Almost every day, it seems, there's a story saying that either (1) Factor X is associated with increased diabetes risks or (2) people with diabetes are at increased risk of Condition Y.
Today's Risk Factor of the Day is arsenic. A study found that people with diabetes have higher levels of arsenic in their urine than nondiabetics and people with higher arsenic levels are more likely to get diabetes.
Of course, "is associated with" doesn't mean the same as "causes." Rich people drink more expensive wines than poor people, so drinking expensive wines is associated with wealth. Does this mean if I drank a lot of expensive wine I'd become wealthy?
Of course not. But the headline writers tend to ignore such distinctions. Maybe people with diabetes or people already at risk for diabetes process arsenic differently. Or maybe in parts of the country where there's more arsenic in the water, other lifestyle conditions are unhealthy.
Several other bloggers, apparently as fed up with this news story type of reporting as I am, have recently written about the difference between causation and association. Unfortunately, I suspect the news reporters aren't listening.
Other recent news stories about factors that are associated with increased diabetes risks concern PCBs, sugary drinks, depression, gum disease, having obese diabetic mothers while in utero, a particular liver protein, living in particular neighborhoods, antipsychotic drugs, lack of sleep, a high glycemic index diet, having had gestational diabetes, higher heart rates, waist circumference, C-reactive protein levels, gum disease, smoking, and fruit juices.
Factors that are associated with lower rates include spices, exercise, getting enough vitamin D, omega-3 fatty acids, walking to work, and vitamin C.
Each of these factors may, indeed, have a stasticially significant association with getting diabetes, not getting diabetes, or improving diabetic control. But in many cases, the association is not great, and the numbers are manipulated to make them seem more newsworthy.
For example, let's say that 0.01% of people in a study group who had a particular behavior got diabetes in the next 5 years, and 0.005% of those who didn't have that behavior didn't get diabetes. That means the risk of getting diabetes was doubled by having that particular behavior.
But 0.01% means 1 person in 10,000, which is a pretty low risk even though it's twice the risk of 0.5 people in 10,000 in the other group. Give us a break!
Another type of story that seems to appear every day is the "Diabetics are at increased risk of condition Z."
Recent stories of this type include being at higher risk of heart disease, cancer, cognitive problems including Alzheimer's disease, TB, hospitalization from pneumonia, impaired ankle function, male infertility, having a poor outcome after a heart attack, and having a baby with birth defects.
