What Vague Causes of Death Really Tell us
Author: Wouter Ronsijn
On 8 September 2022, Queen Elizabeth II died, officially from ‘old age’. Today, deaths from old age are rare, because it is a very imprecise term, possibly masking an underlying, more specific disease. In the past, however, vague terms such as those were a common cause of death.
In the middle of the nineteenth century, a death in a Belgian village often left more questions than answers. When an infant died suddenly, a clerk might attribute it to ‘convulsions’. When an elderly woman passed away quietly in her home, the cause recorded often was ‘old age’. These vague terms, written into municipal cause of death registers, were often the only official trace of what caused the end of someone’s life.
What are Ill-Defined Causes of death
What is Dementia
Dementia is a syndrome of cognitive decline, affecting memory, reasoning, language, and daily activities. It results from neurodegenerative diseases like Alzheimer’s.
Where Does Dementia occur?
Dementia primarily affects older adults, making it prevalent in high-income countries with ageing populations. The global burden is increasing.
Diagnosis and Treatment
Diagnosis is based on clinical assessment, with neuroimaging improving accuracy. There is no cure, so treatment focuses on symptom management and support.
Prevention and Control
While incurable, modifiable risk factors like hypertension and inactivity are linked to increased risk. Public health efforts focus on prevention.
Vague terms such as the above describe how somebody died but fail to indicate the precise cause. Every death, however, has a cause – a process or event that brings a person’s life to an end. That cause can be natural, such as a disease, or external, such as an accident or intentional harm. For governments, understanding these causes has been essential. Only by knowing what people die from can authorities design health policies that save people’s lives.
However, identifying the precise cause of death is not self-evident. It requires information about the deceased and medical knowledge about possible causes, which were often lacking in the past. When Belgium started collecting data on causes of death in 1851, many municipalities had no doctors. Causes of death were based on declarations by family members or friends, or determined by local police officers. Even where doctors were involved, they often lacked diagnostic tools such as laboratory tests or X-ray technology.
Medical knowledge itself was limited. Some doctors also refused to cooperate, because they believed that patient-doctor confidentiality prevented them from making causes of death public. In addition, people felt embarred about certain causes of death, such as syphilis or cancer, which is why doctors sometimes wrote down vague terms instead. As a result, many deaths were recorded with ill-defined or unknown causes. Ill-defined causes include symptoms rather than actual diseases, such as convulsions, dropsy or fever, or vague terms such as ‘old age’, ‘natural causes’ or ‘sudden death’. Or the record may simply state that the cause ‘could not be determined’.
Over time, advances in medical knowledge and diagnostic technologies, the introduction of an International Classification of Diseases (ICD) in 1900, and the gradual introduction of confidential cause-of-death reporting (between 1930 and 1960 in different parts of Belgium) improved the accuracy and reliability of the data.
A Brief History
Belgium’s system of cause-of-death registration emerged from crisis. In the aftermath of the devastating cholera epidemic of 1849, the government introduced nationwide registration in 1851. Since then, municipalities needed to record the cause of each death within their territory and to submit annual summaries to the National Statistical Committee in Brussels.
In practice, registration started slowly. Many municipalities initially failed to submit reports or submitted reports with many vague or unclear causes of death. As a result, a large number of deaths were recorded with ill-defined causes: reports mentioned symptoms rather than diseases, attributed deaths to ‘old age’, described them as ‘sudden’, or admitted that the cause was unknown.
It took until 1861 for all municipalities to start sending in cause-of-death reports to the National Statistical Committee in Brussels. Yet municipalities still used a wide variety of terms, in three different languages (French, Dutch, Latin) in these reports. Some of these terms had unclear or ambiguous meanings. That made it difficult for the National Statistical Committee to interpret the reports and compile an accurate national overview. To deal with that problem, Belgium introduced a standardized list of causes of death in 1867 (see below). It contained 116 terms that could be used to describe why someone had died, and was meant to make the terminology more uniform. It was simplified in 1874 to contain 33 categories. In 1903, it was replaced by the first International Classification of Diseases (ICD)
Standardized list of causes of death, 1867
The share of deaths attributed to ill-defined causes depended heavily on the involvement of doctors, who were much better equipped than ordinary people to identify the correct cause of death in precise terms. Several cities therefore appointed doctors to verify causes of deaths or required medical certificates before burials were allowed. Yet not all municipalities had the financial means to appoint doctors, leaving significant gaps in coverage, especially in rural areas.
Medical confidentiality remained a major problem for cause-of-death registration in Belgium long into the twentieth century. Doctors faced a conflict between their duty to report causes of death and the confidential relation they had with their patients. In response, the National Health Inspection introduced anonymous cause-of-death-bulletins in the province of East Flanders in 1930. Antwerp followed in 1949, and the rest of the country in the 1950s. After the introduction of medical confidentiality, the share of deaths attributed to a vague cause such as ‘old age’ dropped. However, anonymity also made it more comfortable for doctors to admit ignorance. The share of deaths attributed to unknown and ill-defined causes other than old age remained stable.
Trends over Time
When Belgium started collecting data on causes of death, it also published overviews gathering the results of these efforts. Unfortunately, these overviews did not always contain the number of deaths attributed to ill-defined causes. Notably, there is no information on the number of deaths from ill-defined causes from 1870 to 1887. Deaths attributed to ‘old age’ were not listed separately from 1870 to 1902.
Deaths from 'Old Age and from other ill-defined or unknown causes in Belgium, ca. 1850 - 2000
In the early years of registration, a very large share of deaths had ill-defined causes. Together, deaths attributed to old age, symptoms or incomplete causes, or unknown causes made up over 40 per cent of deaths. This share dropped rapidly in the 1850s and 1860s, mainly because the proportion of deaths with unknown causes declined rapidly. More municipalities sent in reports with causes of deaths, and within municipalities, more deaths received a cause. Deaths attributed to ill-defined or unknown causes dropped from about 17.5 per cent in the early 1850s to about 6 per cent in the late 1860s. It fluctuated around that level for the rest of the twentieth century. The only exception were the two world wars. During the First World War, no national cause-of-death data were recorded, although some local records exist. In 1940 and 1944, the first and last year of the Second World War, over 15 per cent of deaths had ill-defined causes.
In the 1850s and 1860s, about 7.5 per cent of deaths were due to old age, rising to about 10 per cent in the first half of the 20th century. This share declined sharply in the 1950s, . From about 1960 onwards, deaths attributed to old age accounted for less than 2.5 per cent of all deaths in most years.
The early overviews of the 1850s and 1860s also included a broad category of deaths attributed to symptoms or incomplete causes. These made up between 15 and 17.5 per cent of deaths. Most of these were deaths attributed to convulsions, dropsy, apoplexy but also complications from childbirth. These specific causes were no longer listed separately after 1869. Some of them were later included among well-defined causes (such as childbirth-related deaths), others in the category of ill-defined causes.
By the late 1990s, the share of deaths attributed to all kinds of ill-defined causes (including old age and symptoms) reached an all-time low. It was less than . In recent years, this share is again on the rise, from about 5 per cent in 2010 to almost 8 per cent in 2022.
From strong regional differences to national uniformity
Deaths attributed to old age and other ill-defined and unknown causes were common in all parts of Belgium. Until about 1950, districts in the south and east of Belgium (provinces Luxemburg, Namur and Limburg) counted many deaths attributed to old age (sometimes 20-30 per cent of all deaths). Apart from that, there is no clear consistent geographical pattern. Districts with high levels of ill-defined deaths in one period could have much lower levels in later years, and vice versa.
The most striking pattern is that geographical differences became much smaller over time, especially after 1950. Before 1950, some districts had up to 25-30 per cent of their deaths attributed to old age, while others only had about 5 per cent. Likewise, some districts had up to 20 per cent of their deaths attributed to other ill-defined or unknown causes between 1890 and 1930, while others had almost none. By 1970, when medical confidentiality was applied all over Belgium, these shares were much lower, often to below 10 per cent, reflecting improvements in data collection and greater uniformity in registration practices.
Maps: deaths from ill-defined or unknown causes (excl. old age), 1890-1990 (percent of all deaths)
Unknown/ Ill Defined causes
Maps: deaths from ill-defined or unknown causes (excl. old age), 1890-1990 (percent of all deaths)
Old Age
Source: Mortality and population data, HISSTER and LOKSTAT-databases, Quetelet Center, Ghent University/STATBEL.
Who died of ‘unknown’ or ‘ill-defined’ causes?
Cause-of-death data from the city of Antwerp allow us to examine the individual characteristics of people whose deaths were recorded with ill-defined causes. In the nineteenth and twentieth century, four major groups stand out in Antwerp: Each group displays distinct characteristics.
Convulsions, the rapid, uncontrollable shaking of the body, can be a symptom of several illnesses, including infectious diseases causing diarrhea and dehydration. In the nineteenth century, convulsions were the largest group among the ill-defined causes, but they nearly disappeared in the twentieth century. These deaths occurred most often among infants (below 1 year old) or small children (1 to 4 year old). They were more common in poor and middle class families than among the elite. Over time, these social differences diminished, until deaths from convulsions all but disappeared after the middle of the twentieth century.
Deaths attributed to old age were, unsurprisingly, concentrated among the elderly. Social differences are also visible in this group, but with a remarkable shift in time. Before the mid-nineteenth century, deaths from old age were more common among the elite, a difference that disappeared in the second half of the nineteenth century Among the deaths from old age were in reality probably many cancer deaths, which initially struck the elite more often, since elite people smoked more tobacco and drank more alcohol. In contrast, around the beginning of the twentieth century, deaths from old age became more common among those not belonging to the elite. After the middle of the twentieth century, this social difference disappeared, as the number of deaths due to old age declined considerably. Deaths from old age were more often women, and more often people who were unmarried, divorced or widowed, although these differences were not always clear.
Deaths attributed to convulsions and old age clearly clustered in people of particular age groups: the young and the old. In contrast, deaths attributed to had less clear age profiles. Both tended to count more people of middle age. Deaths from unknown causes counted more people not belonging to the elite, whereas other ill-defined causes showed little or no social differentiation.
Conclusions
For more than a century and a half, Belgium has sought to record why its people die. Behind each entry in the registers lies a human life, often reduced to a few carefully chosen, or hastily written words. Knowing what is killing people allows the government to take targeted measures to improve the length and quality of life. Initially, these data were often vague or incomplete, and many deaths had ill-defined or unknown causes.
Major improvements occurred in the 1850s and 1860s and again in the 1950s, when the share of deaths attributed to old age or other ill-defined causes fell sharply. These improvements were driven by advances in medical knowledge, increased involvement of doctors and the introduction of anonymous reporting. As a result, a larger share of deaths received more precise causes and geographical differences in the quality of cause-of-death data have become much smaller.
Historically, ill-defined causes clustered at the beginning and end of life and more often involved people outside the elite. Despite all efforts to improve registration, in recent years a small and rising minority of deaths continue to have ill-defined causes.
Further Reading
Do you want to know more about our research? You can explore our published articles below.
- To be published
Literature
- Alter, George, and Ann Carmichael. 1996. ‘Studying Causes of Death in the Past: Problems and Models’. Historical Methods: A Journal of Quantitative and Interdisciplinary History 29 (2): 44–48.
- Hacha, Tom. 2025. ‘Van Arts Tot Ambtenaar. De Administratieve Praktijk Achter Het Antwerpse Doodsoorzakenregister En de Stedelijke Sterftestatistieken (1820-1946)’. HistoriANT: Jaarboek Voor Antwerpse Geschiedenis
- Hacha, Tom. Forthcoming. From Death to Data: The History of Cause-of-Death Registration in Belgium (1820–1960).
- Velle, Karel. 1985. ‘Statistiek en sociale politiek: de medische statistiek en het gezondheidsbeleid in België in de 19de eeuw.’Belgisch Tijdschrift voor Nieuwste Geschiedenis 16 (1–2): 213–42.
- Reid, Alice, Eilidh Garrett, Chris Dibben, and Lee Williamson. 2015. ‘“A Confession of Ignorance”: Deaths from Old Age and Deciphering Cause-of-Death Statistics in Scotland, 1855–1949’. The History of the Family 20 (3): 320–44.
Sources
- HISSTER database, Quetelet Center, Ghent University/STATBEL
- LOKSTAT database, Quetelet Center, Ghent University/STATBEL.
- Documents statistiques de la Belgique (1851-1867)
- Annuaire statistique de la Belgique (1868-1869)
- Mouvement de la population (State Archives Belgium, F 1160, agatha.arch.be) (1888-1959)
- Individual mortality data, STATBEL (1961-1997)
