Robert L. Stout

NT-proBNP Predicts All-Cause Mortality in a Population of Insurance Applicants, Follow-up Analysis and Further Observations

OBJECTIVE - Further refine the independent value of NT-proBNP, accounting for the impact of other test results, in predicting all-cause mortality for individual life insurance applicants with and without heart disease. METHOD - Using the Social Security Death Master File and multivariate analysis, relative mortality was determined for 245,322 life insurance applicants ages 50 to 89 tested for NT-proBNP (almost all based on age and policy amount) along with other laboratory tests and measurement of blood pressure and BMI. RESULTS - NT-proBNP values ≤75 pg/mL included the majority of applicants denying heart disease and had the lowest risk, while values >500 pg/mL for females and >300 pg/mL for males had very high relative risk. Those admitting to heart disease had a higher mortality risk for each band of NT-proBNP relative to those denying heart disease but had a similar and equally predictive risk curve. CONCLUSION - NT-proBNP is a strong independent predictor of all-cause mortality in the absence or presence of known heart disease but the range of values associated with increased risk varies by sex.

2014 CRL Build Study of Life Insurance Applicants.

Objective .- Determine the impact of build on insurance applicant mortality accounting for smoking, laboratory test values and blood pressure. Method .- The study consisted of 2,051,370 applicants tested at Clinical Reference Laboratory between 1993 and 2007 with build and cotinine measurements available whose body mass index (BMI) was between 15 and 47. Vital status was determined as of September, 2011 by the Social Security Death Master File. Excluded from the primary study were applicants with HbA1c values ≥6.5%, systolic BP ≥141 mmHg, albumin values ≤3.3 g/dL or total cholesterol values ≤130 mg/dL. Relative mortality was determined by Cox regression analysis for bands of BMI split by age, sex and smoking status (urine cotinine positive). Results .- A majority of applicants had BMI >24 (overweight or obese by WHO criteria). After the exclusions noted above, relative mortality does not increase by >34% unless BMI is <20 (<18 for female non-smokers age 18 to 59) or BMI is >34. BMI values in the range of 22 to 24 and 25 to 29, overall, had similar and the lowest relative risks. For most nonsmokers, risk was lowest in the lower of these two BMI bands but for smokers (and non-smoking males age 60 to 89) risk was lowest in the higher BMI band. Additional analysis showed limited reduction in relative risk by accounting for all laboratory test values as well as continuing the exclusions. Eliminating the exclusions resulted in only a modest increase in relative risk because the mortality rate of the reference band increased as well. Conclusion .- After excluding elevated HbA1c and blood pressure (associated with high BMI) and low albumin and cholesterol (associated with low BMI) which are usually evaluated separately, mortality varies by a limited degree for BMI 20 to 34. Accounting for the mortality impact of other test values, in addition to the exclusions noted, reduced mortality associated with high BMI to a limited extent, but had little impact on mortality associated with low BMI.

Hemoglobin Screening Independently Predicts All-Cause Mortality.

Objective .- Determine if the addition of hemoglobin testing improves risk prediction for life insurance applicants. Method .- Hemoglobin results for insurance applicants tested from 1993 to 2007, with vital status determined by Social Security Death Master File follow-up in 2011, were analyzed by age and sex with and without accounting for the contribution of other test results. Results .- Hemoglobin values ≤12.0 g/dL (and possibly ≤13.0 g/dL) in females age 50+ (but not age <50) and hemoglobin values ≤13.0 g/dL in all males are associated with progressively increasing mortality risk independent of the contribution of other test values. Increased risk is also noted for hemoglobin values >15.0 g/dL (and possibly >14.0 g/dL) for all females and for hemoglobin values >16.0 g/dL for males. Conclusion .- Hemoglobin testing can add additional independent risk assessment to that obtained from other laboratory testing, BP and build in this relatively healthy insurance applicant population. Multiple studies support this finding at older ages, but data (and the prevalence of diseases impacting hemoglobin levels) are limited at younger ages.

2014 CRL Blood Pressure Study of Life Insurance Applicants.

Objective .- Define the relative mortality risk by systolic (SBP) and diastolic blood pressure (DBP) in a relatively healthy cohort split by age and sex with adjustment for smoking status, other findings and admitted heart disease history. Method .- Blood pressure (BP in mm Hg), build, laboratory studies and limited medical history are collected when people apply for individual life insurance. Information on 2,472,706 applicants tested by Clinical Reference Laboratory from 1993 to 2007 was utilized with follow-up for vital status using the September 2011 Social Security Death Master File identifying 31,033 deaths. Data was analyzed by SBP and DBP split by age and sex accounting for smoking and for BMI, urine protein/creatinine ratio and history of heart disease in a Cox multivariate survival analysis. Separate analysis by admitted hypertension history was also conducted. Results are presented by SBP and DBP for 4 age-sex groups with and without added covariates beyond age and smoking status. Results .- Relative mortality progressively increased by SBP level from the 90 to 119 band (down to 80 in younger women) upward with little additional impact by DBP. Addition of covariates beyond age and smoking resulted in a 5% to 10% reduction in relative risk. Although high DBP had limited impact, a pulse pressure/SBP ratio >½ identified 1% of applicants at high mortality risk, with little difference in risk for ratios ≤½. Hypertension history with current BP control was associated with a 10% to 25% increase in relative mortality risk as compared to those with similar BP but no such history. Conclusion .- Increasing SBP is closely associated with increasing relative mortality, starting from the lowest SBP. Increasing DBP has little additional impact, but a pulse pressure/SBP ratio >½ is a potent marker of increased risk as well. Accounting for build and other laboratory findings reduces risk modestly. A history of hypertension with current control increases risk.