Background-Monitoring oesophageal pH conventionally detects "acid reflux" (pH less than 4). The pH of the gastric contents determines whether or not reflux can be detected. Aim-To monitor gastric and oesophageal pH simultaneously in order to determine the eVect of milk feeds on gastric pH and how this would influence interpretation of the oesophageal pH record. Methods-Milk fed infants for whom oesophageal pH monitoring was requested underwent simultaneous gastric and oesophageal pH monitoring using a dual channel pH probe. Results-Twenty of 24 records were technically satisfactory. Mean reflux index was 1.0%, range 0.0-4.0%. Gastric pH was less than 4 for 24.5% (range 0.6-69.1%) of the total time. The average time the gastric pH was greater than 4 after feeds was 130 minutes (range 29-212 minutes). The corrected reflux index (limited to the time the gastric pH was less than 4) was 2.6% (range 0.0-11.0%). Conclusion-The pH of the gastric contents may be greater than 4 for prolonged intervals, during which oesophageal pH monitoring using current criteria cannot detect reflux nor correlate it with clinical events. A low reflux index may reflect prolonged buVering of gastric acidity rather than the absence of reflux. (Arch Dis Child 2001;84:273-276)

The school-based behavioural adjustment at 7-8 years of a cohort of 243 prematurely born, very low birthweight (< 1501 g) children and their normal birthweight controls is reported. The findings indicate that the children born preterm (both male and female) were rated by their teachers as expressing more behaviour problems than their controls, and were less well adjusted to the school environment. The deficits noted in the preterms applied across the social classes, with no amelioration noted in preterms of higher social class. It is speculated that the problem behaviours reflect a failure in self-regulatory functions.

In recent years it has been shown that alteration in heart rate variability can be used for the objective assessment of autonomic function in adult diabetic patients. Using a microcomputer based system for on-line monitoring and analysis of heart rate variability by power spectral analysis, 100 children with Type 1 (insulin-dependent) diabetes mellitus were studied. A highly significant negative correlation was identified between heart rate variability and duration of diabetes (r = -0.88, p less than 0.0001). The mean heart rate variability in patients with diabetes of duration 3 years or more was significantly lower in comparison to age-matched control subjects. A highly significant negative correlation was evident between heart rate variability and mean HbA1 in patients with duration of diabetes of 5 years or more. A mean HbA1 over 10% during this period was associated with the greatest reduction in heart rate variability. Serial measurements of heart rate variability in diabetic children may be of predictive value prior to the onset of symptomatic neuropathy.

Whether very-low birthweight (VLBW less than or equal to 1500 gm) children differ from normal birthweight (NBW greater than 2500 gm) children with respect to social (as opposed to intellectual) competence has been a relatively neglected issue. The social competence at school age of 183 VLBW children was therefore compared with that of 183 NBW children born at the same hospital matched for age, gender, social class, parity, and maternal age. A multi-informant, multidefinitional approach to social competence was adopted involving teacher, (same-gender) peer, and self-ratings of the 366 children's levels of social maladjustment, social skill, and peer acceptance. VLBW children from the upper, middle, and lower social classes received significantly higher mean sadness/unhappiness scores (even with IQ covaried) than their NBW counterparts. Mean scores for the VLBW group were also higher for social withdrawal, and lower for both social skill and peer acceptance. Possible antecedents and consequences of such group differences in affect and sociability are suggested.