OR WAIT null SECS
These data suggest that heart health, as determined by the LC9 assessment, is positively correlated with pulmonary health.
Individuals with better cardiovascular health, as determined by LC9, tend to have stronger pulmonary health, according to recent findings.1
These findings resulted from a cross-sectional study that looked at data from the National Health and Nutrition Examination Survey (NHANES). The research was authored by Haolin Shi and Xiuhua Ma, from the Capital Medical University Daxing Teaching Hospital in Beijing.
Shi and Ma highlighted that there has been a lack of research that has explored the link of LC9 with multifactorial elements of lung health. The investigators’ new analysis implemented the criteria from prior published NHANES-related data to determine patients’ LC9 scores.2
“This is the first investigation into the connection between LC9 and lung health, and applied a more innovative ML to build predictive models,” Shi and Ma wrote.1 “It is hypothesized that LC9 may be positively correlated with lung health, and is better at predicting.”
The investigative team noted that the National Center for Health Statistics oversees the NHANES, with nationally representative data being collected by those involved through a probability-based, multistage stratified sampling approach. All data that Ma and Shi gathered are publicly available via the NHANES database, allowing for independent analyses.
NHANES was designed to evaluate US residents in terms of health and nutritional status. The team used pooled NHANES data spanning from 2005 to 2012, focusing on individuals aged 40 and older who had complete data on both lung health and LC9-derived scores.
The dataset initially used by the investigators included 40,790 individuals. After excluding 26,079 participants, given their lack of data related to pulmonary health data, and excluding another 4250 who had been missing LC9 scores, their final sample was made up of 11,061 participants. As NHANES 2005–2006 did not include information related to lung function, analyses necessitating complete assessments of forced expiratory volume in 1 second (FEV₁) and forced vital capacity (FVC) were restricted to data from the period between 2007 - 2012.
Nine metrics were integrated into the LC9 scores. There were 4 behavioral factors, including physical activity, diet, sleep, and tobacco exposure. Five health-related measures are also included: blood lipids, body mass index (BMI), blood glucose levels, mental health status, and blood pressure.
To evaluate any link between LC9 scores and patients' pulmonary health levels, Shi and Ma looked at linear associations through multiple regression models. They also used Restricted Cubic Spline (RCS) analyses to detect any potential non-linear trends. Any associations seen in this analysis were confirmed through subgroup analyses and interaction tests.
They also developed a machine learning model that had been based on the Light Gradient Boosting Machine (LightGBM) algorithm, using LC9 scores to predict outcomes in study participants' lung health. They found that 1725 subjects of the 10,461 included in the cardiovascular health category were classified as having low cardiovascular health.
They reported that 7,476 had moderate health and 1,260 as high levels of health. Consequently, Ma and Shi concluded that there was a clear positive association between higher LC9 scores and better lung health.
This trend was shown to be consistent over the score's individual components. Additionally, their RCS models demonstrated non-linear associations between LC9 scores and respiratory outcomes. Such outcomes included asthma, chronic cough, and chronic obstructive pulmonary disease (COPD).
In the investigators' LightGBM model, strong performance in predicting lung health was demonstrated. There were high values for area under the curve (AUC), as well as strong accuracy and specificity. Their SHAP analysis suggests that exposure to nicotine, BMI, and depression were the most influential LC9 variables contributing to the model’s predictive ability.
“There is a positive correlation between LC9 score and lung health. The LightGBM model combined with LC9 was excellent for its prediction,” they concluded.1 “Clinicians should consider using LC9 as a guiding framework to manage patients’ lifestyles and improve both cardiovascular and respiratory health.”
References