OBJECTIVE: To determine the effects of weight variability on cardiovascular
DESIGN: Longitudinal study (followed from 1990 to 1998),
SUBJECTS: Five-hundred and eighty nonsmoking Japanese male office workers a
ged 25-49 y.
MEASUREMENTS: Baseline levels and slopes of seven selected cardiovascular r
isk factors (systolic and diastolic blood pressure, total and high-density
lipoprotein cholesterol, triglyceride, uric acid and hemoglobin A(1c)) and
body mass index (BMI) variables (baseline BMI, BMI-slope, the root mean squ
are error (BMI-RMSE), and the BMI-slope x BMI-RMSE interaction).
RESULTS: From the multiple regression analyses, the slopes of the six cardi
ovascular risk factors, except hemoglobin A(1c), were most strongly related
to the baseline level of each cardiovascular risk factor and BMI-slope, Ne
ither BMI-RMSE nor the interaction of BMI-RMSE with BMI-slope was related t
o these cardiovascular risk factor slopes. As for hemoglobin A(1c), BMI-slo
pe and the interaction of BMI-RMSE with BMI-slope were not significantly re
lated to hemoglobin A(1c) slope but baseline BMI and BMI-RMSE were. Further
more, hemoglobin A(1c) slope values were significantly higher among those w
ho cycled at >4.0 kg than among those who did not.
CONCLUSION: The only cardiovascular risk factor associated with BMI variabi
lity was hemoglobin A(1c). Weight variability had less or little impact on
cardiovascular risk factors compared with BMI-slope and baseline BMI.