DISCUSSION
To our knowledge, this is the first study to explore the mediating
effect of glucose level on the linking between advanced maternal age and
offspring birthweight. We found the positive linear association between
maternal age and macrosomia, and the total influence of maternal age on
offspring birthweight was mediated 4% to 16% by preconception FPG
level, which was more obvious among women with a history of adverse
pregnancy outcomes (mediating 18%).
We confirmed that maternal age had a linear association with the risk of
macrosomia, which was in line with most previous studies (28, 29). We
noticed that a majority of studies defined various cutoff values of
advanced maternal age taking all the adverse pregnancy outcomes into
consideration, such as 35 or 40 years old (30, 31). In view of
macrosomia, we firstly applied cubic restricted splines and found that
the age of 30 years would be the reasonable cutoff value for advanced
maternal age, which was consistent with Xiaolei’s definition with the
join point regression (32).And, based on the cutoff value, women with
advanced age would have a 66% increase in the risk of macrosomia.
We identified that the underlying pathway of the influence of advanced
maternal age on offspring birthweight involved the preconception FPG
level. Advanced maternal age was associated with higher preconception
FPG level, which was consistent with the prior result that preconception
diabetes mellitus had advanced age (24). E Cosson et.al proposed that
the preconception hyperglycemia indicated the FPG was already increased
during early pregnancy and might persist in late-pregnancy while insulin
resistance increased (20), resulting the increasing risk for macrosomia
(19). Salman et.al also reported that the pre-pregnancy impaired fasting
glucose level was associated with increased risk for gestation diabetes
mellitus (33), which indicating the mediating effect of glucose level
might be persistent during pregnancy. Our study supported those prior
results and confirmed the real relationship among preconception FPG
level, maternal age and offspring birthweight. Notably, although the
proportion of macrosomia among older reproductive-age women was low, the
total effect would be large with the increasing number of delaying
motherhood in China and the worse prognosis of macrosomia. In addition,
it’s a challenge to monitor glucose level for women during pregnancy
(34, 35), and the glycemic control before pregnancy among this risk
group would be implication.
Interestingly, we further found that the mediating effect of glucose
level was more obvious among older women with a history of adverse
pregnancy outcomes. In our study, 94.2% adverse pregnancy outcomes were
spontaneous and induced abortion. It was reported spontaneous abortion
could increase the prevalence of insulin resistance, which resulting in
the glucose intolerance and increasing the transport of glucose (36,
37). Similarly, induced abortion associated with metabolic syndrome such
as glucose intolerance in women with age ≥ 40 years (38).
Glucose
intolerance in older women with a history of adverse pregnancy outcomes
might to some extent explained the increased mediated effect of glucose
level on birthweight of advanced maternal age. Our finding suggested
that we should pay more attention to the preconception glycemic control
before pregnancy for older women with a history of adverse pregnancy
outcomes.
There are some limitations in our study. First, the 2-hour plasma
glucose level and HbA1c concentration were not detected. Second, we did
not get access to the glucose level during pregnancy. However, this may
not change the mediating effect of glucose level since that the high
glucose level indicated the increasing glucose level during early
pregnancy (20). Third, the glycemic control status among
reproductive-age women was unknown, which may underestimate the risk
(24). Finally, the dietary patterns were not available, further study
need to be explored.