INTRAFAMILIAL AND MIDPARENTAL-CHILD CORRELATIONS AND HERITABILITY ESTIMATES OF ANTHROPOMETRIC MEASUREMENTS IN PRADER-WILLI SYNDROME FAMILIES

Abstract

To determine the effects of familial background on anthropometric dimensions in Prader-Willi syndrome (PWS), we measured weight; height; sitting height; longitude and breadth of the head, hands, and feet; head, arm, and calf circumferences; and triceps and subscapular skinfolds in 28 individuals with the syndrome and their natural parents. Midparental-child correlations were significant for height and foot length, with heritability estimates of 0.52 and 0.68, respectively. Significant mother-child correlations were found for weight, height, foot length, and minimal frontal diameter for the total group; in addition, hand length and breadth, and calf and arm circumferences were significant for the patients age 12.5 years or under. These data provide evidence for maternal effects on several physical characteristics in PWS, particularly in younger patients.

The Prader-Willi syndrome (PWS), generally sporadic in occurrence, is characterized by infantile hypotonia, early childhood obesity, mental deficiency, short stature, small hands and feet, and hypogonadism.¹,² A deletion of the proximal long arm of chromosome 15 is identified in about one-half of patients.^2–5 The incidence of PWS has been estimated at 1 in 25,000 live births.¹

To date, only one detailed anthropometric study of patients with this syndrome has included longitudinal data but no familial data.⁶ We report for the first time intrafamilial and midparental-child correlations and heritability estimates on several anthropometric measurements on 28 families in order to gain a better understanding of the effects of familial background on these variables in individuals affected with PWS.

MATERIALS AND METHODS

Patients with PWS in this study were diagnosed on the basis of infantile hypotonia, early childhood obesity, hypogonadism, delayed psychomotor development or mental deficiency, or both, small hands and feet, and short stature. A total of 28 individuals and their natural parents were included. The patients were 17 males (10 with 15q11–13 deletion and 7 with normal chromosomes) and 11 females (7 with 15g11–13 deletion and 4 with normal chromosomes) ranging in age from 0.7 to 26.9 years (8 ≤ 7 yrs; 8 > 7 yrs but ≤12.5 yrs; 5 > 12.5 yrs but < 18 yrs; and 7 ≥ 18 yrs) with a mean age of 12.5 ± 7.3 years at the time of examination. The average maternal age at the time of examination was 38.9 ± 9.4 years (range 24.0–63.0 yrs) and the average paternal age was 42.2 ± 10.6 years (range 25.0–66.0 yrs).

The anthropometric measurements were made according to standard techniques as presented by Weiner and Lourie⁷ and consisted of the following: weight, height (or length), sitting height, total hand length, middle finger length, hand breadth, foot length, foot breadth, head circumference, head length, head breadth, minimal frontal diameter, arm circumference, calf circumference, triceps skinfold, and sub-scapular skinfold. Up to age 2 years, length was measured using a horizontal calibrated board with the child in the supine position. Thereafter, height was measured with an anthropometer. Skinfold measurements were obtained to the nearest half millimeter with a Lange skinfold caliper. Circumferences were obtained to the nearest millimeter with a steel tape. Weight was measured using a balanced beam scale.

Z scores for patients with PWS and their parents were computed using normative standards to control for age and sex effects. Normative standards for Z score computation of the anthropometric variables were as follows: minimal frontal diameter in individuals over age 7 years⁸; weight, height, arm circumference, triceps and subscapular skinfolds in adults^9–13; triceps and subscapular skinfolds in children under age 6 years¹⁴; and sitting height, hand length and breadth, middle finger length, foot length and breadth, calf and head circumferences, head length and breadth, weight, and height in children under age 18 years.¹⁵ Z scores were calculated from the formula:

$$Z=(X-\mathrm{X̅})/SD,$$

where SD is the standard deviation of the normal control, X̅ is the mean of the normal control, and X is the measurement of the individual. For each set of parents, midparental Z scores were calculated by averaging maternal and paternal Z scores for each anthropometric variable.

Pearson product-moment correlation and linear regression analyses were undertaken to study the relationship between anthropometric Z scores in the patients and their mothers and fathers. Midparental-child correlations, regressions, and heritability estimates (${\mathrm{h}}^2=\mathrm{r}/\sqrt{1/2}$) were computed for each anthropometric variable.^16,17

RESULTS

Table 1 shows the average Z scores for each anthropometric variable in the three parental categories and in the children with PWS. Significantly smaller foot length (< − 2 SD) and larger triceps and subscapular skinfold thicknesses (>2 SD) were observed for the patients with PWS compared with the parents. The average height, sitting height, finger length, hand breadth, foot breadth, and head breadth were also relatively lower (average Z score < − 1.75 for patients), while arm circumference was increased (average Z score 1.80 for patients) compared with parental Z scores. The average Z scores for the anthropometric variables for the mothers and fathers were close to the expected mean value of zero.

Table 1.

Parent, Midparent, and Child Z Score Data

Variable	N	Child Mean ± SD	Father Mean ± SD	Mother Mean ± SD	Midparent Mean ± SD
Weight	27	1.50 ± 2.76	0.65 ± 0.95	−0.05 ± 0.93	0.30 ± 0.69
Height	28	−1.77 ± 1.30	0.51 ± 0.98	−0.06 ± 0.65	0.22 ± 0.62
Sitting height	16	−1.78 ± 1.89	−0.25 ± 1.63	−0.32 ± 0.88	−0.29 ± 0.97
Hand length	27	−1.55 ± 1.05	0.51 ± 1.08	0.43 ± 0.91	0.47 ± 0.73
Finger length	27	−1.89 ± 1.31	−0.23 ± 0.96	0.28 ± 0.83	0.02 ± 0.65
Hand breadth	26	− 1.76 ± 1.64	0.62 ± 0.90	0.48 ± 1.06	0.55 ± 0.75
Foot length	27	−2.45 ± 1.30	0.04 ± 1.01	0.08 ± 0.86	0.06 ± 0.72
Foot breadth	22	− 1.77 ± 0.57	−0.44 ± 0.68	−0.26 ± 0.76	−0.33 ± 0.57
Head circumference	27	− 0.95 ± 1.18	0.16 ± 0.86	−0.06 ± 0.85	0.05 ± 0.70
Head length	27	− 0.62 ± 1.04	0.01 ± 0.92	−0.12 ± 0.92	−0.06 ± 0.78
Head breadth	27	− 1.98 ± 0.87	0.11 ± 0.50	−0.41 ± 0.80	−0.15 ± 0.49
Minimal frontal diameter	21	− 1.05 ± 1.06	0.08 ± 1.05	0.80 ± 2.14	0.44 ± 1.35
Arm circumference	27	1.80 ± 2.23	0.44 ± 0.90	0.06 ± 1.10	0.25 ± 0.67
Calf circumference	26	0.87 ± 1.86	0.32 ± 1.00	0.08 ± 0.99	0.20 ± 0.73
Triceps skinfold	27	2.25 ± 1.36	0.07 ± 0.77	−0.02 ± 0.70	0.02 ± 0.56
Subscapular skinfold	26	2.77 ± 2.50	0.01 ± 0.71	0.14 ± 1.02	0.08 ± 0.68

Table 2 shows marital, midparental, parent, and child correlation and heritability data for each anthropometric variable, as well as representative heritability estimates from the literature of normal children and their parents. No significant mother-father correlations were found in our sample with the exception of head length. No father-child correlations were found, but mother-child correlations were significant (p < 0.05) for height, weight, foot length, and minimal frontal diameter for the total group; in addition, hand length and breadth, and calf and arm circumferences were significant in patients age 12.5 years or under. Significant positive midparent-child correlations were observed for height and foot length. Midparent-child heritability estimates were greater than 0.50 for height, sitting height, foot length, and minimal frontal diameter. Foot length had the highest heritability estimate and triceps skinfold had the lowest.

Table 2.

Marital, Parent-Child, and Midparent-Child Correlation and Heritability Estimates and Comparisons with Published Data

		Correlation				Heritability
Variable	N	Mother- Father (r)	Mother- Child (r)	Father- Child (r)	Mid- parent- Child (r)	Mid- parent- Child (${\mathrm{h}}^2=\mathrm{r}/\sqrt{1/2}$)	Parent- Child26,a (h2 = VA/VP)	Twin Data23,b(h2 = rM − rD/ 1 − rD)	Parent- Child19 (${\mathrm{h}}^2=\mathrm{r}/\sqrt{1/2}$)		Mid- parent-Child27 (${\mathrm{h}}^2=\mathrm{r}/\sqrt{1/2}$)
									Mother- Child	Father- Child
Weight	27	0.08	0.39c	0.04	0.29	0.41	0.64	0.69	0.50	0.44	0.48
Height	28	0.11	0.38c	0.22	0.37c	0.52	0.82	0.88	0.52	0.53	0.68
Sitting height	16	0.10	0.44	0.23	0.39	0.55	—	0.72	0.51	0.55	0.59
Hand length	27	0.06	0.34	0.11	0.29	0.41	—	0.82	0.51	0.37	—
Finger length	27	0.05	0.18	0.20	0.27	0.38	0.80	0.88	—	—	—
Hand breadth	26	0.15	0.37	− 0.01	0.26	0.37	—	0.80	0.59	0.28	—
Foot length	27	0.19	0.62d	0.16	0.48c	0.68	—	0.81	0.44	0.48	—
Foot breadth	22	0.26	0.31	0.06	0.27	0.38	—	—	0.50	0.45	—
Head circumference	27	0.34	0.10	0.25	0.21	0.30	—	0.74	0.38	0.56	—
Head length	27	0.41c	−0.03	0.26	0.14	0.20	0.55	0.54	0.27	0.56	—
Head breadth	27	0.07	0.08	0.22	0.18	0.25	0.61	0.72	0.43	0.36	—
Minimal frontal diameter	21	0.36	0.43c	0.14	0.39	0.55	0.67	0.61	0.49	0.17	—
Arm circumference	27	−0.12	0.13	0.03	0.13	0.18	0.50	0.62	0.49	0.48	0.32
Calf circumference	26	0.08	0.37	−0.13	0.16	0.23	0.46	0.75	0.50	0.60	0.44
Triceps skinfold	27	0.16	0.16	0.04	0.12	0.17	—	0.63	0.28	0.32	0.28
Subscapular skinfold	26	0.21	0.26	0.23	0.31	0.44	—	0.60	—	—	0.25

^aThe V_A and V_P are additive genetic variance and phenotypic variance, respectively.

^bThe r_M and r_D are intraclass correlations for monozygous and dizygous twins, respectively.

^cP < 0.05.

^dP < 0.001.

DISCUSSION

No evidence of assortative mating for anthropometric characteristics could be detected in these families in whom a child had PWS. The Z score data do not indicate short stature, small hands and feet, obesity, or dolichocephaly as characteristics in the parents of affected children. The study showed no evidence of father-child correlations, but several mother-child correlations (weight, height, foot length, minimal frontal diameter) were observed in the total group, while additional mother-child correlations were calculated in the younger patients. The data provide evidence for maternal effects on several body measurements, particularly in younger patients, for body habitus (weight and height), minimal frontal diameter, circumferences, and hand measurements. Previous studies of normal children showed similar evidence of maternal effects.^18,19

The maternal effects on several body measurements is of interest in view of recent DNA findings of maternal disomy of chromosome 15 in several patients with PWS with normal chromosomes.²⁰ Correlation studies were undertaken to identify if maternal effects were more prevalent in the nondeletion PWS group than in the deletion group, but no significant differences were found in our 11 patients with non-deletion and 17 with the deletion versus their parents. Therefore, the maternal disomy of chromosome 15, if present in the nondeletion PWS, apparently does not affect the overall correlations compared with deletion PWS.

The maternal effects on several anthropometric measurements could also be an example of genetic imprinting^20–22 or genetic information that behaves differently if donated by the mother instead of the father. Additional research is necessary to understand better the maternal influences of certain characteristics and the role of gene activity in this process.

Midparent-child correlations were significant (p < 0.05) for height and foot length, with heritability estimates of 0.52 and 0.68, respectively. Sitting height and minimal frontal diameter both had heritability estimates of 0.55, although the correlations on which these estimates are based did not reach statistical significance. These physical characteristics are apparently influenced by parental factors, even though PWS is a condition characterized by short stature, small hands and feet, obesity, and dolichocephaly.

Midparent-child correlations differed among the anthropometric measurements, indicating variable degrees of parental influence over these morphologic characteristics, with significant or borderline significant correlations for longitudinal body measurements such as height, sitting height, and foot length. Significant correlations between parents and children were not observed for the measurements involving soft tissue such as skinfolds. Therefore, linear measurements demonstrated the highest heritability while breadth measurements, circumferences, and skinfolds generally had lower heritabilities. These results agree with previously reported studies of populations of healthy children and their parents.^23–27 The heritability data indicate that a child with PWS who has taller than average parents may be expected to be taller than the average child with the syndrome. A previous report, however, suggested that most affected individuals do not exhibit growth in height consistent with their familial background.²⁸

The heritability estimates in this study were generally lower than those reported in other studies of populations of healthy children and their parents.^16,23,25,26 This may be an indication that familial influences on anthropometric dimensions in individuals with PWS are tempered by the effects of the basic genetic defect in this condition.