I have been struck by several news stories over the last few months: not so much by the facts reported but by the difficulty in understanding what lies behind them. They led me to wonder whether a less medically informed reader might be confused or draw a wrong conclusion. How do we decide? Here are two examples.

1. Pollution from cars kills more people in Britain than traffic accidents

This is based on taking mathematical models of air pollution in Britain and relating these concentrations to estimates of the risks of death derived from studies of populations in cities in relation to changes in levels of pollutants. The data on which these latter are based are reasonably robust in that they have been repeated in many different studies and come up with similar risks – the more particulate matter or nitrogen oxides we are exposed to, the greater our risk of dying from a heart attack or lung disease.

However, the deaths recorded are almost exclusively among older people who already have heart or lung disease, though they may not necessarily know it. Thus the 3,000+ deaths each year attributable to vehicle-derived air pollution are an estimate and represent deaths occurring over an unknown period (best estimated in months on average) before they would have been expected in people who may not have been very far from death in any case. In contrast, the 3,000 or so annual deaths from car accidents are predominantly among young people and represent a precise number of individuals, most of whom would have been in good health.

This is an example of like not being compared with like in order to make a point, the point being that air pollution is a public health problem that needs to be addressed more urgently. The more appropriate comparison would have been an estimate of the life years lost from the two causes, but the case would then have been less strong because car accidents are the major cause of death in young people in Britain. Even the well-intentioned misuse statistics and, were I a road safety campaigner, I should point out that in 2010 over 23,000 road users were killed or seriously injured, of whom 2,500 were under the age of 16.

If the problem were a matter of priorities for a politician who has to make judgements on spending our money, one can see that the advocates of lower speed limits and better control of traffic would win by arguments involving death and maiming of children when opposed by partial contributions to deaths of elderly people from air pollution. But the air pollution advocates would bring in the associations of air pollution with asthma attacks in children, implying that asthma in children is caused by pollution. However, it is not a matter of priorities, since we are talking about a single cause of two separate consequences; vehicles cause both injury and, indirectly, illness through pollution, so appropriate control of them takes care of both. Even better, black carbon, the most important vehicle emission, is now known to be a major influence on climate change. The more we reflect on these matters when we purchase a vehicle or drive it, the better.

2. Vitamin D prevents dementia (and lots of other things)

When I went to medical school, back in the 1950s, vitamin D was known to be necessary to enable our bones to calcify properly: if we did not eat enough we got rickets, with malformed limbs and easily broken bones. Rickets was already rare in Britain, since we were all recommended to swallow cod liver oil as children, and I only saw its consequences in older folk. Now it is coming back, particularly among immigrant families and there has been a recent sad case in which parents have apparently been accused of injuring their child because of a failure by doctors to recognise the vitamin deficiency causing the broken bones.

In fact, remarkably low levels of vitamin D are found in people living in Scotland and in northern latitudes generally. While most levels are sufficient to prevent rickets and the equivalent condition in adults, osteomalacia, it has been suggested that low intake may contribute to the causation of several other conditions such as prostate cancer, asthma and now dementia. These suggestions come from epidemiological studies that show an association between the levels in the blood and the risks of these conditions, implying that insufficient vitamin D (or some other substance closely associated with it) is increasing our susceptibility to them. The important implication is that a general increase in the amounts in our blood would lead to a reduction in the incidence of these diseases in the population and that if we go back to taking cod liver oil or pills with vitamin D our personal risks would reduce, but this has not been proven. Causation and statistical association are not the same thing.

Vitamin D is a unique vitamin in that we can make it ourselves by the action of sunlight on cells in our skin. The story is intriguing. Our remote ape-like ancestors had white skins and a hairy body. As they evolved as bipeds, with the greater exercise capacity for hunting that this allowed on the plains of Africa, they needed more efficient ways of losing heat and started to lose their hair and develop more sweat glands. Since this is likely to have led to serious problems from sun exposure, including skin burning and heat exhaustion, further evolution produced an advantage to those with the protection of darker skins and a greater density of sweat glands.

Those who then migrated north found themselves in an environment where in winter the sun’s ultraviolent radiation was not always sufficient to stimulate the production of enough vitamin D, especially through a dark skin. We became paler again, and only by eating enough in the diet could we keep rickets at bay. The best source is oily fish, though it is also present in nuts and meat, and this may explain in part our ability to survive in northern latitudes, the Nordic, Lapps, Inuit and even the Scots happily having had available abundant fish for our diet.

Does vitamin D prevent dementia, asthma and various cancers, though? It certainly is being found to have multiple interesting biochemical properties. It would be good to know, and I have spent a lot of my last 20 years trying to answer a part of this question. It is not simple. As far as I know there are three ways: studying people with the condition in question and comparing them with others who don’t have it; studying well people and following them forward in time while observing their diets; or feeding a group of people the vitamin over a long time and seeing what happens.

The numbers of people required for these sorts of studies and the time and cost required to carry them out usually make only the first of these options viable, and this unfortunately leads to conclusions that are relatively insecure, since any associations found, even if strong, may be caused by some related factor that has not been taken account of. For example, people who have low vitamin D levels tend to be poorer than those with high levels and are exposed to a generally poorer diet and more of other factors like pollution that might increase their risks of disease. It is only when a number of large studies from different places come up with similar results that the science begins to look solid and it becomes easier to get grants to do the clinching intervention studies. We are at that point now with respect to asthma but far from it with respect to dementia and cancers.

What therefore should we make of claims about environmental causes of disease in the media? Bear in mind the difficulties in finding clear-cut answers to such questions while recognising that behind the claim there may be a grain of truth. However, if you changed your diet every time you read of some new food scare or wonder foodstuff you would never maintain the good balanced long-term diet that is most likely to be of the greatest health benefit to the greatest number. So, eat as varied and balanced a diet as you can afford and if you live in Scotland evolutionary reasoning suggests that there may be a general health benefit in taking a small supplement of cod liver oil or similar if you don’t eat oily fish (but no one has proved it yet).

Professor Anthony Seaton is an emeritus professor in the school of medicine and dentistry at the University of Aberdeen