Cancer trends · Population data · Belgium
Why Are Cancer Diagnoses Still Rising?
Twenty years of Belgian data show how population growth and ageing can conceal falling cancer rates—and where the increase persists after adjustment.
A rising cancer count is an unsettling statistic. With advances in prevention, earlier diagnosis and treatment, it can seem that the annual number should be moving in the opposite direction. When it keeps increasing, the natural suspicion is that progress has been overstated.
But that number answers a particular question: how many diagnoses were recorded in a population of a given size and age? It cannot, by itself, tell us whether the risk of diagnosis has increased at comparable ages. And it cannot tell us whether treatment is helping people live longer after diagnosis.
Belgium provides a useful way to examine the distinction. The Belgian Cancer Registry has national incidence data from 2004, and Statbel provides the population counts needed to adjust for demographic change. This analysis follows observed diagnoses through 2024, using the registry’s October 2026 extract. [1] [2]
First, separate the count from the rate
Belgium recorded 58,456 new cancer diagnoses in 2004 and 78,697 in 2024: an additional 20,241 diagnoses per year. These totals exclude non-melanoma skin cancer and may include more than one primary cancer in the same person.
Over the same period, the population grew from 10.40 million to 11.76 million, an increase of 13.2%. The proportion aged 65 or older rose from 17.1% to 20.0%. A larger population generates more diagnoses even if individual risks are unchanged. An older population can generate more again, because cancer incidence is strongly age-dependent. [1] [2]
Dividing diagnoses by population size gives the crude incidence rate. That still rose by 19.0%. To account for ageing, we need a further step: calculate the rate within each age group, then give those groups the same fixed weights in every year. This is age standardisation. It asks how recorded incidence would compare if the age distribution were held constant. [3]
| Population | Count | Rate |
|---|---|---|
| Men | +32.0% | −4.2% |
| Women | +37.9% | +13.1% |
| Both sexes | +34.6% | +5.4% |
Calculated from BCR diagnoses and Statbel populations using the European Standard Population 2013, with 18 age bands ending at 85+. Changes compare the endpoints; they are not fitted annual trends. Methods below.
The contrast is substantial. Male diagnosis counts rose by almost a third, while age-standardised incidence fell slightly. Female counts rose more, and an increase remained after age adjustment. The combined rate also increased, but by much less than the raw count.
Men
Counts +32.0% · age-adjusted rate −4.2%
Women
Counts +37.9% · age-adjusted rate +13.1%
How much of the increase is demographic?
Age standardisation shows how the rate changed. A separate calculation can divide the increase in diagnosis counts into three components: population growth, a change in age composition, and a change in diagnosis rates within age groups.
These components interact. For example, an increase in an age-specific rate produces more additional diagnoses in a larger population. To avoid assigning that interaction according to an arbitrary calculation order, this analysis averages all six possible orders of changing the three factors. The calculation is performed separately for men and women, then added together.
Population growth
41.9%A larger population, including the different growth rates of the male and female populations.
Older age composition
46.8%A greater concentration of people in ages with higher cancer incidence.
Changed diagnosis rates
11.3%The remaining change in recorded incidence within the age groups.
Demography therefore accounts for 88.7% of the increase in this decomposition. This percentage refers to the additional diagnoses between 2004 and 2024; it is not the proportion of all cancers “caused by ageing”.
A related calculation makes the scale easier to grasp. Apply the age- and sex-specific diagnosis rates from 2004 to the actual population of 2024, and the result is approximately 76,468 diagnoses—already close to the 78,697 observed. This fixed-rate comparison and the decomposition share out interactions differently, so their residuals are slightly different.
“Changed rates” needs careful interpretation
The 11.3% component combines changes in underlying disease occurrence, screening, diagnostic intensity, classification and registration. Changes in age within the grouped categories, especially 85+, can also contribute. It is an accounting residual, not a measurement of a particular biological cause.
Lung cancer makes the smoking question clearer
Male lung cancer diagnoses rose from 5,510 to 5,779, an increase of just 4.9%. Looking only at those counts, it would be easy to conclude that little had improved. Yet the male age-standardised lung incidence rate fell by 23.0%.
The female pattern was very different. Diagnoses rose from 1,554 to 3,902, and age-standardised incidence increased by 102.3%—approximately doubling even after the age adjustment.
Lung cancer: opposite trends after age adjustment
These opposing trends are consistent with the different historical smoking patterns of men and women and the delayed effects of tobacco exposure. Today’s smoking prevalence does not describe the lifetime exposure of people diagnosed today. The European Commission’s cancer information service specifically notes that female lung incidence can continue rising despite lower current smoking prevalence because of smoking in previous decades. [4]
This does not calculate how many Belgian cancers were prevented by smoking reduction. It does show why an aggregate count is a poor way to test whether prevention has helped. It also shows why combining the sexes can hide important differences.
“Cancer” contains several different trajectories
The same count–rate distinction appears elsewhere. Colorectal diagnosis counts increased slightly in both sexes, while age-standardised incidence fell appreciably. Prostate diagnoses increased despite a lower age-standardised rate at the end of the period. Female breast incidence changed much less after age adjustment than the count suggests.
| Cancer / sex | Count | Rate |
|---|---|---|
| Lung · men | +4.9% | −23.0% |
| Lung · women | +151.1% | +102.3% |
| Colorectal · men | +2.3% | −27.9% |
| Colorectal · women | +3.7% | −16.5% |
| Breast · women | +25.2% | +3.3% |
| Prostate · men | +28.8% | −9.1% |
Source: the same October 2026 BCR extract. Endpoint changes can conceal intervening rises and falls; the prostate series, for example, did not decline continuously.
Detection also changes the shape of the curves. The Flemish colorectal screening programme began in October 2013. A BCR-coauthored study of residents aged 50–79 found an initial incidence peak in 2014, particularly in early-stage cancers, followed by a decline. The timing of age-specific peaks followed the phased screening invitations. [5]
That is a reminder that a temporary increase in recorded incidence can accompany efforts to find existing disease earlier. The national series alone cannot quantify the screening programme’s contribution.
A twenty-year result is not a permanent rule
The demographic share depends on the years compared. It is 88.7% for 2004–2024, 83.5% if the starting year is moved to 2005, and 78.0% for 2019–2024. Demography remains the larger component in these comparisons, but “89%” should not become a universal constant.
Nor does the full-period male decline mean that male incidence keeps falling. The all-cancer male age-standardised rate rose by 1.5% between 2019 and 2024. The dip in 2020 also requires care: the registry documented substantial disruption to cancer diagnosis during the pandemic. A drop in diagnoses during that disruption cannot be read as an abrupt reduction in underlying cancer risk. [6]
What these numbers say about progress
Belgium’s growing cancer count represents a real burden: more people need investigation, treatment and follow-up. Demographic adjustment does not make those needs disappear. It changes the explanation for why they increased.
At the same time, age-adjusted improvements in several cancer categories coexist with worsening trends in others. The combined age-standardised rate rose by 5.4%, and the female rate rose by 13.1%. There is room to recognise progress while taking those increases seriously.
Treatment progress needs a different assessment. Incidence counts diagnoses; survival describes what happens after them. A treatment that extends life does not erase the diagnosis already recorded. Longer survival can also increase prevalence—the number of people living after a cancer diagnosis—without implying the same increase in new diagnoses. This analysis measures neither overall survival nor cure. [1] [7]
The useful question is which measure changed, for whom, and over what period. Belgium’s data show a rising healthcare burden, a large demographic contribution, and distinct trends within individual cancers. Those findings provide a firmer basis for discussing progress than the annual diagnosis count alone.
Methods, scope and reproducibility
This is an original descriptive calculation using public BCR and Statbel data, not an official BCR estimate of demographic contributions or a causal study. Observations cover Belgium from 2004 through 2024; the module’s marked 2025 estimates were excluded. A single October 2026 extract was used because historical registrations can be revised.
“All cancers” includes ICD-10 C00–C43 and C45–C97 plus MDS/MPN, excluding C44 non-melanoma skin cancer. Counts refer to diagnoses, not necessarily unique individuals. Lung cancer is C34. Population denominators are Statbel legal residents on 1 January of each incidence year.
Rates were standardised to the European Standard Population 2013 in 18 groups: 0–4, 5–9, …, 80–84 and 85+. All 42 recalculated male/female all-cancer rates reproduced the live registry module’s two-decimal values. Some PDF fact sheets separate the oldest ages further and therefore report slightly different rates. The grouping here remains consistent across years. [3]
Both-sex rates were calculated from pooled age-specific counts and populations, rather than adding male and female rates. They can still reflect changes in sex composition within age groups. The three-factor decomposition averages all six factor orders (Shapley allocation), separately by sex, then sums the contributions. Its population-growth component includes the different male and female population growth rates.
An independently implemented calculation reproduced the decomposition, and an offline rerun reproduced the annual results exactly. The main analysis did not estimate confidence intervals. The choice of endpoints, broad age groups, detection changes and registration practices limit causal interpretation.
Analysis checked 21 October 2026. The accompanying local analysis folder contains the annual CSVs, archived source tables, scripts, sensitivity comparisons and validation records.
Sources and references
- Belgian Cancer Registry. Drawing module and methodological notes; interactive data tables. October 2026 extract. Primary source for the diagnosis counts and site-specific rates used here.
- Statbel. Population by age and sex. National legal resident populations on 1 January, used as the incidence denominators.
- Belgian Cancer Registry. Cancer incidence projections: methodology (2026), section 3.3. Source for the start-of-year denominator and 18-group ESR2013 convention; projected case numbers were not used in this article.
- European Commission / European Cancer Information System. European cancer data confirm lung cancer in women as a major public health issue (2026). Context on historical tobacco exposure and delayed incidence trends; not a Belgium-specific causal estimate.
- Tran and colleagues. Impact of a six-year existing screening programme using the faecal immunochemical test in Flanders. International Journal of Environmental Research and Public Health 20, 1654 (2023). Population study of colorectal incidence and stage patterns in Flanders.
- Belgian Cancer Registry. Cancer in Belgium 2020. Context on disrupted diagnosis during COVID; historical publication totals may differ from the revised extract used here.
- Belgian Cancer Registry. Cancer fact sheet: all cancers, 2024. Separate presentations of incidence, survival and prevalence.
%20(3).png)
No comments:
Post a Comment