Relative contributions of maternal Western‐type high fat, high sugar diets and maternal obesity to altered metabolic function in pregnancy

Human epidemiological and experimental animal studies have shown that maternal obesity (MO) combined with high energy Western‐style high fat, high sugar (HFHS) obesogenic diets can have adverse effects on mothers and fetuses during pregnancy and lactation and predispose offspring to later life metabolic dysfunction. Thus, understanding maternal, placental and fetal mechanisms of responses to this nutritional challenge is needed to determine efficacious interventions before, during and after MO pregnancy. Most controlled animal studies to evaluate developmental programming effects of MO and HFHS diets have been conducted in rodent models. Many differing periods of maternal exposure to obesogenic diets have been studied experimentally. When interpreting data for translation to human pregnancy it is important to consider the precise dietary components (macro‐ and micronutrients), food intake and extent and duration of MO and maternal HFHS dietary intake before conception and during pregnancy.

A major value of the study by Musial et al. (2017) in this issue of The Journal of Physiology is that it is one of very few designed to examine HFHS diet effects during pregnancy on maternal glucose and insulin and feto‐placental glucose metabolism. The majority of studies induce MO prior to pregnancy, some from the time mothers are weaned (Zambrano et al. 2010; Howie et al. 2013; Saben et al. 2014; Vega et al. 2015). Inducing adiposity early in the mother's life addresses consequences of the growing prevalence of childhood obesity and attempts to mimic the common problems of excessive BMI in women before they become pregnant, the most common human origin of programming by MO and obesogenic diet. The approach used in this study is a reductionist one aiming to throw light on specific maternal pregnancy mechanisms modified by MO and HFHS in pregnancy alone.

Obesity and HFHS diets may exert pre‐pregnancy effects through adverse actions on maternal reproductive tissues such as the ovary prior to conception (Igosheva et al. 2010) or through one or more of the well‐known effects of obesity on cardiometabolic function. There is therefore merit in a reductionist approach to evaluate exposures restricted to pregnancy only as attempted here. This report follows up a previous study by these authors (Sferruzzi‐Perri et al. 2013) reporting placental phenotype in the same model at the same gestational ages, 16 and 19 days’ gestation (dG; term, 20.5 days). Most studies on developmental programming by MO focus on offspring outcomes. It is a strength of the current study that it evaluates maternal metabolism employing a maternal hyperinsulinaemic–euglycaemic clamp, the gold standard for evaluating insulin function. This is a difficult procedure to conduct in mice. Investigators measured maternal glucose tolerance and insulin resistance, and feto‐placental glucose metabolism. Maternal HFHS increased hepatic insulin sensitivity while decreasing it in maternal skeletal muscle and white adipose tissue. MO mothers showed increased adiposity and reduced glucose production and tolerance. Placental weight and feto‐placental glucose consumption were reduced. Interestingly decreases in some key genes observed at 16 dG were normalized by 19 dG showing the importance of studying different periods of pregnancy.

The HFHS diet did not alter maternal corticosterone plasma levels. In contrast when maternal overnutrition is introduced from the time mothers are themselves weaned, corticosterone is elevated before pregnancy (Vega et al. 2015), at 19 dG (Rodriguez et al. 2012) and at the end of lactation, (Nathanielsz et al. 2013; Vega et al. 2015). The differences in corticosterone responses may be attributable to differences in diet, degree of obesity and/or duration of exposure. Since glucocorticoids play a major regulatory role in multiple maternal, placental, fetal and neontal developmental processes, differences in glucocorticoid function in different MO models may be of great use in determing different mechanisms and overall significance for maternal and offspring outcomes.

Maternal age affects responses in pregnancy. In this study mothers were 8–12 weeks of age, roughly corresponding to 16–20 years of human life (Dutta & Sengupta, 2016). Much growth and development is still occurring in female mice over this period. Experiments in sheep show that maternal age affects pregnancy outcome (Wallace et al. 2001). For example, adolescent mothers who are growing fast themselves compete with their fetuses for nutrients. A systematic study of effects of programming challenges, MO and others, at different maternal ages would be of great value especially if combined with metabolic and molecular end points such as reported here.

The exact nature of the diet is of great importance – especially when translating to the different obesogenic diets eaten by women before and during pregnancy. Addition of dietary sugar is essential to produce obesity – largely because its hedonistic effects increase food intake. When fed a diet of fat alone to produce obesity, most mammals reduce their food intake and do not become obese. This need for high sugar intake in experimental models parallels the elevated consumption of high glycaemic index carbohydrates that occurs in the developed and developing world since junk food and confectionary products are major dietary components. The findings of this study are consistent with a recent clinical intervention in pregnant women consuming a low glycaemic index diet, reduced sugar‐sweetened beverages and reduced saturated fats from 15 to 18 weeks’ gestation. This diet reduced gestational weight gain and maternal adiposity and decreased skin fold thickness (Poston et al. 2015).

In the future, ease of monitoring maternal and fetal blood and tissues with metabolomics approaches will provide much information and overcome concerns in all studies to date on confounds arising from rebalancing the diet. In this study, the HFHS diet includes 17% protein while the control diet contains 20%. The authors do not present maternal food and caloric intake data necessary to determine if the differences in maternal, placenta and fetal weight are due to nutrient deficiency or the diet. In their previous paper using this model (Sferruzzi‐Perri et al. 2013), HFHS food intake was reduced but caloric intake was similar. Fat and sugar intake are higher and, very importantly, protein intake lower (almost 50% lower) in mice on HFHS, which is likely to explain the lower maternal, placental and fetal weight.

In summary, this study adds valuable information on an HFHS diet's effects on maternal physiology, largely ignored in rodent models in favour of programming studies on fetuses and offspring, and thus points the way to important future approaches.

Additional information

Competing interests

No competing interests declared.

Author contributions

Both authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.