The impact of maternal high fat nutrition on the next generation: food for thought?

The adverse impact of maternal gestational undernutrition on the development and long-term health of offspring has now been studied extensively, both in humans and animal models, with a range of reported postnatal consequences including increased risk of cardiovascular disease and type II diabetes, as well as behavioural alterations. However, the increasing prevalence of obesity both in Western and rapidly developing societies raises the question of whether maternal overnutrition may hold similar risks for the developing fetus and its future health. Howie and colleagues (2009) reported recently in The Journal of Physiology that a high fat (HF) diet fed to female rats either throughout life, including pregnancy and lactation, or solely during pregnancy and lactation, resulted in significant changes in offspring phenotype. These included altered birth weight and postnatal growth, increased fat deposition, and changes in leptin and insulin homeostasis. Importantly, their study has shown for the first time that a HF diet consumed solely during pregnancy and lactation was equally as detrimental to the development of offspring as if the mother had experienced a life-long overconsumption of fat (Howie et al. 2009), emphasising the particular importance of maternal diet during pregnancy.

To study the effects of maternal HF diet on offspring, rat dams were given either a control standard chow diet (17% kcals as fat) throughout life, including pregnancy and lactation; a maternal high fat (MHF) diet (45% kcals as fat) throughout life, including pregnancy and lactation; or fed standard chow until mating at which point they were switched to the HF diet throughout pregnancy and lactation (PLHF). Offspring from each maternal manipulation were then fed either a HF or control diet post-weaning. Immediately prior to mating, MHF dams were significantly heavier than controls, with increased body fat mass and fat to lean ratio, demonstrating the impact of a HF diet on pre-conception body composition. However, no significant differences in maternal weight gain during gestation were observed, leading the authors to propose that dietary composition during pregnancy rather than weight gain per se is the greater risk factor for offspring health (Howie et al. 2009). The inclusion of a further experimental group in which dams receive the HF diet only up to the time of conception, would help to confirm this.

Birth weights of pups from either maternal HF dietary regime were significantly reduced in comparison to controls despite no reported difference in litter size between groups. Variable effects on birth weight following gestational HF diet have been reported previously, thus the authors suggest that differences in fatty acid composition between the HF diets used in this and previous studies may be responsible (Howie et al. 2009). Alterations in uterine blood flow or placental transport capacity, as reported in rodent and sheep undernutrition models, may affect the availability of glucose or specific amino or fatty acids to the fetus resulting in altered fetal metabolism and an ensuing growth-restricted phenotype. Whether increased abdominal fat deposition in HF-fed dams could physically restrict fetal growth in the HF uterus is also of interest. Gestational analysis of maternal body fat deposition in the PLHF dams by the DEXA method used in the study could help to address this. Additionally, we feel that variation between this and other studies in the timing and duration of HF-diet intake prior to and during gestation should be considered.

Curiously, HF feeding during pregnancy increased the length of gestation by around 24 h, irrespective of the maternal diet prior to conception. The authors highlight a similarity to humans, where the production of prostaglandins required for parturition is altered by n-3 long-chain fatty acid consumption, resulting in increased gestation lengths (Szajewska et al. 2006). Interestingly, the HF diet fed throughout life also resulted in an observed reduction in oestrus cycling, mirroring obesity-associated poly-cystic-ovarian-syndrome in humans which results in significant changes in circulating steroid levels and impaired fertility. It is therefore also attractive to ask what the influence of the HF diet might be on quantitative measures of reproductive fitness. Investigation into steroid hormone levels and responsiveness may provide further insight into the observed impairments in oestrus cycling and delayed parturition. Since litter sizes were not different, developmental capacity post-fertilisation appears unaffected. Another route to explore regarding the increased gestational duration is whether there is slower development during HF embryonic/fetal development, given that reduced cell proliferation during preimplantation development has been reported in association with low birth weight after protein undernutrition in rats (Kwong et al. 2000). Analysis of blastocyst development and fetal allometry could be used now to address this question in the HF model.

Despite the similarity of the postnatal obese phenotypes observed in MHF and PLHF offspring, mechanisms underlying this programmed response were not detailed in the current study. The dietary interventions used in this study encompass several critical developmental windows, including oocyte maturation, fertilisation and pre-implantation embryo development, as well as fetal growth and postnatal development to weaning; these have all been shown previously to be differentially sensitive to both in vitro and in vivo environments (reviewed by Watkins et al. 2008). Focusing future work on identifying specific windows of development that are sensitive to maternal HF diet may assist in the identification of mechanisms resulting in the reported offspring phenotypes, and could determine whether the comparable phenotypic outcomes from the MHF and PLHF interventions arise via similar or distinct pathways.

A significant increase in caloric intake was seen in PLHF and MHF dams from postnatal day 10 until weaning, in association with increased weaning weights of the PLHF and MHF offspring. It is unclear whether the increased weight of these offspring is dependent on the increased maternal consumption, and it would therefore be of interest to separate the effects of the HF diet during gestation and lactation. Following weaning, offspring of all HF-fed dams displayed significantly increased postnatal weight gain and body fatness, independent of postnatal diet, despite only the PLHF male offspring demonstrating increased caloric intake (up to 60 days postnatally). This is characteristic of the catch-up growth phenotype seen in low birth weight offspring from intrauterine growth restriction and undernutrition models, leading to an increased growth trajectory into adulthood. Additionally, MHF and PLHF offspring had increased bone mineral content, and although the origins of this phenotype were not explored it probably represents an increase in fat-dependent calcium absorption. Leptin and insulin levels were reduced in offspring from MHF and PLHF dams at postnatal day 2 relative to controls, but were significantly raised by day 175 both with and without offspring HF diet. The similar caloric intake between treatment groups supports the proposal that the offspring of HF-fed dams may have leptin insensitivity (Howie et al. 2009), since increased levels of circulating leptin would normally be expected to reduce appetite and increase activity levels. Activity levels of the offspring were not reported, but it would certainly be of interest now to see if increased sedentary behaviour in these animals is contributing to the maintenance of their overweight phenotype.

Maternal HF diet exposure resulted in fatter offspring, regardless of postnatal diet. However, male offspring from dams fed a HF diet appeared to be more resistant to the effects of postnatal HF feeding than offspring from control dams, suggesting that these offspring had made adaptive responses (Gluckman & Hanson, 2004) to the HF environment in utero, preparing them for predicted high fat overnutrition in postnatal life. In contrast, offspring from control dams did not make such developmental adaptations and as a consequence of the resulting mismatch demonstrated an increased total percentage body fat in response to postnatal HF feeding compared to those from mothers fed the HF diet (Howie et al. 2009). It is interesting that this observation was not reproduced in the female offspring, suggesting a sex-specific sensitivity to HF feeding in these animals that appears to be independent of leptin and insulin response. Sex-specific responses to maternal diet are also reported in undernutrition models, highlighting the similarity between under- and overnutrition models which, as the authors suggest, may indicate common underlying mechanisms. Further investigation into the physiological and molecular characteristics of these offspring will hopefully shed light on the nature of the adaptive response occurring during fetal development in a MHF or PLHF pregnancy as well as the source of sex-specific differences.

Overall, the authors found that a high fat diet during pregnancy and lactation was detrimental to the future health of offspring, irrespective of the maternal diet prior to conception (Howie et al. 2009), supporting the hypothesis that the early stages of embryonic and fetal life are particularly sensitive to the maternal environment. This suggests that consideration of diet during pregnancy and lactation may be more important for the development of offspring than the mother's dietary history, and is of particular relevance to humans given reported excess maternal and postnatal nutritional intake in many modern societies. Investigations into the precise mechanisms behind these changes in offspring phenotype are now required. It would also be pertinent to investigate whether specific windows within the gestational period are differentially sensitive to gestational high fat feeding, given that this has been shown previously in the context of maternal protein undernutrition (Watkins et al. 2008).