Table 5.
Associations between DBP exposure (measured by MBP) and cognition.
| Reference; study confidence; N | Age at outcome measure | Median exposuresa | Exposure IQR (or as specified) | Cognition overall effect estimate β (95% CI) | Cognition in boys β (95% CI) | Cognition in girls β (95% CI) |
|---|---|---|---|---|---|---|
| Included in meta-analysis (age ≤4 yrs, cognition measured using Bayley, medium or high confidence) | ||||||
| Polanska et al. (2014); medium; 150 | 2 yr | 4 ng/mL | 130 (95th) | −0.5 (−1.8, 0.7) | NR | NR |
| Kim et al. (2011); medium; 460 | 6 mo | 17 ng/mL | 5.4‒41.1 | −0.5 (−1.2, 0.1) | −0.9 (−1.8, −0.05)* | −0.2 (−1.2,0.8) |
| Gascon et al. (2015); high; 367 | 1 yr | 31 μg/g | 20‒49 | 1.2 (−0.9,3.3)b | 1.5 (−1.6,4.6)b | 1.4 (−1.5,4.2)b |
| 4 yr | 0.8 (−1.1,2.6)b | 1.7 (−1.4,4.7)b | 0.4 (−2.0,2.8)b | |||
| Doherty et al. (2017); medium; 258 | 1–3 yr | 35 ng/mL | 202 (95th) | 0.7 (−0.7,2.0) | 1.7 (0.1,3.3)* | −2.8 (−5.0,−0.5)* |
| Whyatt et al. (2012); medium, 319 | 38 ng/mL(GM) | 0.2‒785 (range) | −1.1 (−2.6, 0.4) | 0.3 (−2.0,2.6) | −2.7 (−4.7,−0.7)* | |
| Huang et al. (2015); medium, 251 | 2–3 yr | 78 μg/g (GM) | NR | 0.9 (−2.0,3.8) | NR | NR |
| Tellez-Rojo et al. (2013); medium; 135 | 2‒3 yr | 86 ng/mL | NR | 0.3 (−1.0, 1.7) | 0.5 (−1.3,2.4) | −0.2 (−2.2,1.8) |
| Not included in meta-analysis | ||||||
| Kim et al. (2017); medium; 175 | 6 yr | 30 ng/mL | NR | 4.0 (−1.3,9.2) | NR | NR |
| Factor-Litvak et al. (2014); medium, 328 (same study as Whyatt et al., 2012) | 7 yr | −2.7 (−4.3, −1.1)* | −1.9 (−4.3,0.6) | −3.2 (−5.4,−0.9)* | ||
| Li et al., 2019; high; 253 | 5, 8 yr | 26 wk gest 23 ng/mL | 9–49 | 0.6 (−0.9,2.2) | NR | NR |
| 4 yr 22 ng/mL | 11–43 | 2.1 (0.3,3.9) | NR | NR | ||
| Nakiwala et al. (2018); medium; 452 | 5 yr | 45 ng/mL | 12–444 (5th-95th) | −0.09 (−0.2,0.04) | NR | NR |
| Huang et al. (2015); medium, 251 | 2, 5, 8, 11 yr | 78 μg/g (GM) | −0.2 (−2.5, 2.1) | NR | NR | |
Maternal exposure measured during gestation unless otherwise noted.
Upon request, author provided re-calculated results with 1n-transformed exposure and β representing a ln-unit increase, to improve comparability with other studies
One low confidence study (Kim et al., 2018) not shown in table because results were incompletely reported.
All studies used ln-transformed exposure and βs represent 1 ln-unit increase, except for Gascon, which used log2-transformation and β represent doubling of exposure.
Results that support an adverse association are shaded. This represents one or more of the following: p < 0.05, β ≥ 0.5, or exposure-response trend across categories of exposure. NR = not reported. GM = geometric mean.
p < 0.05.