weather and drowning risk.

Materials and Methods

We conducted a retrospective case-crossover analysis of all unintentional drowning deaths in Ontario, Canada from 1999 to 2009. Demographic data were obtained from the Office of the Chief Coroner. Weather data were obtained from Environment Canada. We used the pair-matched analytic approach for the case-crossover design to contrast the weather on the date of the drowning with the weather at the same location one week prior (control period).

Results

We identified 1243 drowning deaths. The mean age was 40 years, 82% were male, and most events (71%) occurred in open water. The pair-matched analytic approach indicated that temperatures exceeding 30°C were associated with a 69% increase in the risk of outdoor drowning (OR = 1.69, 95% CI 1.23–2.25, p = 0.001). For indoor drowning, however, temperatures exceeding 30°C were not associated with a statistically significant increase in the risk of drowning (OR = 1.50, 95% CI 0.53–4.21, p = 0.442). Adult men were specifically prone to drown in hot weather (OR 1.67, 95% CI 1.19–2.34, p = 0.003) yet an apparent increase in risk extended to both genders and all age groups.

Conclusion

Contrary to popular belief, hot weather rather than cold stormy weather increases the risk of drowning. An awareness of this risk might encourage greater use of drowning prevention strategies known to save lives.     

Temperature-induced excess mortality in Moscow,Russia

After considering the observed long-term trends in average monthly temperatures distribution in Moscow, the authors evaluated how acute mortality responded to changes in daily average, minimum and maximum temperatures throughout the year, and identified vulnerable population groups, by age and causes of death. A plot of the basic mortality–temperature relationship indicated that this relationship was V-shaped with the minimum around 18°C. Each 1°C increment of average daily temperature above 18°C resulted in an increase in deaths from all non-accidental causes by 2.8%, from coronary heart disease by 2.7%, from cerebrovascular diseases by 4.7%, and from respiratory diseases by 8.7%, with a lag of 0 or 1 day. Each 1°C drop of average daily temperature from +18°C to −10°C resulted in an increase in deaths from all non-accidental causes by 0.49%, from coronary heart disease by 0.57%, from cerebrovascular diseases by 0.78%, and from respiratory diseases by 1.5%, with lags of maximum association varying from 3 days for non-accidental mortality to 6 days for cerebrovascular mortality. In the age group 75+ years, corresponding risks were consistently higher by 13–30%. The authors also estimated the increase in non-accidental deaths against the variation of daily temperatures. For each 1°C increase of variation of temperature throughout the day, mortality increased by 0.3–1.9%, depending on other assumptions of the model.     

Skin histology and its role in heat dissipation in three pinniped species

ABSTRACT: BACKGROUND: Pinnipeds have a thick blubber layer and may have difficulty maintaining their body temperature during hot weather when on land. The skin is the main thermoregulatory conduit which emits excessive body heat. Methods: Thorough evaluation of the skin histology in three pinniped species; the California sea lion-Zalophus californianus, the Pacific harbor seal-Phoca vitulina richardsi, and the Northern elephant seal-Mirounga angustirostris, was conducted to identify the presence, location and distribution of skin structures which contribute to thermoregulation. These structures included hair, adipose tissue, sweat glands, vasculature, and arteriovenous anastomoses (AVA). Thermal imaging was performed on live animals of the same species to correlate histological findings with thermal emission of the skin. Results: The presence and distribution of skin structures directly relates to emissivity of the skin in all three species. Emissivity of skin in phocids (Pacific harbor and Northern elephant seals) follows a different pattern than skin in otariids (California sea lions). The flipper skin in phocids tends to be the most emissive region during hot weather and least emissive during cold weather. On the contrary in otariids, skin of the entire body has a tendency to be emissive during both hot and cold weather. Conclusion: Heat dissipation of the skin directly relates to the presence and distribution of skin structures in all three species. Different skin thermal dissipation patterns were observed in phocid versus otariid seals. Observed thermal patterns can be used for proper understanding of optimum thermal needs of seals housed in research facilities, rescue centers and zoo exhibits.     

Shifts in the seasonal distribution of deaths in Australia, 1968–2007

Studies in temperate countries have shown that both hot weather in summer and cold weather in winter increase short-term (daily) mortality. The gradual warming, decade on decade, that Australia has experienced since the 1960s, might therefore be expected to have differentially affected mortality in the two seasons, and thus indicate an early impact of climate change on human health. Failure to detect such a signal would challenge the widespread assumption that the effect of weather on mortality implies a similar effect of a change from the present to projected future climate. We examine the ratio of summer to winter deaths against a background of rising average annual temperatures over four decades: the ratio has increased from 0.71 to 0.86 since 1968. The same trend, albeit of varying strength, is evident in all states of Australia, in four age groups (aged 55 years and above) and in both sexes. Analysis of cause-specific mortality suggests that the change has so far been driven more by reduced winter mortality than by increased summer mortality. Furthermore, comparisons of this seasonal mortality ratio calculated in the warmest subsets of seasons in each decade, with that calculated in the coldest seasons, show that particularly warm annual conditions, which mimic the expected temperatures of future climate change, increase the likelihood of higher ratios (approaching 1:1). Overall, our results indicate that gradual climate change, as well as short-term weather variations, affect patterns of mortality.     

Thermosensory intensity and affect throughout the perceptible range

The thermosensory system was evaluated psychophysically in 12 healthy volunteers, spanning the full range of tolerable temperatures. Subjects provided ratings of (1) perceived thermal intensity, (2) perceived pleasantness or unpleasantness, and (3) perceived pain intensity after placing either one hand or foot in a temperature controlled water bath. Of particular interest were the interrelationships among the three perceptual measures, and differences between heat and cold. The relationship between perceived intensity and (un)pleasantness was different for hot vs cold stimuli. Specifically, for a given perceived thermal intensity, cold stimuli were rated as less pleasant or more unpleasant than hot stimuli. Similarly, for a given pain intensity, cold stimuli were rated as more unpleasant than hot stimuli. As warm temperatures increased and as cold temperatures decreased, stimuli were perceived as being unpleasant before they were perceived as being painful. The difference in transition temperatures for unpleasantness vs pain for heat averaged 1.4 degrees C, while the same difference for cold averaged 5.6 degrees C. Thus, there was a fourfold difference in the range of unpleasant but non-painful cold vs hot temperatures. Pain intensity and unpleasantness ratings were significantly higher for heat stimuli applied to the foot vs hand. In contrast, there was no significant body site difference for pain intensity or unpleasantness ratings of cold stimuli. All of these results reveal important differences in the processing of cold vs hot stimuli. These differences could be exploited to differentiate processing relevant to discriminative vs affective components of somesthetic perception, in both the innocuous and noxious ranges.     

Association between temperature and death in residential populations in Shanghai

The study is focused on patterns of daily deaths in Shanghai for the period from 1 May 1979 to 30 April 1980. From May to September the deaths in all age groups are lower, but increase gradually from October and reach to a peak in February. This confirms results found in other countries, namely the death rate is increased in winter. The peak for the population aged over 70 is the highest of the three different age groups. Correlation analyses were carried out on three temperature parameters (daily minimum, maximum and mean temperatures) and six categories of death (heart disease, coronary heart disease, cerebrovascular disease, cancer, respiratory disease and total deaths). The results reveal that the average daily temperature is very significant for the six categories of death. There are three correlations: straight line relationship, parabolic relationship and exponential relationship. These different types arise from the different morbidity rates. Death from the different disease is also increased during days when the daily maximum temperature is over 35° C or the daily minimum temperature is below 0°C. This shows, in general, that days of extreme temperature lead to an increase in the death rate.     

Deaths from heat-stroke in Japan: 1968–1994

Global warming is increasingly recognized as a threat to the survival of human beings, because it could cause a serious increase in the occurrence of diseases due to environmental heat during intermittent hot weather. To assess the direct impact of extremely hot weather on human health, we investigated heat-related deaths in Japan from 1968 through 1994, analyzing the data to determine the distribution of the deaths by age and their correlation to the incidence of hot days in summer. Vital Statistics of Japan, published by the Ministry of Health and Welfare of Japan, was the source of the heat-related mortality data employed in this study. Meteorological data were obtained from the District Meteorological Observatories in Tokyo and Osaka, the two largest cities in Japan. Heat-related deaths were most prone to occur on days with a peak daily temperature above 38°C, and the incidence of these deaths showed an exponential dependence on the number of hot days. Thus, even a small rise in atmospheric temperature may lead to a considerable increase in heat-related mortality, indicating the importance of combating global warming. Furthermore, half (50.1%) of the above-noted deaths occurred in children (4 years and under) and the elderly (70 years and over) irrespective of gender, indicating the vulnerability of these specific age groups to heat. Since a warmer climate is predicted in the future, the incidence of heat waves will increase, and more comprehensive measures, both medical and social, should be adopted for children of 4 years and younger the elderly to prevent heat-related deaths in these age groups. Received: 20 January 1999 / Accepted: 15 June 1999     

Plasma cholesterol,coronary heart disease,and cancer in the Renfrew and Paisley survey.

The relation between plasma cholesterol concentration and mortality from coronary heart disease, incidence of and mortality from cancer, and all cause mortality was studied in a general population aged 45-64 living in the west of Scotland. Seven thousand men (yielding 653 deaths from coronary heart disease, 630 new cases of cancer, and 463 deaths from cancer) and 8262 women (322 deaths from coronary heart disease, 554 new cases of cancer, and 395 deaths from cancer) were examined initially in 1972-6 and followed up for an average of 12 years. All cause mortality was not related to plasma cholesterol concentration. This was largely a consequence of a positive relation between cholesterol values and mortality from coronary heart disease being balanced by inverse relations between cholesterol and cancer and between cholesterol and other causes of death. These changes were highly significant for coronary heart disease and cancer in men and significant for coronary heart disease and other causes of death in women. The inverse association between cholesterol concentration and cancer in men was strongest for lung cancer, was not merely a function of the age at which a subject died, was present for the incidence of cancer as well as mortality from cancer, and persisted when new cases or deaths occurring within the first four years of follow up were excluded from the analysis.