The Office for National Statistics (ONS) provisional figures for excess winter deaths (EWDs) provide an insight into the rise in mortality during the colder months.
Published on 23 November, the statistical bulletin displays the figures by gender, age, region and cause.
In 2015/16, an estimated 15% more deaths occurred in winter months than non-winter months, giving a figure of approximately 24,300. This excess winter mortality (EWM) is in line with the average trend across England and Wales.
The most recent figures have however decreased from 2014/15, when excess deaths occurring around winter was at unusually high levels.
Across all age groups, the levels of excess winter mortality were similar, with those aged 0 – 64 affected more than in previous year and those aged 85+ affected less.
In 2015/16, there were more excess winter deaths for females than males, a trend which follows that of previous years. For both genders, the level of excess winter deaths decreased from 2014/15.
Respiratory diseases were the cause of over one-third of all excess winter deaths in 2015/16 across England and Wales.
Observing the levels of excess winter deaths over time, there is not a clear geographical pattern. For 2015/16 however, the highest levels of excess winter mortality were in Wales, with an index of 17%. The East of England in contrast, had the lowest excess winter mortality index at 13%.
Commenting on the fall in excess winter mortality was Dr Annie Campbell, from ONS: “One of the key factors behind the lower excess mortality this winter was a fall in the number of deaths among the elderly. This was mainly due to the most prevalent strain of the flu virus impacting younger people rather older people, who are more at risk.”
During the winter months, it is common for mortality numbers to increase. It has been found that historically speaking, above-average mortality usually occurs between December and March.
Thus in relation to the data, winter is defined as this period, and is compared with the average levels of mortality which occur in August to November and the subsequent April to July.
2015/16 returns to average trends
Excess winter deaths returned to average trends in 2015/16. The significant fall in comparison with 2014/15 can be mainly explained by the above average figures for the period, as opposed to uncommonly low figures in 2015/16. Another cause may have been the predominant influenza virus strain which significantly impacted young adults and had a reduced effect on the elderly.
Trends for the levels of excess winter deaths over time are not smooth and are commonly subject to large fluctuations. Since 1950/51, the levels of excess winter deaths have slowly decreased and this trend has gradually levelled off during recent years. Rather than the start of an upward trend, the 2014/15 figures were instead a fluctuation.
Increased morality extends beyond winter
From August 2015 to July 2016, the peak in daily deaths was later than reported in recent years, occurring in late January 2016.
The 5-year average indicates the pronounced increase in mortality typically occurring over the winter months. For 2015/16, the mortality growth was not as prominent over this period. Daily deaths were lower than the 5-year average during December.
Around the 19th July, there was an unusual peak in mortality. Higher than average temperatures were recorded by trends in provisional temperature data during this period. Temperatures recorded by the Met Office Hadley Centre indicate that the 19th July observed the highest annual levels, reaching a maximum temperature of 31 degrees. Thus, the high temperatures may be somewhat responsible for the high mortality at this time. Studies indicate that vulnerable groups of people such as the elderly and infants, may be more at risk of death during a heat wave.
The calculation of excess winter deaths can be affected by an increase in daily deaths during non-winter months. High levels during this time can make the excess winter death rates appear low, even if the figures are high. An increased number of daily deaths during non-winter months can conceal the increased mortality over the winter period.
EWM mostly affects females and elderly
From the excess winter deaths which occurred over 2015/16, 11,400 were males (47%) and 12,900 were females (53%). This higher proportion of females may be partly due to females tending to live longer, with 65% of those over the age of 85 being female.
The EWM index compares the number of people who dies during winter with those who died in non-winter months. This is displayed as a percentage of the average number of deaths which occurred in the non-winter months. Across both genders, those aged 85+ had the highest EWM index; approximately 17%. The next highest EWM index for males was for those aged 65 or younger (EWM 15%).
In years prior to this, an increase in EWM index has occurred across both genders from those aged 0 – 65 to those aged 85+. For 2015/16, this trend was not as prominent as each age group possessed a similar EWM index.
The EWM index fell significantly from 2014/15 to 2015/16 across all age groups, except from those aged between 0 – 64 which displayed significant increases in EWM index for both genders. The EWM index has for both genders roughly halved, where older age groups are concerned. For males aged 85+, the EWM index has decreased from 36% in 2014/15 to 17% in 2015/16 – a fall of 53%. The EWM index for females in the same age group also decreased, but from 42% in 2014/15 to 17% in 2015/16 – a drop of 60%. The lower levels of impact the influenza strain had upon the elderly may be a likely cause of this.
For females of all ages, the EWM index dropped to 15% compared with the 14% for males of all ages. These percentages are down from 31% and 23% in 2014/15 respectively.
Predominant Influenza strain impacts young more than elderly
Complications may cause an influenza infection to become life threatening. Those most susceptible to developing these complications are the elder and those with underlying health conditions which may subsequently result in death. One of the main causes of excess winter deaths in 2015/16 was respiratory disease, with 41% more respiratory deaths occurring in the winter months than the non-winter months. Of all excess winter deaths, respiratory disease was the underlying cause of 8,600 from 24,300 (35%). Of this figure, 72% were from those aged 75+.
Within the community, there were moderate levels of influenza activity across England and Wales in 2015/16 according to a Public Health England Report. This activity was also said to have peaked at a later point than in the previous year.
In 2015/16, the highest number of weekly deaths took place in the third week of 2016 with 11,459 deaths. The number remained high until week 11 where 11,125 deaths occurred. This coincides with the peak in influenza-like illness (ILI) consultation rate, and after this time, both the consultation rate and the number of weekly deaths declined.
The specific flu strain (influenza A(H1N1) pdm09) had a significant impact of children and younger adults; the highest numbers of ILI consultation rates were for those aged 5 – 14 (37 for every 100,000) and those aged 15 – 44 (37 per 100,000). This mirrors the pattern of the 2010/11 influenza strain although in that case the levels of influenza were much higher.
Despite the similarity in levels of influenza, in 2015/16 the EWM was lower than in 2014/15. This is likely to relate to the strain of influenza which was prominent in each season. Over 2014/15, the specific strain proved particularly infectious for older people, which is a group already at risk. The predominant strain in 2015/16 however, had much less impact on the elderly and was instead responsible for ICU admissions and hospitalisations in younger people.
For those aged between six months and 64 with one or more underlying clinical risk factor, Public Health England also reported a decrease in flu vaccination uptake. This fell by 5% between 2014/15 and 1015/16. There was a similar trend in Wales also, where flu vaccination uptake fell from 49% in 2014/15 to 47% in 2015/16 for the same group.
Disease and EWDs
Respiratory diseases had an EWM index of 41% in 2015/16, meaning this remained the most prominent underlying cause of excess winter deaths. This figure has fallen from the 2014/15 period, but is still much greater than the 2013/14 period, which was 30% lower. 8,600 from the 24,300 excess winter deaths occurred because of a respiratory disease, making up more than a third at 35%. The largest proportion of these EWDs was as a result of pneumonia.
One of the leading causes of death in 2015 was circulatory diseases, which also accounted for just under a quarter of excess winter deaths at 24%. This EWM index for circulatory diseases fell by 9% from 2014/15 to 2015/16. In comparison with respiratory diseases or dementia and Alzheimer’s, the seasonal impact is not as strong for circulatory diseases.
The main cause of death in 2015 was dementia and Alzheimer’s. It accounted for 4,300 deaths out of the 24,300 excess winter mortality deaths, a proportion of 18%, and there was a 23% more deaths of this sort during the winter period, in comparison with non-winter months.
Temperature and EWDs
Studies have indicated that as the temperature falls, the number of deaths increases, although this alone does not give a full picture in relation to excess winter mortality. In comparison with previous years, the typical seasonal peak in deaths during 2015/16 was more understated.
Although at 1,584 deaths January had the highest mean daily death rate from all the months, February and March also had high levels at 1,570 and 1,575 respectively. This is inconsistent with the 5-year average.
Data observed over the winter period supports the relationship between excess mortality rates and cold weather. Temperatures observed from November to February were above average, especially so in December. In comparison to the 5-year average, December had the most negative difference in terms of death numbers and temperature.
However, higher levels of EWDs are not always coupled paired with colder winters. This indicates that factors aside from temperature influence winter mortality trends such as disease.
Following the high EWD rates observed in 2014/15, the 2015/16 levels returned to the usual range for this decade. Between 1990/00 and the most recent period, the excess winter death rates have been between 20,000 and 30,000, with the figure increasing every 3 – 4 years. The 2015/16 period EWDs fall within this range.
Geographical patterns for EWDs
For the 2015/16 period, the region with the highest EWM index was Wales at 17%, then followed by both Yorkshire and the Humber as well as the South West at 15%. The region with the lowest EWM index was the East of England at 13%.
Contrasting with other mortality statistics such as drug-related deaths and avoidable mortality, there is not a clear “north-south divide” where excess winter deaths are concerned. For Wales, the EWM index was much higher than the average for England and Wales as well as in comparison to any region in England within 2015/16. For 2014/15 however, Wales had the lowest EWM index and was also the third lowest in 2013/14. Over time therefore, fluctuations in figures such as this indicate there is not a consistent regional pattern.
In every region from the 2014/15 period to the 2015/16 one, the EWM has significantly fallen. This rate however, remained higher than that witnessed in 2013/14 period.
EWM causes
On a European scale, rates of excess winter mortality have been seen to vary widely. Countries such as Finland and Germany with low winter temperatures observed low rates of EWM in comparison to rates in Portugal and Spain. For England and Wales, seasonal variation in mortality is high.
Countries with milder winters are likely to have higher levels of winter mortality. The reasons for this may be that the people who reside there take less precautions against cold weather. As the temperature is low for a shorter proportion of the year, homes are also likely to be less thermal efficient and make retaining heat more difficult during the winter. In England, low indoor temperature has been associated with higher EWM rates as a result of cardiovascular disease.
The cold may also bring about physiological issues, which can for vulnerable people be a real issue. This can include for example, increased blood pressure as well as thrombosis as a result of cold-induced haemoconcentration. Immune system resistance may also be reduced against respiratory infection.
Although the correlation between excess mortality and colder weather does exist, lower temperatures explain only a small part of the fluctuation in death rates during winter. Findings indicate that influenza as well a temperature play significant roles in predicting winter mortality levels.
