Table 6.
Statistical comparison of multivariable associations between methods for collecting Global Diet Quality Score (GDQS) data and outcomes related to nutrient adequacy and metabolic risk among Thai adults: GDQS app compared with GDQS-24 compared with GDQS-food-frequency questionnaire
| Outcome | GDQS app |
GDQS-24 |
GDQS-FFQ |
P-diff |
|||||
|---|---|---|---|---|---|---|---|---|---|
| Per 1 SD | P-trend | Per 1 SD | P-trend | Per 1 SD | P-trend | GDQS app vs. GDQS-24 | GDQS app vs. GDQS-FFQ | GDQS-24 vs. GDQS-FFQ | |
| Continuous outcomes (statistic: estimated marginal mean, 95%CI) | |||||||||
| Probability of protein adequacy computed from 24HR (energy-adjusted), % | 3.7 (1.1, 6.2) | 0.025∗ | 1.4 (–1.2, 4.0) | 0.405 | 1.6 (–1.0, 4.2) | 0.308 | 0.250 | 0.344 | 0.595 |
| Mean probability of micronutrient adequacy computed from 24HR (energy-adjusted), % | 2.8 (1.9, 3.7) | <0.001∗ | 2.3 (1.5, 3.2) | <0.001∗ | 1.6 (0.7, 2.6) | 0.001∗ | 0.302 | 0.149 | 0.297 |
| Probability of protein adequacy computed from FFQ (energy-adjusted), % | 2.3 (–0.3, 5.0) | 0.171 | 0.8 (–1.9, 3.6) | 0.769 | 5.5 (2.8, 8.2) | <0.001∗ | 0.340 | 0.207 | 0.155 |
| Mean probability of micronutrient adequacy computed from FFQ (energy-adjusted), % | 2.1 (0.7, 3.6) | 0.018∗ | 1.9 (0.5, 3.4) | 0.009∗ | 7.1 (5.8, 8.5) | <0.001∗ | 0.324 | 0.014∗ | 0.028∗ |
| BMI, kg/m2 | –0.4 (–0.8, 0.0) | 0.045∗ | –0.1 (–0.5, 0.2) | 0.788 | –3.4 (–42.0, 35.2) | 0.971 | 0.165 | 0.158 | 0.391 |
| MUAC, cm | –0.5 (–0.8, –0.1) | 0.017∗ | –0.2 (–0.5, 0.1) | 0.340 | –0.1 (–0.4, 0.3) | 0.858 | 0.193 | 0.140 | 0.310 |
| WC, cm | –1.1 (–2.1, –0.2) | 0.037∗ | –0.6 (–1.5, 0.4) | 0.484 | –0.4 (–1.4, 0.5) | 0.442 | 0.228 | 0.259 | 0.553 |
| Fat mass, % | –0.9 (–1.4, –0.3) | 0.002∗ | –0.4 (–1.0, 0.2) | 0.370 | –0.6 (–1.1, 0.0) | 0.054 | 0.141 | 0.322 | 0.502 |
| SBP, mm Hg | –1 (–2, 0) | 0.355 | –1 (–2, 1) | 0.686 | –1 (–2, 1) | 0.960 | 0.424 | 0.382 | 0.644 |
| DBP, mm Hg | –1 (–2, –0) | 0.021∗ | –1 (–2, 0) | 0.127 | –1 (–2, 0) | 0.580 | 0.405 | 0.276 | 0.427 |
| Hb, g/L | –0.4 (–1.4, 0.7) | 0.470 | –0.8 (–1.9, 0.2) | 0.367 | 0.3 (–0.8, 1.4) | 0.290 | 0.782 | 0.338 | 0.273 |
| Total cholesterol, mg/dL | –3 (–7, 0) | 0.079 | –2 (–5, 2) | 0.339 | –3 (–6, 1) | 0.076 | 0.548 | 0.641 | 0.388 |
| LDL cholesterol, mg/dL | –3 (–7, 0) | 0.087 | –2 (–5, 2) | 0.469 | –2 (–5, 2) | 0.121 | 0.343 | 0.684 | 0.384 |
| HDL cholesterol, mg/dL | 1 (0, 2) | 0.032∗ | 1 (–0, 2) | 0.280 | 1 (–0, 2) | 0.234 | 0.238 | 0.427 | 0.547 |
| TG, mg/dL | –10 (–19, –0) | 0.113 | –11 (–20, –1) | 0.047∗ | –9 (–19, 0) | 0.204 | 0.308 | 0.676 | 0.362 |
| FPG, mg/dL | –2 (–4, 1) | 0.743 | –2 (–5, 1) | 0.070 | 1 (–2, 4) | 0.264 | 0.253 | 0.281 | 0.167 |
| HbA1C, % | 0.0 (–0.2, 0.1) | 0.892 | –0.1 (–0.2, 0.0) | 0.152 | 0.0 (–0.1, 0.1) | 0.630 | 0.274 | 0.391 | 0.233 |
| Number of MetS components (range: 0–5) | –0.2 (–0.3, 0.0) | 0.063 | –0.1 (–0.2, 0.1) | 0.382 | –0.1 (–0.2, 0.0) | 0.434 | 0.255 | 0.463 | 0.476 |
| 24-h urinary sodium, mg | 69 (–69, 212) | 0.649 | –48 (–184, 94) | 0.532 | –71 (–207, 74) | 0.503 | 0.310 | 0.476 | 0.748 |
| 24-h urinary potassium, mg | 117 (62, 172) | <0.001∗ | 59 (4, 113) | 0.061 | 98 (43, 156) | <0.001∗ | 0.260 | 0.568 | 0.309 |
| Sodium screener score (range: 0–100) | –1.4 (–2.1, –0.7) | 0.001∗ | –0.9 (–1.6, –0.2) | 0.040∗ | –0.1 (–0.8, 0.6) | 0.723 | 0.310 | 0.100 | 0.133 |
| Binary outcomes (statistic: multivariable odds ratio, 95%CI) | |||||||||
| BMI ≥25 kg/m2 | 0.89 (0.75, 1.05) | 0.128 | 0.98 (0.83, 1.16) | 0.950 | 0.87 (0.73, 1.03) | 0.180 | 0.217 | 0.691 | 0.322 |
| High MUAC | 0.83 (0.7, 0.98) | 0.020∗ | 0.94 (0.79, 1.11) | 0.553 | 0.9 (0.75, 1.07) | 0.283 | 0.142 | 0.265 | 0.418 |
| Abdominal obesity | 0.8 (0.67, 0.94) | 0.009∗ | 0.93 (0.79, 1.1) | 0.664 | 0.9 (0.75, 1.07) | 0.200 | 0.141 | 0.333 | 0.439 |
| Waist-to-height ratio >0.5 | 0.80 (0.67, 0.96) | 0.008∗ | 0.88 (0.73, 1.05) | 0.272 | 0.94 (0.78, 1.12) | 0.479 | 0.239 | 0.248 | 0.557 |
| Hypertension | 0.87 (0.73, 1.03) | 0.154 | 0.88 (0.74, 1.04) | 0.119 | 0.96 (0.8, 1.14) | 0.656 | 0.990 | 0.335 | 0.378 |
| Anemia | 1.08 (0.88, 1.32) | 0.532 | 1.13 (0.91, 1.39) | 0.412 | 1.05 (0.85, 1.3) | 0.961 | 0.841 | 0.649 | 0.499 |
| Raised LDL cholesterol | 0.96 (0.81, 1.14) | 0.899 | 0.90 (0.76, 1.08) | 0.451 | 0.95 (0.79, 1.14) | 0.207 | 0.340 | 0.322 | 0.497 |
| Reduced HDL cholesterol | 0.93 (0.77, 1.11) | 0.419 | 0.90 (0.75, 1.08) | 0.291 | 0.96 (0.80, 1.16) | 0.881 | 0.444 | 0.426 | 0.719 |
| Raised TG | 0.94 (0.79, 1.12) | 0.886 | 1.01 (0.85, 1.20) | 0.732 | 0.88 (0.73, 1.05) | 0.243 | 0.450 | 0.483 | 0.356 |
| Raised FPG | 0.88 (0.74, 1.04) | 0.385 | 1.03 (0.87, 1.23) | 0.939 | 0.95 (0.80, 1.14) | 0.621 | 0.468 | 0.543 | 0.398 |
| Raised HbA1C | 0.93 (0.70, 1.22) | 0.680 | 0.94 (0.72, 1.23) | 0.631 | 1.08 (0.82, 1.42) | 0.407 | 0.635 | 0.369 | 0.343 |
| MetS | 0.92 (0.77, 1.1) | 0.755 | 1.06 (0.89, 1.26) | 0.424 | 0.87 (0.72, 1.04) | 0.428 | 0.232 | 0.390 | 0.162 |
Values presented as multivariable estimated marginal means or odds ratios (95% CI) associated with a 1-SD positive difference in metrics. Models adjusted for age, sex, education, physical activity, smoking, sleep quality, and study group (Mahidol staff compared with community sample). P-trend: multivariable P for linear trend across metric quintiles. P-diff: P for the difference in linear trends across metric quintiles from the Wald test. ∗ indicates P-trend or P-diff <0.05.
Abbreviations: BMI, body mass index; CI, confidence interval; DBP, diastolic blood pressure; FFQ / -FFQ, food-frequency questionnaire; FPG, fasting plasma glucose; GDQS, Global Diet Quality Score; Hb, hemoglobin; HbA1C, hemoglobin A1C; HDL, high-density lipoprotein; LDL, low-density lipoprotein; MetS, metabolic syndrome; MUAC, midupper arm circumference; SBP, systolic blood pressure; SD, standard deviation; TG, triglyceride; WC, waist circumference; 24HR / -24, 24-h dietary recall.