Machine Learning Techniques for Identifying Lifestyle Factors Associated With Low Back Pain in Adults Aged 50 and Older Using Data From the Korean National Health and Nutrition Examination Survey.
Songhee Ko, Heesung Yang, Namsu Kim and 3 others
PMID 41186069WHAT IT FOUND
In Korean adults aged 50 and older, lifestyle factors were associated with chronic low back pain.
Female sex, older age, activity limitation, stress, low walking, high sitting, smoking, and diet appeared in a predictive model, but it was not strong enough for screening.
Key findings
01Of 5607 adults aged 50 and older, 1320 had chronic low back pain.
02The best model had AUROC 0.721, which the paper's own scale called good, but accuracy 0.683, and the authors described its remaining performance as suboptimal.
03Sex was the strongest predictor, followed by age and activity limitation.
STILL TO COME
How it was doneWhat they foundWhat it means for PTsWhat it means for RNs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The study was cross-sectional, so it cannot show that lifestyle factors caused chronic low back pain. The model was built only from adults aged 50 and older in Korea, so it may not apply to younger people, people in institutions, or other countries. The data came from 2014-2015, so they may not reflect current lifestyles or health policies. Chronic low back pain was self-reported as pain for more than 30 days in the last 3 months, not confirmed by clinical diagnosis. Lifestyle factors were also self-reported, and some relevant predictors such as sleep quality were not included. The authors described the best model's remaining performance as suboptimal, so it is not ready as a clinical screening tool. The chronic low back pain group was oversampled in the training data to address class imbalance.
Declared interests
The authors declare no conflicts of interest.
The easy way to misread this
Do not read the suggested lifestyle cutoffs as proven causes or treatment targets. The study is cross-sectional, and the authors state causality cannot be inferred from SHAP plots.