Maria Glymour and her colleagues at the Harvard School of Public Health conducted their study of the Stroke Belt by using data from the 1980, 1990, and 2000 censuses and mortality data from the National Center for Health Statistics for those same years. One of the strengths of their study was the use of “hard” mortality data, which is based on official death certificates.
They defined the Stroke Belt as being the states of North Carolina, South Carolina, Georgia, Alabama, Mississippi, Tennessee, and Arkansas. These are the seven states included in the federal government’sStroke Belt elimination project.
Using the census and mortality data, the researchers created four “exposure” categories, with the exposure being either birth in the Stroke Belt or residence in a Stroke Belt state as an adult. The exposure groups stack up like this:
Born in the Stroke Belt, adulthood in the Stroke Belt (double exposure)
Born in the Stroke Belt, adulthood out of the Stroke Belt (out-migrants)
Not born in the Stroke Belt, adulthood in the Stroke Belt (in-migrants)
Not born in the Stroke Belt, adulthood out of the Stroke Belt (no exposure)
Crude rates
Once the exposure groups were set, the researchers then calculated the number of stroke deaths per 100,000 people — the crude stroke mortality rate — for each of them. Here are the results:
1980 | 1990 | 2000 | |
Stroke Belt birth, Stroke Belt adulthood (doubly exposed) | 122 | 83 | 74 |
Stroke Belt birth, non-Stroke Belt adulthood (out-migrants) | 115 | 89 | 88 |
Non-Stroke Belt birth, Stroke Belt adulthood (in-migrants) | 63 | 43 | 39 |
Non-Stroke Belt birth, non-Stroke Belt adulthood (nonexposed) | 78 | 52 | 47 |
Here is the same information just for whites:
1980 | 1990 | 2000 | |
Stroke Belt birth, Stroke Belt adulthood (doubly exposed) | 97 | 68 | 66 |
Stroke Belt birth, non-Stroke Belt adulthood (out-migrants) | 86 | 65 | 66 |
Non-Stroke Belt birth, Stroke Belt adulthood (in-migrants) | 60 | 41 | 40 |
Non-Stroke Belt birth, non-Stroke Belt adulthood (nonexposed) | 75 | 51 | 46 |
And for blacks:
1980 | 1990 | 2000 | |
Stroke Belt birth, Stroke Belt adulthood (doubly exposed) | 211 | 128 | 96 |
Stroke Belt birth, non-Stroke Belt adulthood (out-migrants) | 146 | 115 | 116 |
Non-Stroke Belt birth, Stroke Belt adulthood (in-migrants) | 112 | 66 | 37 |
Non-Stroke Belt birth, non-Stroke Belt adulthood (nonexposed) | 112 | 68 | 58 |
Note that in almost every category, the stroke mortality rate is higher for blacks. The lone exceptions are the so-called in-migrants — people who moved to the Stroke Belt after living elsewhere.
Glymour and her colleagues pointed out that excess stroke mortality from Stroke Belt exposure, either at birth or as an adult, declined from 1980 to 2000 in every category. The biggest decline was among blacks who were “doubly exposed” (from a rate of 211 stroke deaths per 100,000 people in 1980 to a rate of 96 stroke deaths per 100,000 in 2000).
More refined numbers
But these are crude mortality rates that don’t take into account sex and age. The researchers made those adjustments. They also expressed the stroke risk as an odds ratio relative to the nonexposed group. In simplified terms, that means calculating how much greater the stroke risk was for people who had spent time in the Stroke Belt (birth, adulthood, or both) compared with the stroke risk for those who were neither born nor lived there as adults — the nonexposed in the charts above.
So, for example, in the chart below, 45% means a 45% greater chance of having a stroke than the nonexposed group. Here are those results for whites:
1980 | 1990 | 2000 | |
Stroke Belt birth, Stroke Belt adulthood (doubly exposed) | 45% | 29% | 34% |
Stroke Belt birth, non-Stroke Belt adulthood out (out-migrants) | 31% | 20% | 20% |
Non-Stroke Belt birth, Stroke Belt adulthood (in-migrants) | 8% | 7% | 11% |
For blacks:
1980 | 1990 | 2000 | |
Stroke Belt birth, Stroke Belt adulthood (doubly exposed) | 55% | 47% | 34% |
Stroke Belt birth, non-Stroke Belt adulthood out (out-migrants) | 20% | 11% | 9% |
Non-Stroke Belt birth, Stroke Belt adulthood (in-migrants) | 13% | 49% | 1% |
The highest group for whites and blacks are the doubly exposed, although by 2000, there is no racial difference: the stroke death rate is 34% higher for both whites and blacks. It’s also interesting that the stroke risk for out-migrant blacks is lower than it is for out-migrant whites in all three years.
Glymour and her colleagues acknowledge that their study cannot answer perhaps the most interesting question: Why is there a Stroke Belt in the first place? Genetic explanations don’t seem to work, because the Stroke Belt effect is true for both blacks and whites, although it’s usually less pronounced in whites. Behaviors or perhaps environmental toxins of some sort are more likely explanations, but nobody knows for sure.
This study does raise new questions about timing and whether there might be something about being born and spending your early childhood (place of birth and place of early childhood are not the same but census information shows that they are closely related) in the American Southeast that leads to a greater chance of having a fatal stroke later in life.
Posts
