Table 4.
Performance of diagnostic measures of lactose malabsorption for detection of lactase non-persistence
| Measurea and substrateb | Cut-off usedc | Sensitivity (%) | Specificity (%) | FP (n) | FN (n) | TP (n) | TN (n) | PPV (%) | NPV (%) | κ indexd |
|---|---|---|---|---|---|---|---|---|---|---|
| Breath H2 | ||||||||||
| Lactose | 20 ppm | 100.0 | 96.2 | 1 | 0 | 14 | 25 | 93.3 | 100 | 0.9*** |
| CON - unadjusted | 20 ppm | 92.9 | 100.0 | 0 | 1 | 13 | 26 | 100 | 96.3 | 0.9** |
| CON - adjusted | 15 ppm | 92.9 | 96.2 | 1 | 1 | 13 | 25 | 82.9 | 96.2 | 0.9*** |
| A2M - unadjusted | 20 ppm | 78.6 | 100.0 | 0 | 3 | 11 | 26 | 100 | 89.6 | 0.8*** |
| A2M - adjusted | 15 ppm | 85.7 | 100.0 | 0 | 2 | 12 | 26 | 100 | 89.7 | 0.8*** |
| Plasma glucose | ||||||||||
| Lactose | 1.11 mmol/L | 64.3 | 69.2 | 8 | 5 | 9 | 18 | 52.9 | 78.2 | 0.3* |
| CON - unadjusted | 1.11 mmol/L | 92.9 | 26.9 | 19 | 1 | 13 | 7 | 40.6 | 87.5 | 0.2 |
| CON - adjusted | 0.83 mmol/L | 78.6 | 57.7 | 15 | 3 | 11 | 11 | 42.4 | 78.6 | 0.2 |
| A2M - unadjusted | 1.11 mmol/L | 64.3 | 38.5 | 16 | 5 | 9 | 10 | 36 | 66.7 | 0.02 |
| A2M - adjusted | 0.83 mmol/L | 57.1 | 46.2 | 14 | 6 | 8 | 12 | 36.4 | 77.8 | 0.02. |
| Urinary galactose/creatinine ratioe | ||||||||||
| Lactose | 0.10 mg/mg | 100.0 | NA | 26 | 0 | 14 | 0 | 35 | NA | 0 |
| CON- unadjusted | 0.10 mg/mg | 100.0 | NA | 26 | 0 | 14 | 0 | 35 | NA | 0 |
| CON - adjusted | 0.08 mg/mg | 100.0 | NA | 26 | 0 | 14 | 0 | 35 | NA | 0 |
| A2M- unadjusted | 0.10 mg/mg | 100.0 | NA | 26 | 0 | 14 | 0 | 35 | NA | 0 |
| A2M- adjusted | 0.08 mg/mg | 100.0 | NA | 26 | 0 | 14 | 0 | 35 | NA | 0 |
Classification was based on the genotype reference standard for lactase non-persistence (LNP; n = 14) relative to lactase persistent (LP; n = 26)
a Breath H2, plasma glucose indicates maximal increase in breath H2 and plasma glucose over baseline for 3 h post lactose and milk ingestion and urinary galactose/creatinine ratio at 3 h post lactose and milk ingestion
b Lactose: 50 g lactose in 250 mL water; CON and A2M: 750 mL of milk
cCut-off values for lactose and unadjusted (CON and A2M) were based on ingestion of 50 g lactose. Adjusted cut-off values were calculated based on ingestion of 750 mL of milk ~ 37.5 g of lactose
dκ represents the Cohen’s kappa index indicating the agreement between lactose malabsorption and genotype
Significance indicated as * p value < 0.05, ** p value < 0.01, *** p value < 0.001
e Galactose/creatinine ratio concentration range was lower than the threshold of 0.10 mg/mg; no lactose malabsorption was detected by the method using the literature threshold of 0.10 mg/mg [28] for both CON and A2M
A2M a2 milk, CON conventional milk, FN false negative, FP false positive, NPV negative predictive value, PPV positive predictive value, TN true negative, TP true positive. FN and FP represent the total number of false negativeand false positive, respectively out of total (N = 40) participants