Renal function assessed by two different formulas and incidence of myocardial infarction and death in middle-aged men and women.

BACKGROUND: Chronic kidney disease predicts mortality in the general population, but less is known about the association with incidence of first myocardial infarction. We evaluated glomerular filtration rates (GFR) estimated by the Modification of Diet in Renal Disease study (GFR-MDRD) equation and the Mayo formula (GFR-Mayo) as predictors of myocardial infarction and death. METHODS: In 571 353 Swedish men and women, undergoing health controls, with mean age 45 years, and no previous myocardial infarction, hazard ratios were calculated to assess the association between renal function and incidence of myocardial infarction and all-cause mortality, respectively. Glomerular filtration rate 60-90, 30-60 and <30 mL per minute per 1.73 m(2), was defined as mildly, moderately and severely decreased GFR, respectively. RESULTS: There were 19 510 myocardial infarctions and 56 367 deaths during 11.6 years of follow-up. Hazard ratios (and 95% confidence intervals) for myocardial infarction, using GFR-Mayo were 1.11 (1.06-1.16) for mildly, 1.32 (1.18-1.48) for moderately and 2.54 (1.90-3.40) for severely decreased GFR. The corresponding figures for GFR-MDRD were 1.01 (0.96-1.05), 1.23 (1.14-1.32) and 2.49 (1.85-3.35). Mortality was increased at all levels of reduced GFR-Mayo and at moderately or severely decreased GFR-MDRD. CONCLUSIONS: Already mildly decreased GFR increase the risk of myocardial infarction and death in the general population. The association with adverse outcomes is stronger when GFR-Mayo rather than GFR-MDRD is used to assess renal function.

Profile comparisons of physical growth for monozygotic and dizygotic twin pairs.

Three different methods of analysing similarity of physical growth curves within MZ and DZ twin pairs are compared. The concordance of the curves for each pair can be measured by either a multiple regression coefficient (R2) or the square root of the sum of squared differences between corresponding deviation points (D). The latter seems to be a more appropriate measure of profile contour similarity, mainly because the distance between curves is allowed for. An intra-class correlation based on a longitudinal analysis of variance seems to be a useful measure of average profile similarity for comparing groups of individuals. The distributions of the D measures, calculated for all MZ and DZ twin pairs, show MZ pairs to have significantly smaller D-values for both height and weight, on average, than DZ pairs. The intra-class correlations also show a higher concordance in developmental profile for MZ in comparison to DZ.