Table 2.
HIV+ and ID | HIV+ and iron‐sufficient nonanaemic | HIV− and ID | HIV− and iron‐sufficient nonanaemic | p‐Valuesa | |||
---|---|---|---|---|---|---|---|
n = 43 | n = 41 | n = 44 | n = 38 | HIV | Iron status | HIV × iron status | |
Faecal calprotectin (µg/g) median (IQR)b | 22 (7–57) | 6 (3–18) | 20 (3–75) | 5 (1–50) | 0.85 | 0.006 | 0.77 |
50–200 µg/g, n (%) | 7 (17) | 4 (10) | 8 (18) | 7 (18) | 0.36 | 1.00 | 0.36 |
>200 µg/g, n (%) | 4 (10) | 4 (10) | 5 (11) | 2 (5) | 0.99 | 0.50 | 0.45 |
Intestinal fatty acid–binding protein (pg/ml), median (IQR) | 859 (482–1164) | 799 (366–1110) | 633 (457–1010) | 935 (603–1219) | 0.47 | 0.44 | 0.09 |
Abbreviations: HIV, human immunodeficiency virus; ID, iron deficient; IQR, interquartile range.
Non‐normally distributed outcome variables were log‐transformed prior to analysis. Associations of the factors HIV and iron status with gut health markers were assessed using two‐way analysis of covariance for continuous variables and two‐way logistic regression analysis for categorical variables, adjusting for age, sex, ethnicity and deworming.
Total study population n = 162, HIV+ and ID n = 42, HIV+ and iron‐sufficient nonanaemic n = 39 and HIV− and ID n = 43, because insufficient stool sample sizes were provided in four cases.