Visit-to-Visit Blood Pressure Variability as a Risk Factor for All-Cause Mortality, Cardiovascular Mortality, and Major Adverse Cardiovascular Events Among American Indians: the Strong Heart Study

We screened 3,501 eligible participants in the Strong Heart Study and 2,346 and eligible participants in the Strong Heart Family Study. From the Strong Heart Study we excluded 1,044 participants because their visits were urgent (hospitalizations, emergency room visits, urgent care visits, ambulance trips, or pregnancy clinical visits), leaving 2,457 eligible participants. From these we further excluded 517 participants with insufficient data on blood pressure, 262 because they had fewer than 8 measurements overall and 255 because they had fewer than 8 measurements in the first 5 years of data collection. These exclusions left a total of 1,940 participants from the Strong Heart Study. From the 2,346 participants selected from the Strong Heart Family Study, we excluded 136 because of urgent blood pressure measurements only, leaving 2,210 eligible participants. From these, we further excluded 798 participants because of insufficient data on blood pressure: 223 had fewer than 8 blood pressure measurements overall, 574 had fewer than 8 blood pressure measurements in the first 5 years of data collection, and 1 had 8 identical blood pressure measurements. These exclusions yielded a total of 1,412 participants from the Strong Heart Family Study. The final combined number of participants from both studies was 3,352.

Figure 1.
CONSORT Diagram for participants with 8 nonurgent blood pressure measurements in the first 5 years of available medical records from the National Data Warehouse of the Indian Health Service. Study participants were drawn from the Strong Heart Study (26) and the Strong Heart Family Study (27) of American Indians residing in Arizona, Oklahoma, North Dakota, and South Dakota (1997–2003). [A text version of this figure is available.]
We screened 3,501 eligible participants in the Strong Heart Study and 2,346 and eligible participants in the Strong Heart Family Study. From the Strong Heart Study we excluded 1,044 participants because their visits were urgent (hospitalizations, emergency room visits, urgent care visits, ambulance trips, or pregnancy clinical visits), leaving 2,457 eligible participants. From these we further excluded 517 participants with insufficient data on blood pressure, 262 because they had fewer than 8 measurements overall and 255 because they had fewer than 8 measurements in the first 5 years of data collection. These exclusions left a total of 1,940 participants from the Strong Heart Study. From the 2,346 participants selected from the Strong Heart Family Study, we excluded 136 because of urgent blood pressure measurements only, leaving 2,210 eligible participants. From these, we further excluded 798 participants because of insufficient data on blood pressure: 223 had fewer than 8 blood pressure measurements overall, 574 had fewer than 8 blood pressure measurements in the first 5 years of data collection, and 1 had 8 identical blood pressure measurements. These exclusions yielded a total of 1,412 participants from the Strong Heart Family Study. The final combined number of participants from both studies was 3,352

Figure 2.
Box plots of the quartiles of the SD of systolic blood pressure and diastolic blood pressure for 8 nonurgent blood pressure measurements taken at clinic visits (urgent visits were defined as hospitalizations, emergency department visits, urgent care visits, ambulance trips, and pregnancy clinical visits and excluded) within a 5-year period closest to the dates of the Strong Heart Study (26) examination 3 (1997–1999) or the Strong Heart Family Study (27) examination 1 (2001–2003) of American Indians residing in in Arizona, Oklahoma, North Dakota, and South Dakota.

For SD of systolic blood pressure, there is clear and early separation of the 4 survival curves with lower survival with each increasing quartile (P <.001). For SD of diastolic blood pressure, survival curves remain close together with curves for quartiles 1 and 2 showing moderate separation from curves for quartiles 3 and 4 (P <.001).

Figure 3.
Survival curves for all-cause mortality, by quartiles, of the SD of systolic blood pressure (A) and diastolic blood pressure (B) for 8 blood pressure measurements taken at nonurgent clinic visits (urgent visits were defined as hospitalizations, emergency department visits, urgent care visits, ambulance trips, and pregnancy clinical visits and excluded) within 5 years, over 23 years of follow-up (1999–2022) for American Indians residing in Arizona, Oklahoma, North Dakota, and South Dakota.