Skip to main content

Table 1 Identification of metabotypes

From: Optimisation of a metabotype approach to deliver targeted dietary advice

Clustering variables

Metabotype 1 (n = 71)

Metabotype 2 (n = 97)

Metabotype 3 (n = 39)

p value*

p value**

Triacylglycerol (mmol/L)

0.75 (0.59, 1.00)3

0.77 (0.61, 0.93)3

1.73 (1.36, 2.53)1,2

6.9 × 10–30

9.5 × 10–24

Total cholesterol (mmol/L)

5.00 (4.20, 5.40)2

3.90 (3.50, 4.30)1,3

5.50 (4.80, 6.00)2

1.7 × 10–28

2.4 × 10–19

HDL-cholesterol (mmol/L)

1.90 (1.74, 2.02)2,3

1.38 (1.28, 1.57)1,3

1.14 (0.98, 1.41)1,2

4.1 × 10–33

7.2 × 10–22

Glucose (mmol/L)

4.80 (4.40, 5.00)3

4.80 (4.50, 5.05)3

4.90 (4.70, 5.40)1,2

4.0 × 10–3

0.057

  1. Clustering variables are from fasting plasma samples. Values are presented as median (percentile 25, percentile 75). Underlined values represent the highest median across the metabotypes. Italic values represent the lowest median across the metabotypes. Superscript numbers denote where the differences lie across the metabotypes; for example, 1means significantly different from metabotypes 1. *Analysis of variance with Bonferroni post hoc test. **General linear models adjusted for age and sex