Vitamins and Minerals
Accurate estimates of sodium (Na) intake are critical for advancing our understanding of Na and cardiovascular disease. The aim of this study was to utilize data from a controlled feeding study to assess the agreement between estimated 24hUNa, calculated from several timed spot urine collections, and known Na intake.
Methods : This secondary analysis used data from the control phase of a potassium intervention study in normotensive men and women (n=39) to avoid the potential confounding factor of dietary potassium manipulation. Participants followed the controlled diet containing on average 3463 mg Na/d for 3 days. On day 4, participants ate 3 meals in a clinic setting and collected urine every 2 hours starting with a fasting second void. The 24-hour urine collection was completed with the first morning void the following day. Spot urine samples were analyzed for Na and estimated 24hUNa was calculated from all samples using the INTERSALT and Tanaka equations, and from only the fasting second collection using the Kawasaki equation. The relationship between estimated 24hUNa and Na intake was assessed using Spearman correlations, and agreement between estimated 24hUNa and intake was assessed using Wilcoxon Signed Rank Tests and Bland Altman plots.
Results : For all timed spot urine samples, estimated 24hUNa using INTERSALT and Kawasaki were significantly correlated with Na intake (all p< 0.05), whereas estimates using Tanaka were not significantly correlated with intake (all p >0.05). For all timed spot urine samples, estimated 24hUNa using INTERSALT was significantly different from Na intake (all p< 0.05). All estimates of 24hUNa using Tanaka were significantly different from Na intake (all p< 0.05) except from estimates from the 2-hour sample (p=0.172) and the first void at 24 hours (p=0.102). Estimate of 24hUNa using Kawasaki was not significantly different from intake (p=0.635). Bland Altman plots showed high levels of bias, especially at higher and lower levels of intake.
Conclusions : These results suggest that, regardless of timing of collection and choice of equation, estimated 24hUNa is not a valid indicator of actual Na intake. New equations that directly estimate Na intake are needed.
Funding Sources :
The parent study was funded by the Alliance for Potato Research and Education (APRE).