Who am I?
I am Robert Hoenselaar. A 44-year old research journalist and dietitian.
In the past, I published in the scientific literature on the relation between saturated fat intake and cardiovascular disease. Showing that the scientific evidence on this topic was misrepresented to the general public. Which, it still is!
A lot of unused scientific data, readily available to the general public.
For several years, I am aware that there is a large amount of health data available. E.g., throught the World Health Organization, the World Bank and the Food and Agriculture Organization of the United Natations. The data give us yearly numbers for several health indicators and outcomes for every country in the world with available data. The health indicators include different foods consumed, health care, income, clean drinking water, etc. The health outcomes include the prevalence of diseases or the mortality rates from them. It is possible to correlate/link these numbers to each other. What can give additional insights in where we are as a human race.
But this is rarely done. We see some publications where one or two health indicators are linked to each other or to a certain health outcome. But the data seems cherry picked and is often poorly controlled for confounders.
Untill now, I did not really dive into this data. But the time has come to get it out to everybody who has an interest in these correlations.
How should we interprete these correlations?
The correlations are what we call “ecological studies”. Many people think that results from these studies cannot give us any usefull insights. As did I, in the past.
But this would be a false conclusion. We can correlate everything to each other. We just should be careful interpreting the meaning of these correlations.
Ecological studies cannot truly give us evidence that something we do or consume increases or decreases risk of a certain disease/type of death. But, it can show us that perceived assumptions about things we think may be false. When a health indicator is supposed to be (one of the) main determinant(s) of a health outcome, we should always be able to see this back in an ecological study.
For example, we are told that vaccinations are the cause of the decrease in an infectious disease. Then, we should always see that higher vaccination rates decrease the risk of catching this disease over time. Especially, when we have a fair number of comparisons. There is no way around it.
For another example, take a look at the next graph which I created years ago. It shows the correlation between per capita availability (consumption) of animal fats and the DALY’s for cardiovascular diseases. DALY stands for “disability-adjusted life years” or: burden of disease for cardiovascular diseases. The higher a dot is in the graph, the higher the burden of disease:
This correlation does not prove that higher availability of animal fats increases or decreases the DALY’s for cardiovascular diseases over time. But, if we assume high consumption of animal fats increases cardiovascular disease risk/mortality, then why do we see the highest burden of these diseases in the countries with the lowest past consumption of these fats? Let’s assume, you think animal fats increase cardiovascular disease risk/mortality. Then this graph shows you there need to be other factors contributing to the disease that are far more relevant than aminal fats.
Here you have 2 examples where ecological studies can somewhat “(dis)prove” a certain idea. Which can give us very useful insights.