Cohort data. Sample characteristics. Table 3 shows the number of data points for weight, height and BMI for classic A-T (extracted dataset) and mild patients, subdivided by age.
Growth centile curves for classic A-T. Additional Figs. S1 – S10 in Additional File 2 are fitted growth curves for height, weight, and BMI. For ease of use, we made two sets of growth charts. For height and weight, they are Birth – 36 months and 2 – 18 or 20 years. For BMI they are ages 2 – 20 years. Centile values range from 3 to 97.
Additional Figs. S11 – S16 show individual data points used for construction of the charts.
A freely available calculator is on the Forgotten Diseases Research Foundation’s website.[29] It returns A-T-specific centiles and z-scores for height by age in months or years and months. To facilitate comparison with the unaffected population, data from the CDC centiles are also returned. The calculator is free to use, and the foundation does not see or retain any data entered into it. Calculators for weight and BMI will be added in the future.
Growth centiles and comparisons to CDC data. Impaired growth was a consistent finding in classic A-T. Although weight, length/height and BMI tracked with the CDC curves in the toddler years, faltering started early and continued through the teenage years. Figures 1 – 3 show median height, weight, and BMI in A-T compared to the CDC median. In each case, the A-T median falls further below the CDC median with increasing age.
Height faltering started sooner and was more severe in females. Female median height reached the CDC’s 3rd centile at age 6.0 and stayed close to it until crossing below it at age ~11.5. Male height ony reached the CDC’s 3rd centile around age 16.5, and did not cross below it (see Additional File 2). Some recovery of centiles occurred in females (Additional Fig. S17 in Additional File 4).
Males appeared to have an abbreviated adolescent weight spurt. In females, weight gain with time was nearly linear. Weight faltering occurred in both sexes, started sooner in female patients. The female median crossed below the CDC 3rd at age 10.5, compared to age 14.5 in males (Additional File 2). Female BMI did not fall below the CDC’s 3rd centile, while male BMI went below the 3rd centile around age 16 (Additional File 2). Relatively higher BMI in females may have been related to smaller relative stature.
Comparisons with mild A-T. Birth. Median gestational age in classic A-T was 39 weeks in both sexes. Twenty-two infants were premature (33 to 36 weeks), and twelve were born at 42 or more weeks. The remaining 171 infants were born between 37 and 41 weeks. Thus, 10.7% were born before term, 83.4% at term, and 5.9% post-term. These figures align well with CDC statistics on pre- and post-term birth rates in the general US population.[30]
All 26 mild A-T patients were born at term (averages: 39 weeks in males, 40 weeks in females, and 40 weeks in the entire cohort).
Birthweight. The median birthweight centile was 27 in the classic A-T cohort and 61 in the mild cohort (Table 4). A calculation using sample size, mean, and standard deviation indicated that this analysis had 99.99% power to reject the null hypothesis of equal means between the groups.
There were striking differences between the mild and classic groups: 72% of classic A-T patients had birthweights below the 50th centile, while 71% of mild A-T patients had birthweights at or above it (Fig. 4).
Birth length. Birth length and gestational age in weeks were known for 108 classic A-T patients (52 females, 56 males). Average birth length in both sexes was at the 48th centile based on the Fenton growth charts[26], and spread relatively evenly across the four quartiles (Fig. 5). The figure shows the group as a whole, given that there was no difference between the sexes.
Birth length in mild A-T was known for 6 females and 7 males. This sample size is too small for a meaningful analysis, but mean length for the group was at the 62nd centile.
Postnatal growth. In this analysis, we examined height and BMI from after birth until adulthood. Weight for age was not included in this analysis: given the generally short stature of classic A-T patients, we believed that BMI was a better indicator of nutritional status.
Height. Fig. 6 compares median height in classic and mild A-T patients with the CDC median. Faltering was evident in the classic cohort, and was less pronounced in mild A-T. By adulthood, 88% of women and 62% of men with classic A-T had heights below the CDC’s 10th centile, and only 1 had a height above the 50th. In mild A-T, 3 adult men and 1 woman had heights above the 80th CDC centile, with one man having a height above the 99th. Importantly, the centile drop in mild A-T may have been due to lack of childhood and adolescent growth data for some of the mild A-T patients, including the tallest patients.
The WHO defines stunting as a height at least two standard deviations below the mean. This condition was common in classic A-T. In classic A-T patients age 13 or older, stunting occurred in 67% of females and 41% of males (it was therefore 1.5 times more common in females; Additional Fig. S18 in Additional File 4). Of 26 mild A-T patients over age 13, only one was stunted.
A sex-specific height difference appears to exist in mild A-T as well as in classic A-T. Additional Fig. S19 in Additional File 4 shows that at ages 10 and above, male heights were close to the CDC average, while female heights were well below it.
In spite of this fact, females with mild A-T were considerably taller than those with classic A-T. Additional Fig. S20 in Additional File 4 shows median height in females with mild A-T as a function of age. Faltering occurred before age 2, but appeared to level off after that time. We were not able to make a similar figure for males, due to sparse data.
Height and mortality in classic A-T. To explore the link between growth and early mortality, we subdivided classic A-T patients into two groups: an early death cohort (died by age 15.0) and a longer-surviving cohort (survived past age 25.0). We compared CDC height z-scores for age 10-14 in these two groups and in mild A-T patients. Fig. 7 shows that patients who died young were shortest, with mild A-T patients being the tallest.
Effect of infection on height. We used medical records to classify the 162 extracted data set patients into 4 groups based on frequency and severity of infections before their 10th birthdays See Methods and Table 1 for criteria. Briefly, patients in groups 1 and 2 were designated as being less susceptible to infection, and patients in groups 3 and 4 were designated more susceptible. We classified 144 patients using this system (17 were too young and data was insufficient data for 1 other).
We compared height in older patients aged ≥13 years (88 total patients). The median height z-score was the same in both groups, at -2.08 SD. The mean in the less susceptible group was
-2.07 SD, and -2.17 in the more susceptible group (difference not statistically significant).
In addition, 55% of stunted patients had high susceptibility to infection and 45% had low susceptibility. These proportions were essentially the same among patients were not stunted: 54% had high susceptibility to infection and 46% had low susceptibility.
BMI in patients without gastrostomy tubes. We performed an examination of BMI in patients without gastrostomy tubes as a way of examining the effects of A-T in the absence of surgical interventions aimed at improving nutrient intake. Median BMI peaked between the 3rd and 6th birthdays and steadily declined thereafter (Fig. 8, dashed violet line). Faltering did not occur in mild A-T (solid green line). In classic A-T, there was no significant difference between males and females until later ages, when male BMI recovered slightly (Additional Fig. S21). When we examined BMI in classic A-T patients with and without feeding tubes, the differences were not statistically significant.
As classic A-T patients aged, an increasing proportion met or exceeded the WHO definition of moderate malnutrition, which is a BMI of 2 or more SDs below the mean (Fig. 9, solid violet bars). The pattern was different in mild A-T (dotted green bars).
BMI in many classic A-T patients was far below the -2 SD cutoff, particularly in adolescence. We had BMI measurements for 78 patients aged between 11 and 20. Twenty-two of them had BMI below -3 SDs (28.2%; WHO-defined severe malnutrition).
We also looked at minimum and maximum BMI centiles in classic A-T. The minimum values were always low, and fell with time to 8 SD by age 17 (table 5). Maximum values hovered around 2 until age 15, and then declined thereafter to –0.03 SD by the 20s. This pattern was not observed in mild A-T patients (green line in Fig. 6).
Some classic A-T patients were overweight or mildly obese. For example, from ages 11 – 20, 6 patients had BMIs between the 86th and 98th centiles, meeting the definition for overweight or obesity. We had data for three of these patients after age 21; their highest BMIs after that age had declined to the 71st, 49th, and 45th centiles. Overweight in classic A-T occurred most frequently at ages 3 – 5 (17% of children in this age group), and did not occur in adults. No mild A-T were obese, 5 were overweight before age 15, and none over 15 were overweight.