Table 3.
Effect of LBW on type 2 diabetes risk later in life with mediation of established biomarkers
| Mediator | n | Effect decompositiona
|
Proportion mediated (% of total effect) | |
|---|---|---|---|---|
| Mediated effectb | Effect not mediatedc | |||
| Insulin sensitivity and beta cell function | ||||
| WHR | 3,033 | 1.06 (0.97, 1.16) | 1.61 (1.22, 2.17) | 14 |
| HOMA-IR | 2,379 | 1.22 (1.02, 1.49) | 1.32 (0.95, 1.88) | 47 |
| HOMA-beta | 2,379 | 0.93 (0.84, 1.03) | 1.86 (1.34, 2.64) | –d |
| Leptin and leptin receptor | ||||
| Leptin | 1,312 | 0.99 (0.95, 1.01) | 2.17 (1.44, 3.35) | – |
| Soluble leptin receptor | 1,312 | 0.99 (0.96, 1.01) | 2.17 (1.41, 3.35) | – |
| Sex steroids and SHBG | ||||
| Total oestradiol | 1,311 | 1.00 (0.97, 1.02) | 2.16 (1.41, 3.35) | 0 |
| Free oestradiol | 1,311 | 1.00 (0.94, 1.05) | 2.16 (1.42, 3.39) | – |
| Total testosterone | 1,311 | 1.00 (0.94, 1.06) | 2.16 (1.40, 3.39) | – |
| Free testosterone | 1,311 | 1.05 (0.95, 1.17) | 2.11 (1.38, 3.30) | 8 |
| SHBG | 1,311 | 1.16 (1.03, 1.33) | 1.97 (1.25, 3.10) | 24 |
| Inflammation | ||||
| TNFα-R2 | 2,370 | 1.01 (0.97, 1.05) | 1.65 (1.17, 2.29) | 3 |
| IL-6 | 2,376 | 1.01 (0.95, 1.07) | 1.63 (1.15, 2.31) | 2 |
| hsCRP | 2,384 | 1.01 (0.95, 1.08) | 1.64 (1.17, 2.31) | 3 |
| Endothelial function | ||||
| E-selectin | 2,372 | 1.12 (1.02, 1.23) | 1.56 (1.10, 2.21) | 25 |
| ICAM-1 | 2,354 | 0.99 (0.91, 1.07) | 1.67 (1.19, 2.40) | – |
| VCAM-1 | 2,377 | 0.99 (0.93, 1.04) | 1.68 (1.20, 2.36) | – |
| Cellular ageing | ||||
| Leucocyte telomere length | 3,028 | 0.99 (0.96, 1.01) | 1.68 (1.29, 2.27) | – |
| Blood pressure | ||||
| Systolic | 3,048 | 1.03 (1.01, 1.07) | 1.61 (1.25, 2.16) | 8 |
| Diastolic | 3,046 | 1.00 (0.98, 1.01) | 1.65 (1.28, 2.22) | – |
Effects are shown as ORs (95% CIs). CIs were calculated using bootstrapping. Models were adjusted for age, race/ethnicity, cigarette smoking, alcohol consumption, physical activity, dietary fibre intake, dietary glycaemic load, BMI and family history of diabetes
OR for type 2 diabetes for those who had LBW, comparing the risk if the biomarker level was what it would have been with low vs normal birthweight
OR for type 2 diabetes in persons who had LBW vs those who had normal birthweight if the biomarker level was what it would have been with normal birthweight
Proportion mediated is undefined because the point estimate of the mediated effect is not in the same direction as the total effect