Maternal and cord blood adiponectin levels in relation to post-natal body size in infants in the first year of life: a prospective study

This prospective study is the first to report the influence of maternal and cord blood adiponectin levels on neonatal weight-, height- and BMI-for-age Z score at birth and their changes at 3, 6 and 12 months of age. The main finding of our study is that cord blood adiponectin was positively associated with anthropometric measures at birth but inversely correlated with weight-for-age Z score changes from baseline during the first year of life after adjustment for potential covariates. Maternal blood adiponectin concentrations were significantly lower than those of the umbilical blood and showed a positive relationship with height-for-age Z score during the first year of life.

We confirm data from previous observational studies showing that cord blood adiponectin levels were several folds higher than those seen in adults [17, 18]. The recent study of Luo et al. observed a statistically significant positive correlation in adiponectin levels in maternal and fetal circulation (r?=?0.3, P??0.0001) [19]. A significant positive correlation in adiponectin levels in maternal vs. fetal circulation adjusted by covariates was also detected in our data (r?=?0.172, P?=?0.004). Genetic variation in the adiponectin gene or its regulating elements may be responsible for this maternal-fetal correlation. In contrast, only nonstatistically significant positive correlations have been reported in most previous studies [6, 20, 21]. Insufficient study power in some studies (n?=?51–74) [20, 21] and differences in the adjustment for potential confounders may partly explain the null findings in previous studies.

The positive correlation between cord blood adiponectin and fetal anthropometrics in our study is in agreement with some studies [6, 8, 21] but not all [22, 23]. Several factors may explain the lack of consistency in this regard. In the study of Lindsay et al., the absence of a significant association may be explained by the confounding effect of the existence of pre-disease states in some of the individuals included in the populations that were studied [9]. The other studies were limited by a small sample size (44–74), and thus they may have missed associations due to a lack of statistical power [20, 22, 23]. Additionally, studies in adults have indicated that the correlations between adiponectin and adiposity are modulated by ethnicities [12]. Whether ancestral background also accounts for disparities in the relationship between adiponectin and obesity in neonates needs to be further elucidated.

A few prospective studies have been conducted on the role of cord and maternal blood adiponectin levels in the postnatal growth of infants. In a sample of 588 children participating in the prospective prebirth cohort study Project Viva, Mantzoros et al. reported that cord blood adiponectin was inversely correlated with change in weight-for-length Z score (r?=??0.12; P?=?0.05) and weight-for-age Z score (r?=??0.10; P?=?0.04) from birth to 6 months but was not associated with change in length-for-age Z score (r?=?0.03; P?=?0.61) [10]. In agreement with the findings of Mantzoros et al. [10], we also found that neonates with a higher cord blood adiponectin gained less weight during the first year of life.

Most previous epidemiologic studies investigating the links between adiponectin levels and obesity in adults and children have demonstrated a negative relationship [24]. In fetal tissue, adiponectin is secreted not only by adipocytes, but also muscle and vascular cells, whereas in adult humans it is exclusively secreted by adipose tissue [25]. Fetal adipose tissue is also composed mainly of small newly differentiated adipocytes that lack the factors that are responsible for the inhibition of adiponectin production. Several lines of evidence suggest that senescent cells could accumulate in fat tissue with chronological aging and that these cells might contribute to age-related fat tissue inflammation and dysfunction [26]. Additionally, the available data suggest that brown adipose tissue (BAT) is more prevalent in children than in adults [27]. It is under strikingly different hormonal regulation than in white adipose tissue [28]. Moreover, adiponectin secretion from omental but not from sc adipocytes was negatively associated with measures of adiposity [29, 30]. The different origination of adiponectin, a decreasing sc adipose tissue/visceral adipose tissue ratio, the atrophy of brown adipose tissue and metabolic function changes in adipocytes with increasing age may in part explain the switch from a positive correlation between adiponectin and weight at birth to a negative correlation later in life [21, 28, 31]. In the Nurses’ Health Study, elevated adiponectin levels at baseline were associated with greater weight gain in healthy women [32], which is contrary to our findings in infants. The authors of the Nurses’ Health Study hypothesized that elevated adiponectin levels in humans was a sign of healthy adipose tissue and its capacity to adapt to more fat accumulation [32]. However, cellular stress and adipocyte overutilization with ageing led to metabolic dysfunction [26]. The variation in the metabolic function of adipocytes such as fat storage, the secretion and response to the modulation of adiponectin may collectively contribute to the discrepancies between infants and adults. In addition, BAT is especially abundant during infancy and its activation protects against weight gain [33]. Additional studies are required to determine whether BAT modulates weight changes by altering the secretion of adiponectin. Furthermore, some evidence suggested that adiponectin could reduce food intake and increase energy expenditure through action on the hypothalamus, which also provided a potential explanation for the negative association between cord adiponectin and weight gain [34].

The association between maternal adiponectin concentration and fetal growth are less clear. A negative correlation with birth weight has been reported by some studies [7], but not others [35, 36]. In this study, maternal adiponectin level was not a determinant of fetal size at birth, but was positively correlated with the height-for-age Z score during the first year after adjustment for cofounders. In theory, maternal adiponectin is unable to pass through the placental barrier due to its large molecular weight. However, Aye et al. [37] and Rosario et al. [38] have reported that adiponectin infusion in pregnant mice downregulates placental amino acid transporter activity and expression and decreases fetal growth. Thus, it is possible that maternal adiponectin circulation potentially modulates placental function, which, in turn, may affect the intrauterine environment of the foetus and the postnatal growth of infants. Further human studies are warranted to verify this hypothesis.

The strength of our study is its prospective design and repeated measures of anthropometric parameters in infants. We are the first to report the associations between adiponectin concentrations in maternal and umbilical blood and infant growth in early infancy in a Chinese population. However, our study has some limitations. First, not all of the subjects consented to the assessment of umbilical adiponectin status; thus, the results may be subject to potential bias. Second, the results may not be easy to generalize to other ethnic groups. Third, similar to all other previous studies, we did not determine fat mass directly, and did not measure the centrality of fat distribution. Thus, it remains unknown whether there is a relationship between adiponectin and visceral fat in infants. Finally, anthropometric measures during the follow-up had about 5.3–6.0 % missing data in our study. However, analyses after multiple imputation of the missing values yielded similar findings and thus our findings were not biased by the missing data.