| Reference | Population | Methods | Estimated 24‐h Na excretion (extrapolation method) Mean ± SD | Measured 24‐h Na excretion (reference) Mean ± SD | Correlation | Comparison | Bland–Altman analysis |
|---|---|---|---|---|---|---|---|
|
Kawasaki et al. (1993) Japan |
159 healthy M and W (20–79 years) (Data set used to derive the equation) |
Collection 3–5 × 2nd morning voiding urine collected within 4 h after the 1st voiding but before breakfast + Kawasaki equation to estimate 24 h Na Reference: 3–5 × 24‐h urine (same days) |
NR | NR | 0.728 | ||
| 91 healthy M and W (40–67 years) |
Collection 1 × 2nd morning voiding urine collected within 4 h after the 1st voiding but before breakfast + Kawasaki equation to estimate 24‐h Na Reference: 1 × 24‐h urine (same day) |
NR | NR | 0.531 |
No statistically significant difference was found between mean estimated 24‐h Na and measured 24‐h Na |
||
| 15 healthy M and W (21–54 years) |
Collection 3 successive × 2nd morning voiding urine collected within 4 h after the 1st voiding but before breakfast + Kawasaki equation to estimate 24‐h Na Reference: 3 successive × 24‐h urine (same days) Procedure repeated 4–6 times |
NR | NR | 0.821 |
No statistically significant difference was found between mean estimated 24‐h Na and measured 24‐h Na |
||
|
Tanaka et al. (2002) Japan |
295 M and 296 W (20–59 years) from INTERSALT (Data set used to derive the equation) |
1 × casual spot urine + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h urine collection (started after the casual collection) |
178.9 ± 36.2 mmol/day | 187.2 ± 65.8 mmol/day | 0.54 |
Difference between measured 24‐h Na and estimated 24‐h Na 8.3 mmol/day |
|
|
336 M and W (20–59 years) (Validation data set) |
1 × casual spot urine + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h urine collection (started after the casual collection) |
154.5 ± 32.7 mmol/day | 178.5 ± 59.5 mmol/day | 0.32 |
Difference between mean estimated 24‐h Na and mean measured 24‐h Na, by quintile of estimated 24‐h Na Q1: 45.4 mmol/day Q2: 27.6 mmol/day Q3: 25.6 mmol/day Q4: 21.4 mmol/day Q5: 0.5 mmol/day |
||
|
Brown et al. (2013) 15 countries from North America and Europe (INTERSALT) |
2,948 M and W (20–59 years), from 29 cities (Data set used to derive the equation) |
1 × casual spot urine + INTERSALT equation to estimate 24‐h Na Reference: 1 × 24‐h urine collection (undertaken after the casual collection) |
NR |
Mean (mix, max) M: Belgium, Charleroi: 147.2 mmol/day Poland, Krakow: 240 mmol/day W: Germany, Cottbus: 117.8 mmol/day Italy, Bassiano: 167.5 mmol/day |
M: 0.51 W: 0.52 |
||
| 2,745 M and W (20–59 years), from 29 cities (Validation data set) |
1 × casual spot urine + INTERSALT equation to estimate 24‐h Na Reference: 1 × 24‐h urine collection (undertaken after the casual collection) |
M: USA, Hawaii: 176.3 mmol/day Poland, Krakow: 223.6 mmol/day W: Iceland, Reykjavik: 126.6 mmol/day Italy, Bassiano: 178.0 mmol/day |
M: USA, Hawaii: 144.2 mmol/day Poland, Krakow: 239.7 mmol/day W: Iceland, Reykjavik: 115.0 mmol/day Italy, Bassiano: 170.6 mmol/day |
M: 0.50 W: 0.51 |
Difference between mean measured 24‐h Na and mean estimated 24‐h Na: M: −1.6 mmol W: +2.3 mmol |
Bland–Altman showed overestimation of excretion at lower levels and underestimation at higher levels for both men and women | |
| Mente et al. (2014) 11 countries | 1,083 M and W (35–70 years) |
1 × fasting morning collection + Kawasaki equation to estimate 24‐h Na Reference: 1 × 24‐h urine collection (previous day) |
4,430 ± 1,253 mg/day | 4,116 ± 1,978 mg/day |
Intraclass correlation coefficient 0.71 (95% CI: 0.65 to 0.76) |
Difference between estimated 24‐h Na and measured 24‐h Na: 313 (95% CI: 182 to 444) mg/day |
Bland–Altman plot evidenced a systematic bias towards overestimation at the lower end and underestimation at the higher end |
|
1 × fasting morning collection + INTERSALT equation to estimate 24‐h Na Reference: 1 × 24‐h urine collection (previous day) |
3,257 ± 860 mg/day | 4,116 ± 1,978 mg/day |
Intraclass correlation coefficient 0.49 (0.29 to 0.62) |
Difference between estimated 24‐h Na and measured 24‐h Na: −872 (−728 to −1016) mg/day |
Bias significantly higher than Kawasaki equation | ||
|
1 × fasting morning collection + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h urine collection (previous day) |
3,569 ± 782 mg/day | 4,116 ± 1,978 mg/day |
Intraclass correlation coefficient 0.54 (0.42 to 0.62) |
Difference between estimated 24‐h Na and measured 24‐h Na: −548 (−408 to −688) mg/day |
Bias significantly higher than Kawasaki equation | ||
| Pfister et al. (2014) UK | 163 M and W (EPIC‐Norfolk study) |
1 × casual spot urine sample + INTERSALT equation to estimate 24‐h Na Reference: Mean of up to 6 × 24 h collections over 1 year |
NR | NR | NR |
Mean of differences between measured and estimated 24‐h Na: −21 mmol/day (95% CI: −32 to −11 mmol/day) |
Variance of the differences reasonably constant, with few outliers and most points lying within the calculated limits of agreement |
| Cogswell et al. (2013) USA | 407 M and W (18–39 y, 52% white) |
1 morning spot urine specimen (first specimen after discarding the first void; 08:30–12:30) +INTERSALT equation to estimate 24‐h Na Reference: 1 × 24‐h collection (same day) |
3,157 ± 891 mg/day | 3,323 ± 1,437 mg/day | 0.47 |
Difference between estimated and measured 24‐h Na: –165 mg (95% CI: −295, −36) |
Overestimation occurred at the low levels of 24‐h Na excretion and underestimation at the high level |
|
1 afternoon spot urine specimen (12:31–17:30) +INTERSALT equation to estimate 24‐h Na Reference: 1 × 24‐h collection (same day) |
3,197 ± 824 mg/day | 3,287 ± 1,408 mg/day | 0.49 | −90 mg (95% CI: −208, 28) | |||
|
1 evening spot urine specimen (17:31–23:59) +INTERSALT equation to estimate 24‐h Na Reference: 1 × 24‐h collection (same day) |
3,178 ± 846 mg/day | 3,298 ± 1,399 mg/day | 0.54 | −120 mg (95% CI: −230, −11) | |||
|
1 overnight specimen (first void collected the next morning after the longest period of sleep (04:00–12:00)) + INTERSALT equation to estimate 24‐h Na Reference: 1 × 24‐h collection (same day) |
3,028 ± 809 mg/day | 3,295 ± 1,400 mg/day | 0.44 | −267 mg (95% CI: −384, −151) | |||
|
Morning as above + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h collection |
3,574 ± 893 mg/day | 3,323 ± 1,437 mg/day | 0.50 | 251 mg (95% CI: NR) | Overestimation occurred at the low levels of 24‐h Na sodium excretion and underestimation at the high levels | ||
|
Afternoon as above + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h collection (same day) |
3,655 ± 870 mg/day | 3,287 ± 1408 mg/day | 0.51 | 368 mg (95% CI: NR) | |||
|
Evening as above + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h collection (same day) |
3,553 ± 820 mg/day | 3,298 ± 1399 mg/day | 0.59 | 255 mg (95% CI: NR) | |||
|
Overnight as above + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h collection (same day) |
3,272 ± 779 mg/day | 3,295 ± 1400 mg/day | 0.47 | −23 mg (95% CI: −141, 95) | |||
|
Morning as above + Kawasaki equation to estimate 24‐h Na Reference: 1 × 24‐h collection (same day) |
4,623 ± 1,471 mg/day | 3,323 ± 1,437 mg/day | 0.52 | 1,300 mg (95% CI: 1,152, 1,300) | Overestimation appeared to occur across low to high levels of 24‐h Na excretion | ||
| Ji et al. (2014) UK/Italy | 915 untreated M and W (297 white, 326 of black African origin and 292 South Asian; 40–59 years old) |
1 × timed morning urine sample + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h urinary collection (previous day) |
NR | NR | From 0.055 in black women to 0.330 in white women | Bland–Altman plots indicated consistent bias with overestimate for low and underestimate for high intakes. The bias was mainly due to the inaccuracy of age, weight and height to predict 24‐h creatinine excretion in the three ethnic groups, particularly in those of African origin | |
|
1 × timed morning urine sample + arithmetic extrapolation to 24‐h scale Reference: 1 × 24‐h urinary collection (previous day) |
NR | NR | From 0.116 in black women to 0.367 in white women | Bias was detected with both Bland–Altman plots and through quintile analyses (underestimate at low levels and overestimate at high levels) | |||
| 148 white M (mean age 58.3 years), in Italy |
1 × timed morning urine sample + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h urinary collection (previous day) |
NR | NR | 0.499 | Bland–Altman plot indicated overestimated values in the low 24‐h urinary Na levels and underestimated values in the high 24‐h urinary Na levels indicating consistent bias | ||
|
1 × timed morning urine sample + arithmetic extrapolation to 24‐h scale Reference: 1 × 24‐h urinary collection |
NR | NR | 0.329 | Arithmetic extrapolation produced underestimated values in the low 24‐h urinary Na levels and overestimated values in the high 24‐h urinary Na levels indicating consistent bias | |||
| Polonia et al. (2017) Portugal | 2,399 M and W (51% W; aged 18–96 years) |
1 × casual urine sample + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h urinary collection (1–7 days before) |
By quintiles of measured 24‐h Na: Q1: 3,755 ± 876 Q2: 3,973 ± 862 Q3: 3,986 ± 803 Q4: 4,150 ± 837 Q5: 4,342 ± 801 |
By quintiles of measured 24‐h Na: Q1: 2,361 ± 280 Q2: 3,121 ± 192 Q3: 3,825 ± 219 Q4: 4,661 ± 279 Q5: 6,298 ± 921 |
0.232 Intraclass correlation coefficient 0.340 (95% CI: 0.285, 0.391) |
Difference between measured and estimated 24‐h Na Overall: 11 mg/day (95% CI: −48.6, 70.6) By quintiles of measured 24‐h Na: Q1:−1,394 ± 905 Q2:−852 ± 875 Q3:−161 ± 832 Q4: 511 ± 867 Q5: 1,956 ± 1,140 |
Bias was detected with both Bland–Altman plots and through quintile analyses (overestimate at low levels and underestimate at high levels). Bland–Altman plots indicated that formula‐based estimates seem to have poorer clarification at higher levels than lower levels |
| 2,399 M and W (51% W; aged 18–96 years) |
1 × casual urine sample + Kawasaki equation to estimate 24‐h Na Reference: 1 × 24‐h urinary collection (1–7 days before) |
By quintiles of measured 24‐h Na: Q1: 4,844 ± 1,402 Q2: 5,171 ± 1,414 Q3: 5,240 ± 1,363 Q4: 5,524 ± 1,398 Q5: 5,864 ± 1,365 |
By quintiles of measured 24‐h Na: Q1: 2,361 ± 280 Q2: 3,121 ± 192 Q3: 3,825 ± 219 Q4: 4,661 ± 279 Q5: 6,298 ± 921 |
0.25 Intraclass correlation coefficient 0.303 (95% CI: 0.050, 0.475) |
Overall: −1,277 (95% CI: −1,346.7, −1,206.4) By quintiles of measured 24‐h Na: Q1: −2,483 ± 1,409 Q2: −2,050 ± 1,413 Q3: −1,416 ± 1,380 Q4: −864 ± 1,408 Q5: 434 ± 1,533 |
Bias was detected with both Bland–Altman plots and through quintile analyses (overestimate at low levels and underestimate at high levels) | |
| 2,399 M and W (51% W; aged 18–96 years) |
1 × casual urine sample + INTERSALT equation to estimate 24‐h Na Reference: 1 × 24‐h urinary collection (1–7 days before) |
By quintiles of measured 24‐h Na: Q1: 3,019 ± 841 Q2: 3,289 ± 848 Q3: 3,467 ± 901 Q4: 3,677 ± 923 Q5: 3,965 ± 894 |
By quintiles of measured 24‐h Na: Q1: 2,361 ± 280 Q2: 3,121 ± 192 Q3: 3,825 ± 219 Q4: 4,661 ± 279 Q5: 6,298 ± 921 |
0.359 Intraclass correlation coefficient 0.457 (95% CI: 0.337, 0.549) |
Overall: 569 (95% CI: 512.4, 624.7) By quintiles of measured 24‐h Na: Q1: −658 ± 852 Q2: −168 ± 857 Q3: 357 ± 914 Q4: 984 ± 930 Q5: 2,333 ± 1,150 |
Bias was detected with both Bland–Altman plots and through quintile analyses (overestimate at low levels and underestimate at high levels) | |
| Zhou et al. (2017) China | 141 M and W (27–64 years) |
1 morning spot urine specimen + Kawasaki equation to estimate 24‐h Na Reference: 1 × 24‐h urinary collection (started after the spot collection) |
246.1 ± 66.8 mmol/day | 220.8 ± 78.5 mmol/day | 0.31 |
Median (95% CI) difference between measured and estimated 24‐h Na 6.4 (−17.5, 36.8) mmol/day The individual absolute difference was > 51.3 mmol/day (3 g salt) in 52.5% of the participants |
|
|
1 morning spot urine specimen + INTERSALT equation to estimate 24‐h Na Reference: 1 × 24‐h urinary collection (started after the spot collection) |
143.6 ± 24.7 mmol/day | 220.8 ± 78.5 mmol/day | 0.25 |
−67.3 (−96.5, −46.9) mmol/day The individual absolute difference was > 51.3 mmol/day (3 g salt) in 63.1% of the participants |
|||
|
1 morning spot urine specimen + Tanaka equation to estimate 24‐h Na Reference: 1 × 24‐h urinary collection (started after the spot collection) |
183.7 ± 39.0 mmol/day | 220.8 ± 78.5 mmol/day | 0.35 |
−42.9 (−59.1, −24.8) mmol/day The individual absolute difference was > 51.3 mmol/day (3 g salt) in 48.3% of the participants |
95% CI: 95% confidence interval; INTERSALT: International Cooperative Study on Salt, Other Factors and Blood Pressure; M: men; Na: Sodium; N/A: not available; NR: not reported; SD: standard deviation; W: women.