Preterm babies born in 1982–1985, randomly assigned to a standard formula designed for term babies, subsequently had lower developmental scores than those receiving a multinutrient enriched preterm formula. Yet, paradoxically, infants randomly assigned donated banked breast milk or the same preterm formula had similar scores, despite the lower macronutrient content of human milk.^1–3

Consideration of the nutrient content of the feeds suggested that taurine was a candidate single nutrient to explain this paradox as term formula contained only a trace, whereas preterm formula and breast milk contained 5 µmol/100 ml (table 1). Furthermore, as taurine is neurotrophic and affects neurotransmission,⁴ the possible explanation was biologically plausible. We therefore explored the hypothesis that taurine was an explanatory nutrient for the benefits of both breast milk and the higher nutrient formula to long term neurodevelopment.

PATIENTS AND METHODS

Neonatal plasma taurine concentrations,⁵ Bayley Scales of Infant Development (corrected for gestational age) at 18 months of age, and the Wechsler Intelligence Scale for Children-revised (WISC-R) at 7 years were available in 157 children (mean (SD) birth weight, 1398 (277) g; gestation, 31 (2.4) weeks). The lowest (minimum) plasma taurine concentration during their hospital stay was obtained for each subject.

RESULTS

Minimum plasma taurine concentrations correlated with corrected Bayley mental development index (r  =  0.28, p < 0.001) and WISC-R arithmetic subtest score (r  =  0.22, p  =  0.006) (fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

 Bayley mental development index at 18 months, arithmetic subtest of WISC-R at 7 years, and minimum neonatal plasma taurine concentration (μmol/l). Taurine, 1st quartile, 20–43; 2nd quartile, 44–55; 3rd quartile, 56–67; 4th quartile, 68–180. Mental development index, mean (SE) 97 (2). Arithmetic score, mean (SE) 9 (0.3).

These relations remained significant after adjustment for possible confounding factors: (a) clinical illness—birth weight, gestational age, weight for gestation, days inpatient, days ventilated; (b) possible undernutrition—plasma concentrations of other amino acids, total protein, urea; (c) amount of intravenous amino acids.

Minimum taurine was not related to the Bayley psychomotor development nor, after the confounding factors had been allowed for, to the other subtests of the WISC-R. Neither maximum nor mean taurine measurements were related to neurodevelopment.

The length of hospital stay also contributed to lower mental development scores, and gestational age to the WISC-R arithmetic scores. After these factors had been allowed for, the relations of minimum taurine to mental development and to the arithmetic subtest remained significant (partial correlations: r  =  0.19, p  =  0.016; r  =  0.18, p  =  0.024). There was no interaction between minimum taurine and hospital stay or gestation.

The positive association of neurodevelopment with own mother’s milk, described previously, was no longer significant after taurine had been allowed for (partial correlations with mental development: r  =  0.03, p  =  0.70; with the arithmetic subtest: r  =  0.09, p  =  0.24).

DISCUSSION

The results support the hypothesis that low taurine status in the neonatal period of preterm babies adversely influences later neurodevelopment, and that the advantages of breast milk are partly due to taurine. Some caution is necessary. This was not a randomised trial. The strengths of the relations, although significant, were modest (r  =  0.28, 0.22), but greater than those seen in this study between either Bayley mental development or the arithmetic subtest of WISC-R with birth weight (r  =  0.17, 0.18) or gestational age (r  =  0.16, 0.19).

These data support the view that taurine is a conditionally essential nutrient, as a dietary supply was required for optimum outcome. It is one more example of short term nutritional differences in the newborn having apparent long term effects.

The relations of taurine to mental rather than motor development, and to arithmetic but not other WISC-R subtests, indicate that transiently low neonatal taurine status (“hypotaurinaemia”) has selective neurodevelopmental effects. Transiently low blood concentrations of other metabolic factors have been shown by us and others to be associated with subsequent reduced developmental scores, notably hypoglycaemia and low thyroid hormone status. The mechanisms for this selection are not known, but there are different concentrations of taurine in different parts of the brain. The selective effects of such early influences on brain function require further study. Other work from this centre has shown that calculation deficits are associated with decreased grey matter in the left parietal cortex, and it may be that taurine is implicated in development and function in this region of the brain.⁶ Therefore more detailed cognitive testing and neuroimaging studies using magnetic resonance imaging are planned as the children get older.

Although further work is needed to test whether taurine is a conditionally essential nutrient for neurodevelopment in healthy term as well as preterm infants, it seems prudent to ensure adequate early taurine intake. Yet, currently, it is not mandatory to add taurine to infant formulas.