TABLE 2.
The distribution of exposure assessment methods used to quantify dietary sodium intake among observational studies by study design1
| Study design | Prospective cohorts | Cross-sectional studies | Case-control studies | All observational studies |
|---|---|---|---|---|
| Urinary measures | ||||
| Multiple 24-h urine collections2 | 93 | 7 | 0 | 16 |
| Single 24-h urine | 7 | 39 | 1 | 47 |
| Multiple spot urine samples4 | 3 | 3 | 0 | 6 |
| Single spot urine | 135 | 28 | 1 | 42 |
| Dietary measures | ||||
| Multiple FFQs | 63 | 0 | 0 | 6 |
| Single FFQ | 12 | 7 | 2 | 21 |
| Multiple, multiple-day diet recalls/records/diaries | 1 | 0 | 0 | 1 |
| Single, multiple-day diet recalls/records/diaries6 | 1 | 1 | 0 | 2 |
| Multiple 24-h diet recalls/records/diaries | 0 | 5 | 0 | 5 |
| Single 24-h recall/record/diary | 1 | 11 | 0 | 12 |
| Totals | 53 | 101 | 4 | 1583 |
n = 157. WLVS, Women's Lifestyle Validation Study.
All multiple 24-h urine collections were collected on nonconsecutive days with the exception of 3 observational studies [1 prospective (127) and 2 cross-sectional (102, 208)].
One prospective cohort study conducted by Cortese et al. (93) used both a dietary and a urinary measurement to estimate dietary sodium intake and is counted in both categories. Sodium excretion was measured using multiple 24-h urine samples from women in the WLVS to correct the sodium intake estimated by FFQ in the study for measurement error. The correction equation was based on a linear regression with energy-adjusted sodium intake assessed by FFQ in the WLVS as exposure and urinary sodium as outcome: [corrected sodium intake = 1455.83 + (0.767* uncorrected FFQ sodium intake)] (93). This study was counted under both categories of multiple nonconsecutive 24-h urine collections and multiple FFQs.
Multiple spot urine samples were collected nonconsecutively in 4 studies (93, 143, 174, 185), whereas the sodium intake was averaged from early-morning urine samples collected on 3 consecutive days in 1 study (162, 163).
Takase et al. (201) instructed participants to “collect overnight urine in a paper cup and to bring in a sample of the urine in a plastic tube.” It is unclear if the overnight collection was timed or if just a spot sample from the overnight urine was used.
All of the studies assessed sodium using a multiple, consecutive-day diet recall (e.g., one 3-d recall from foods eaten on Monday, Tuesday, and Wednesday).