The preterm infant born below 32 weeks of gestational age spends most of the third trimester of development in an environment fundamentally different from that which would have been experienced in utero. The extent to which this alters trajectories of metabolic development is a matter of increasing concern. These infants are now recognised as being at risk of a range of cardio-metabolic abnormalities in later life including hypertension and insulin resistance. We have demonstrated abnormal adipose tissue partitioning by term in extremely preterm infants with increased deposition in the deep subcutaneous and visceral abdominal compartments.1 The determinants of this aberrant development are unknown, but in older age groups it is known to be associated with hepatic steatosis.2

The third trimester is characterised by rapid liver growth. Intrahepatocellular lipids (IHCL) are an important reserve of essential fatty acids, although little is known about the factors that regulate their deposition during early development. Magnetic resonance spectroscopy (MRS) offers the only validated non-invasive method of quantifying IHCL and has become the tool of choice for this purpose in adults. To our best knowledge ¹H MRS has not previously been used to study IHCL in early infancy. The aim of the present study was to evaluate the use of ¹H MRS to assess IHCL content in newborn babies and to compare healthy term-born neonates with adults and infants born preterm.

METHODS

Infants were recruited from the Hammersmith & Queen Charlotte’s and Chelsea and Westminster Hospitals in London. The study was approved by the local research ethics committee and written parental consent was obtained. Preterm infants born at less than 32 weeks’ gestation were studied at term-equivalent age shortly before or after discharge home. They were clinically well and receiving no medications other than vitamin supplements. All infants in the preterm group had received either partial or complete courses of antenatal steroids. Healthy, appropriately grown term infants were studied in the neonatal period. The characteristics of both the preterm and term infants are displayed in table 1. Healthy adult volunteers, recruited for other studies reported elsewhere, were below 30 years of age with a BMI of 20–23.9, undertook regular exercise (2–3 h/week aerobic instruction), had no history of smoking, dyslipidaemia, medications or excess alcohol intake and had normal abdominal adipose tissue content and distribution.

Magnetic resonance spectroscopy and analysis

Infants were scanned in natural sleep in a supine position. The baby was fed, settled and swaddled using sheets and a vacuum sac and foam padding. Heart rate and oxygen saturation were monitored continuously using pulse oximetry. A neonatal paediatrician was present throughout the procedure. MR images were acquired on a Phillips 1.5 T system using a T₁-weighted spin-echo sequence (repetition time of 800 ms, echo time of 16 ms, 20–35 cm field of view, 192×256 matrix, to ensure accurate positioning of the 20×20×20 mm voxel in the liver, avoiding surrounding fatty tissue). Proton MR spectra were acquired from the right lobe of the liver using a PRESS sequence (TR 1500 ms, TE 135 ms, 128 averages) without water suppression. Spectra were analysed in the time domain as previously described with IHCL values adjusted for T₁ and T₂ effects and using hepatic water as an internal standard.3 4 Results are presented as IHCL CH₂/water.

Statistical analysis

Data were analysed using SPSS v 14 (SPSS, Chicago, IL). All results are expressed as medians and interquartile ranges. Differences were assessed using the Mann-Whitney U test.

RESULTS

We studied 26 infants, eight preterm (birth weight 1.33 (1.19–1.44) kg, gestational age 29.7 (29.2–30.5) weeks) and 18 term-born (birth weight 3.59 (3.31–3.70) kg, gestational age 39.9 (39.0–41.7) weeks). The results were compared with data from 32 healthy adult volunteers. Typical ¹H spectra from the livers of a term and preterm infant are shown in fig 1. IHCL contents were: preterm 1.69 (1.04–3.53), term 0.21 (0–0.54) and adult 0.55 (0.08–1.57). Preterm infant IHCL was significantly elevated when compared with that of term-born infants and adults (p = 0.002 and p = 0.025, respectively). Differences between term-born infants and healthy adults did not reach significance. No correlation was observed between IHCL and birth weight, gestational age at birth or SDS weight gain in the preterm group (r = 0.4, 0.5 and 0.3, respectively). Parenteral nutrition was received by preterm infants for a median of 1.9 days (0–4.6).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 (A) MR image from the abdomen of a healthy term infant showing the position of the PRESS voxel. (B) ¹H liver spectra from a term (1) and a preterm (2) infant at term-equivalent age.

DISCUSSION

We have shown that by the age of term, preterm infants have substantially higher IHCL than term-born infants and healthy young adults. As expected adult IHCL was higher than term-born infant IHCL, although this difference was not significant, emphasising the striking increase apparent in the preterm group.

The implications of IHCL accumulation in the preterm baby are unclear. Concern is warranted as in other age groups progression from simple steatosis and steatohepatitis through to cirrhosis is well described. In addition to steatosis, inflammation with hepatocyte destruction is needed for a diagnosis of steatohepatitis. Liver function data are not available for the infants in this study, but future work must examine associations between MR evidence of increased IHCL and abnormal liver enzymes.

Raised IHCL is closely associated with abdominal adiposity,2 obesity, dyslipidaemia and insulin resistance and the preterm infant is increasingly considered to be at risk of these conditions. There are several possible determinants of raised IHCL in the preterm infant. These include protein deficiency, antenatal steroids, systemic illness, infection, parenteral nutrition, a lipid-carbohydrate rich diet, drug toxicity and hypoxia. Alterations in the intestinal microbiome may also modulate host lipid metabolism. Germ free mice experimentally colonised with non-conventional flora develop elevated IHCL. As the preterm infant is reared in an environment conducive to abnormal intestinal colonisation, this opens yet a further avenue of exploration. Our preliminary study is too small to allow us to address these potential causal factors. However, it is worth noting that the preterm infants in this study were predominantly breastfed with human milk received for 41–95% days between birth and discharge. Furthermore, the increase in IHCL did not appear to be associated with parenteral nutrition which was only received in the immediate postnatal period for a short period.

An important consideration must be to establish if these changes persist and to what extent they are related to the several adverse long-term metabolic consequences of extremely preterm birth that have become a matter of recent concern.5 We have shown that acquisition of MRS data in babies is both quick (approximately 5 min) and feasible. Furthermore, it is the only available non-invasive quantitative tool which lends itself to longitudinal measurements of IHCL. Adult data indicate excellent correlation between MRS estimation of IHCL and histological scores obtained by liver biopsy. In addition, there are advantages when compared to fat selective MRI which although able to quantify IHCL has a detection threshold that is 10-fold higher than with ¹H MRS (2% vs 0.2%).

The preliminary observations we report here and our previous work1 suggest the early onset of measurable metabolic derangement in the preterm infant, and argue the necessity of further investigation, intervention and prevention in the neonatal unit.