RESUMO
Morbidity and premature mortality from noncommunicable diseases can be largely prevented by adopting a healthy lifestyle at the earliest possible age. However, tools designed for the early identification of those at risk among young adults are lacking. We developed and validated a multivariable model for the prediction of life expectancy, allowing the early identification of apparently healthy adults at risk of lifestyle-related diseases. We used a cross-sectional approach to calculate life expectancy using data from 38,481 participants of the National Health and Nutrition Examination Survey (1999-2014), aged ≥20 years. A multivariable logistic model was used to quantify the impact of risk factors on mortality. The model included the following lifestyle-related mortality risk factors as predictors: smoking, diet, physical activity, and body mass index. The presence of the following chronic diseases was considered: diabetes, arrhythmia, coronary artery disease, myocardial infarction, stroke, and malignant neoplasms. The model showed a good predictive ability; the area under the receiver operating characteristic curve measure was 0.846 (95% uncertainty interval 0.838-0.859). Life expectancy was determined using the life table method and the period life tables for the US population as the baseline. The results of this model underscore the importance of lifestyle-related risk factors in life expectancy. The difference between life expectancy for 30-year-old individuals with lifestyle characteristics ranked in 90% and 10% of their gender and age groups was 23 years for males and 18 years for females, whereas in 75% and 25%, it was 14 years for males and 10 years for females. In addition to early risk identification, the model estimates the deferred effect of lifestyle and the impact of lifestyle changes on life expectancy. Thus, it can be used in early prevention to demonstrate the potential risks and benefits of complex lifestyle modifications for educational purposes or to motivate behavioral changes.
