Arch Dis Child: first published as 10.1136/adc.59.5.453 on 1 May 1984. Downloaded from http://adc.bmj.com/ on September 3, 2019 at India:BMJ-PG Sponsored.

Archives of Disease in Childhood, 1984, 59, 453-459

Norms for hand grip strength

D G NEWMAN, J PEARN, A BARNES, C M YOUNG, M KEHOE, AND J NEWMAN
Department of Child Health, University of Queensland, Australia

SUMMARY Norms for hand grip strength of healthy children are presented. Sex and age specific
centiles for age 5 to 18 years have been determined using a portable strain gauge dynamometer
with an accuracy of 0O5 N. The test group comprised 1417 healthy, urban school children from a
middle class suburb of Brisbane. Mean maximum grip strength (of four tests, two with each hand)
and mean peak grip strength (best of four tests) were recorded. Mean values of peak grip strength
were 10 to 15% higher than the average maximum grip in all age groups. At all ages girls had a
reduced grip strength compared with boys and although boys manifested a continual,
approximately linear increase in grip strength through all age groups, girls manifested an
approximately linear increase up to 13 years after which mean hand grip usually remained
constant. By the age of 18 years boys had a mean grip strength some 60% higher than girls.
Correlations with height and weight are also presented. 'Handedness' influenced grip strength
and was most noticeable in children aged over 10 years. The clinical use of hand grip strength
centiles for the early indication of neurological and muscular disorders and for following the

Protected by copyright.
natural history of neuromuscular disease is discussed.

Hand grip strength is important as an index of inability to separate isometric from isotonic func-
general health and as a screening test for the tion. A simple, practical, and objective measure of
integrity of both the upper motor neurons and isometric muscle contraction would prove useful in
function of the motor unit. Testing against age the progressive monitoring of muscular strength of
corrected norms for hand grip strength would form a these children.
helpful clinical test in the differential diagnosis of Studies of grip strength (all from the United-
neuromuscular disease and in monitoring the natu- States) have used a variety of dynamometers
ral history of acute and chronic diseases affecting the -pneum,atic and hydrolic manometers, spring
lower motor neuron. loaded resistances, and cable tensiometers. 11-15
Grip strength has come to be regarded as the most Most studies have sampled adults only and none
reliable clinical measure of human strength. It is have included children under 10 years of age.
widely used in adults as an indication of strength in Almost all the dynamometers in clinical use have
fitness testing,23 and as such is seen as the single relatively narrow ranges and test essentially isotonic
item most reasonably representative of total body contraction. Cable tensiometers, despite their many
strength. 6 Occupational therapists have investi- advantages, are cumbersome and not really port-
gated grip strength measures as an index of strength able. The recent development of a portable
and dexterity in university students7 and have used dynamometer, robust yet sensitive, which measures
the test in the assessment of clinical progress after isometric muscle contraction only and has linear
treatment.8 9 It has also been used for more than 30 load response curves at a wide range of loads has
years to assess hand function and general weakness enabled hand grip strength to be measured accu-
in adult patients with rheumatoid arthritis."'1 rately in children' and we report a study to establish
To be useful as a bedside and clinical tool in the hand grip strength norms for Caucasian children.
management of childhood diseases, age centiles for
hand grip strength are required. Tests currently used Subjects and methods
in surveillance of muscular activity in children with
progressive neuromuscular disease (such as the Subjects. The subjects comprised all children
timed stair climb) are complicated by a multiplicity attending the Indooroopilly State School, a primary
of factors such as cardiopulmonary reserve and the school in a middle class suburb of Brisbane (major
453
 Arch Dis Child: first published as 10.1136/adc.59.5.453 on 1 May 1984. Downloaded from http://adc.bmj.com/ on September 3, 2019 at India:BMJ-PG Sponsored.
454 Newman, Pearn, Barnes, Young, Kehoe, and Newman
Table 1 Number of children by age group and sex T s1&s2
Age (yrs) Boys Girls ...& ....S.2..

5 9
lip~ ~
4 G
6 39 44
7 41 30
8 46 45
9 38 45
10 57 58 H Handles: Extensions of steel bar
11 70 52 covered with moulded black
12 71 58 polythene.
13 100 82
14 85 78 T Protective White PVC 48 mm internal
15 79 70 tube sleeve: diameter plumber's pipe.
16 55 58 Affixed to lower bar B2 by
17 31 42 screws P.
18+ 18 12
G Air gap between two arms of hand-piece.
B1 and B2 Bar arms The U-shaped handpieces
urban metropolis of 1 000 000 people) and all pupils were cut from a single plate of
milk steel bar, 12 mm thick.
at the Indooroopilly State High School in the same
area. There were 739 boys and 678 girls. Age and P Screws-pins Attachment of lower hand-
sex distributions are shown in Table 1. The subjects piece B2 to protective sleeve.
were all Caucasian and over 90% were of Anglo S1 and S2 Strain-gauge Affixed at right angles to
Saxon origin. upper arm B1 by Philips
gauge cement (PR 9244/04).

Dynamometer. The dynamometer, which has been L Insulated electrical

leads Connected to Bridge.

Protected by copyright.
described previously and is shown in Figs. 1 and 2,'
operates on the strain gauge principle. It consists of Fig. 2 The hand piece of the strain gauge dynamometer.
a semi rigid steel U shaped handpiece to which a
strain gauge is cemented. When the arms of the
handpiece are squeezed (actual physical movement is a U shaped rigid steel bar in a protective hollow
is very small indeed and does not exceed 8 mm) PVC pipe and two sizes were constructed for use in
changes in electrical resistance are proportional to this study-a smaller one for 6 and 7 year olds and
the force applied. These changes in resistance are the larger one for all other children. Two gauges
measured by incorporating the gauge as one arm of (120 ohm, EA06, 2-065, GF Micromeasurements)
a balanced Wheatstone bridge. There is a peak hold are cemented to the handpiece to provide automatic
reading system whereby rapidly fluctuating forces compensation for environmental, temperature in-
may be frozen at peak by the meter. The handpiece duced changes in resistance. The PVC pipe mount-
ing provides a standard handhold for the contra-
lateral hand/body position during testing.'3 16 With
the subject seated this ensures a standardised
posture for the test which is important. The electri-
cal bridge is a standard commercial model, Strain
Indicator type 1200S with peak hold attachment
(Strainstall, UK). The U shaped handpiece is easily
manufactured from a rod of iron or steel according
to the details shown in Fig. 2. Unlike the common,
commercially available clinical dynamometers (for
example the isotonic, Smedley's type ring grip
instrument), this more accurate instrument records
isometric force which is more appropriate for testing
children. Calibration was performed by hanging
standard weights from each of the handpieces, using
3 kg increments. The functional range of the
instrument-0*5 to 1500 N-encompasses the range
of human hand grip strength from that of a tiny
neonate up to the force deliverable by an Olympic
Fig. 1 The portable, strain gauge dynamometer. weightlifter.
 Arch Dis Child: first published as 10.1136/adc.59.5.453 on 1 May 1984. Downloaded from http://adc.bmj.com/ on September 3, 2019 at India:BMJ-PG Sponsored.
Norms for hand grip strength 455
Procedure. Each school class was given verbal Table 2 Ratio of between subject variance component
instructions and a demonstration before being tested and within subject variance component
and further instructions were provided at the time of
the test. To encourage the children to make their Age (yrs) Boys Girls
best effort in a spirit of competition, each was 5 0-43 1-72
present for the testing of at least one prior subject 6
7
1 51
3-07
1-07
2-17
and was also urged to give a maximum effort. For 8 339 231
the test each subject was seated facing the re- 9 293 3 15
10 398 323
searcher with the equipment mounted on a table 11 2-64 1-73
beside them so that the gauge was visible to both 12 2-51 2-92
and the handpiece, attached to its length of cord, 13 4-51 2-10
14 3-06 2-64
could be passed easily between them. The resear- 15 3-49 1-60
cher demonstrated the hand grip squeeze and the 16
17
1-73
3-08
1-95
1-29
subject was then allowed one practice squeeze. Each 18+ 1-23 1-02
hand was then tested alternately, 10 seconds apart,
and the test repeated, giving two readings for each
hand. It has been shown that in children more than
two squeezes in a series produces a dramatic fatigue groups 7 to 15 years than in older and younger
effect and for this reason we believe that the usual groups. Over all ages there was slightly less con-
three squeeze series (with the mean calculated from sistency among girls than boys.
these) is erroneous. This is especially so in younger Some of the within subject variability was due to
children in our experience. All four readings were differences between right and left grip recordings.
recorded, as well as handedness, height, weight, and When this source of variability was removed from
age. the within subject variance component the ratio was

Protected by copyright.
higher. With the possible exception of five year old
children and those of late teenage the analyses
Results indicated an acceptable degree of consistency in the
hand grip measurements standardised in the format
Repeatability of hand grip measurements. Data on described above. With hand grip strength based on
1417 subjects were analysed. Data for boys and girls four repeat measurements per subject we found that
were analysed separately. Because of volitional even for children outside the 7 to 15 year age group
mode inherent in all strength testing studies, special the data were satisfactory.
interpretive care is needed with results. It is known
from earlier studies that (a) fatigue occurs quickly Age effects on average hand grip recordings. Two
and is universal after five or 6 sequential tests, (b) summary statistics based on the four hand grip
the variation in maximum grip strength at different recordings were calculated; the average and the
test sessions is less than that between sequential maximum of the four recordings. Sample variances
tests in the one sequence, and (c) that maximum and for maximum values were usually larger than the
average strengths (in the sequence of tests) are corresponding figures based on the averages. The
different and should be recorded as separate indices mean value of the maximum hand grips were
of strength. between 10% and 15% higher than the mean value
Hand grip measurements were first analysed to for individual averages for all age groups.
compare the degree of variability between repeat Age related increases in strength are shown in
measurements on the same subjects with the range Fig. 3. Boys showed a continual, approximately
of values found between subjects in the same age linear increase in strength throughout all age
group. This was done by estimating the between and groups. Girls showed an approximately linear in-
within subject components of variance and express- crease up to the age of 13 years after which the mean
ing the figures as a ratio (Table 2). Large values for hand grip remained relatively constant. At all ages
this ratio imply that measurements are repeatable by girls had lower average values than boys and after
individuals and is a desirable feature of this type of puberty this difference increased, until by the age of
research variable. 18 years boys had a mean hand grip 60% higher than
In boys aged 7 to 15 years this ratio took a value girls. Because of (a) the similarity between the
between 2 5 and 4 5. Older and younger boys had averages and peak grip strength and (b) the slightly
smaller ratios implying that individuals are less greater variability associated with peak grip
consistent outside this age range. Girls showed the strengths, mean values only have been used in the
same pattern of greater consistency in the age construction of sex specific centiles.
 Arch Dis Child: first published as 10.1136/adc.59.5.453 on 1 May 1984. Downloaded from http://adc.bmj.com/ on September 3, 2019 at India:BMJ-PG Sponsored.
456 Newman, Pearn, Barnes, Young, Kehoe, and Newman
Boys~ 400.
Mean of (A Girls
,.rnOXImum cri
c
4,00- Man of *0 300-
XP av.Oeraverages
a~~~~~~~~~~~~~da
*a_s _ -- v /it~~~~~~-D-:
300 5 8 11114 15 16 17 18 ~~~~~~~~~~~~
o > _ ,

~~~~~~~~~~~~~~Mean
of
a
Z¢ 200- . . .

V i b averages
100
.0__
in boys and girlsaccording tonage 6 7 8 9 10 11 12 13 14 15 16 17 18
Age (yrs)
5 lb
9 7 10 1 12 t3 1 5 16 17 8
Fig. 4 Median and 5, 25, 75, and 95 centile recordings of
Age Iyrs)
hand grip strength for each year of age in girls.
Fig. 3 Means of average and maximum hand grip strength
in boys and girls according to age. _95th
, 75th
,,, 400. Boysz' __

Variability between subjects' average hand grip g ,' > ~~~~~~~~~25th

recordings. The standard deviations for each of the a.Cdo
300-
C ~~~~5th
peer group distributions increased as the means of .-
'/
% ,
the distributions became larger. Variability ex- a
pressed as coefficients of variations (Table 3) was ca . 200

Protected by copyright.
a
.c
fairly constant for all age groups and both sexes, _M
being between 8% and 15%. For both boys and girls @-
0C4

the most variable age groups were the younger ones p oo-
(up to 8 years of age) and those with the lowest
coefficients of variations were the 9 to 13 year age 6 7 8 9 10 11 12 3 14 15 16 17 18+
groups. Age (yrs)
An indication of the range of hand grip recordings Fig. 5 Median and 5, 25, 75, and 95 centile recordings of
found in this study is given in Figs. 4 and 5, which hand grip strength for each year of age in boys.
show the median and five, 25, 75, and 95 centile
recordings for each year of age, by sex. The five and
95 centile points may be interpreted as normal Influence of height and weight on hand grip. Subject
ranges in the sense that 90% of a similar population height and weight were as strongly associated with
of children would be expected to have hand grip hand grip strength as age (Figs. 6 and 7). This
recordings within these limits. occurred largely because of correlations between
age, height and weight. Even within the age group
classes, however, weight and height showed correla-
Table 3 Coefficients of variation of the average of the four tions with hand grip. For instance the correlation
hand grip recordings by age group and sex
between hand grip and weight within age group
Age (yrs) Boys Girls classes was greater than 0-5 for some and was rarely
(%) (%) less than 0-3.
5 15-9 12 2 These results suggest that at least some of the
6 14-2 14-6 subject to subject variation may be explained if
7 12-9 12 7 either the subjects' heights or weights, or both are
8 11-3 15-5
9 9-8 9-8 taken into account in addition to the subjects' ages.
10 7-7 10-7
10-4
Regression analyses were done to investigate this. It
11 8-8
12 8-9 9-1 was found that linear weight and height terms were
13 8-5 10-6 highly statistically significant for both boys and girls.
14 10-2 11*0 The best of these regression models was one in
15 11*4 11*5
16 13-6 10-8 which both linear height and linear weight terms
17 10-7 11*6 were used. Table 4 summarises the findings in these
18+ 11-5 11-5
models. The implications of height and weight
 Arch Dis Child: first published as 10.1136/adc.59.5.453 on 1 May 1984. Downloaded from http://adc.bmj.com/ on September 3, 2019 at India:BMJ-PG Sponsored.
Norms for hand grip strength 457
400 Table 4 Weight and height regression coefficient and
their standard errors (SE) for average hand grip
Un recordings after first removing year of age effects.
C (Regression coefficients may be used to calculate corrections
0 300 for expected hand grip recordings when a subject's weight
0
u
- Girls and height are known in addition to age)
a.9
Os Coefficient for Coefficient for % of variance
voC 200- weight (SE) (kg) height (SE) (cm) unaccounted for
a by age but
accounted for by
height and weight
(%)
Boys 1-92 (0-26) 1-50 (0-29) 24-7
Girls 0-99 (0-17) 1-13 (0-21) 15-5
For a girl weighing x kg greater than the mean weight for her age and Y cm
nI,
v
taller than the corresponding mean height. the correction to be made to the
100 110 120 130 140 150 160 170 180 expected hand grip in Fig. 2 is 0-99x + 1-13y.
Height (cm)
Fig. 6 Mean recordings of hand grip strength related to
height in boys and girls. Various attempts were made to find other height
and weight based corrections which would improve
further the accuracy of expected hand grip predic-
tions. It was thought that separate linear height and
weight corrections for each age group might achieve
this but, surprisingly, this was not the case. In fact,

Protected by copyright.
the estimated corrections for height and weight for
each group were remarkably similar for all but the
very young and very old age groups. In these age
groups there was some suggestion that height and
weight had slightly smaller influences than at in-
termediate ages. Overall, there was no evidence that
different corrections for each age group would be
beneficial.
We thought that the underlying height and weight
influence on hand grip might be more complex than
had been implied by simple linear corrections, and
in particular that hand grip readings were probably
Weight class (kg)
related to subject strength. Various non-linear
combinations of height and weight were calculated
Fig. 7 Mean recordings of hand grip strength related to as possible crude measures of strength. For instance,
weight in boys and girls. it was conjectured that the ratio of height and weight
might be related to strength by an over-turning
curve. That is, short heavy children at one extreme
effects are best illustrated by example. The coeffi- may be weaker on average than their normally
cients for the girls in Table 4 suggest that for each kg proportioned peers and this might also apply to
that a subject was above or below her age group's those at the other extreme-the tall, thin subjects.
mean weight, her hand grip rose or fell on average A quadratic correction for the height:weight ratio
0-99 of one unit, and for each cm above or below the which allows for such an overturning relation was
average height, the hand grip rose or fell by 1-13 investigated but proved no more advantageous than
units. the more simple correction based on linear weight
Although these height and weight corrections and height adjustments.
were highly significant, they do not account for a
great deal of the subject/subject variation which is Influence of handedness on the right and left hand
unaccounted for by age class (Table 4). It is, grips. The difference between the av rage of the
therefore, debatable whether the slight improve- two right and two left hand grips was c Llcilated for
ment in accuracy compensates for the necessity of each subject. These values are shown in Table 5 and
having to make the height and weight corrections. the extent to which they differ significantly from
 Arch Dis Child: first published as 10.1136/adc.59.5.453 on 1 May 1984. Downloaded from http://adc.bmj.com/ on September 3, 2019 at India:BMJ-PG Sponsored.
458 Newman, Pearn, Barnes, Young, Kehoe, and Newman
Table 5 Mean difjference between the average of the two phiear who tested grip and arm strength of 82% of
right and two left hand grip records. Evidence for a an entire community in Michigan comprising more
difference between hands is indicated by the probability than 6000 subjects.15
that the difference is not different from zero. (Number of Human beinigs have long been and continue to be
subjects in parentheses) fascinated by heir own muscular strength; they seek
Age Right handed subjects Left handed subjects in particular to discover its attainable limits by all
(yrs) manner of weight lifting, athletic, and endurance
Boys Girls Boys Girls feats. Fair grounds have traditionally catered to this
5 3-45 (9) 10-36 (4) - (0) -(() fascination by some measurement device where the
6 4.60 (31)' 1 57 (38) 1-30 (8) -7-77 (6) strong and not so strong may publicly demonstrate
7 -3 01 (36) 5*01 (29)i -2-38 (5) -1)-36 (1) their prowess in raising a strength indicator by the
8 4-66 (45) 0 11 (38) -12-95 (1) -2 13 (7)
9 13-62 (31)§ 11-22 (43)§ -0-96 (7) 0-00 (2) force of a blow. 17 Among the earliest dynamometers
10 2-99 (56) 6-45 (53) -13 42 (1) -16-76 (5) were those devised by the French to test anthropolo-
11 8-07 (64)+ 12-92 (48)§ -20-11 (6)' -6-71 (4)
12 14-34 (65 § 11 63 (49)§ -17-88 (6) -17 50 (9) gical theories of racial differences in strength,'8 and
13
14
9-03 (88)
10-81 (80)t
9.17 (76)
13-36 (75)§
-11.17 (12)
12(07 (5)
-27-93 (6)t
-35-75 (3)t
development of these has continued to the
15 802 (79) 10 25 (68)1 - (0) 10-06 (2) present. 17-21
16 15-77 (54) 13 17 (56)1 -60-33 (1) -70 38 (2) For clinical use it would seem that the average
17 11-44 (29) 13-57 (41) -26 81 (2) -60 34 (1)
hand grip strength based on four alternate hand
18+ 0-00 (17) -4.47 (12) -140-24 (1) - (0)
measurements per subject, classified by age, gives a
P=<0-1; 'P=<0*05; $P=<0-01; iP=<0-001. sufficiently consistent result for most purposes. This
shows an approximately linear increase through all
age groups in boys. Girls record lower hand grip
zero is shown for each age group and sex. Right values than boys at all ages and these values also
handed subjects of both sexes usually recorded follow an approximately linear progression with age

Protected by copyright.
higher right hand readings than left hand ones at all until the thirteenth year when they level out. The
ages but this was most noticeable in children aged 9 discrepancy in hand grip strength beween the sexes
to 17 years. Left handed subjects showed the reverse widens thereafter throughout the teenage years.
and again this was most noticeable in children aged This is in line with the results of Kellor et a!8 who
over 10 years. Fewer significant differences were found that by age 20 men have about twice the grip
found for left handed than right handed children but strength of women. Interestingly, they as well as
this was due to the difference in numbers rather than other researchers have noted that this gap narrows
greater differences in right handed children. Indeed, in later years when women retain more of their grip
the opposite may be true. There was a suggestion strength than men.
that the difference between the left hand and the
right hand in left handed subjects was greater than The authors thank the staff and students of the Indooroopilly
the corresponding reversed difference in right Primary and High Schools, Brisbane for their patience and
handed children. cooperation in this study.

Discussion
References
Although there are numerous published reports on Pearn J, Bullock K. A portable hand-grip dynamometer. Aust
hand strength many cover isolated groups, of Paediatr J 1979;15:107-9.
2 Bookwalter KW. Grip strength norms for males. Research
interest only to specialists in a particular field, or Quarterly 1950;21:249.
sample only small numbers. Studies of a normative 3 Cotton OJ, Johnson A. Use of the T-5 cable tensiometer grip
nature include: Pierson and O'Connell (in Califor- attachment for measuring strength of college men. Research
nia) who tested the grip strength of 299 adult men Quarterly 1968;41:454-6.
4 Wessel JA, Nelson RC. Relationship between grip strength and
drawn from populations representing different age achievement in physical education among college women.
and physical activity requirements;" Schmidt and Research Quarterly 1961 ;32:244-8.
Toews who presented the results of grip strength 5 Tinkle WF, Montoye HJ. Relationship between grip strength
testing for 1128 men and 80 women employed by an and achievement in physical education among college men.
Research Quarterly 1961;32:238-43.
American corporation;'2 Swanson, Matev, and de 6 Heyward V, McCleary L. Analysis of the static strength and
Groot who established grip and pinch strength for relative endurance of women athletes. Research Quarterly
100 adults in an attempt to add to the factual base 1975;48:703-10.
7 Nwuga VC. Grip strength and grip endurance in physical
for reconstructive surgery of the hand;'3 Beasley therapy students. Arch Phys Med Rehabil 1975;56:296-300.
who sampled grip strength in 1524 American adult 8 Kellor M, Frost J, Silberberg N, Iversen I, Cummings R. Hand
men and 1238 women;'4 and Montoye and Lam- strength and dexterity. Am J Occup Ther 1971;25:77-83.
 Arch Dis Child: first published as 10.1136/adc.59.5.453 on 1 May 1984. Downloaded from http://adc.bmj.com/ on September 3, 2019 at India:BMJ-PG Sponsored.
Norms for hand grip strength 459
17
Agnew PJ, Maas F. Hand function related to age and sex. Arch Pearn J. Two early dynamometers. Neurol Sci 1978;37:127-34.
Phys Med Rehabil 1982;63:269-71. 18
Pearn J. Some early experiments on the measurement of human
'1 Myers DB, Grennan DM, Palmer DG. Hand grip function in strength in Port Jackson and Van Diemen's land. Med J Aust
patients with rheumatoid arthritis. Arch Phys Med Rehabil 1978;2:167-9.
1980;61:369-73. 19 Bechtol CO. The use of a dynamometer with adjustable hand
" Pierson WR, O'Connell EK. Age, height, weight and grip spacings. J Bone Joint Surg 1954;34-A:820.
strength. Research Quarterly 1961;33:439-43. 20 Lewey FH, Kuhn WG, Juditski JT. A standardized method for
12 Schmidt RT, Toews RPT. Grip strength as measured by the assessing the strength of hand and foot muscles. Surg Gynecol
Jamar dynamometer. Arch Phys Med Rehabil 1970;51:321-7. Obstet 1947,85:785.
13 Swanson AB, Matev IB, de Groot MD. The strength of the 21 An KN, Chao EYS, Askew U. Hand strength measurement
hand. Bull Prosthet Res 1970;10:145-54. instruments. Arch Phys Med Rehabil 1980;61:366-8.
14 Beasley WC. Efficient estimators of normal adult grip strength.
Arch Phys Med Rehabil 1973;54:573. Correspondence to D Newman, Department of Speech and
15 Montoye HJ, Lamphiear DE. Grip and arm strength in males Hearing, University of Queensland, St Lucia, Q4067, Brisbane,
and females, age 10 to 69. Research Quarterly 1975;48:109-20. Australia.
16 Clarke HH. Recent advances in measurement and under-
standing of volitional muscular strength. Research Quarterly Received 23 January 1984
1956;27:203.

Protected by copyright.
Twenty five years ago
The fertility of mothers of diplegic children and the fate of their conception
T T S INGRAM (Edinburgh)
'The reproductive performance of 76 mothers of diplegic children was compared with that of mothers of
children suffering from other forms of cerebral palsy and of mothers from the general population. The
mothers of diplegic children had had fewer pregnancies than those in the other groups though their average
age was slightly greater. There were fewer conceptions in the years immediately preceding and following the
birth of the diplegic child than at earlier or later periods.
Excluding from consideration the pregnancies which resulted in the birth of the patients, only 62% of
their conceptions had produced healthy siblings who survived at the time of the study. In a high proportion
there had been abnormalities of pregnancy, labour or delivery.
There appears to be an aetiological relationship between diplegia and impaired reproductive performance
of the mother.'
(Now termed 'relative infertility'. Murphy wrote about this in 1947 in his book Congenital Malformations,
and the Baltimore group (Lilienfield and Parkhurst) termed it 'a continuum of reproductive casualty'. Yet
there are still doctors who ascribe every case of cerebral palsy to birth injury-presumably because the child
was born and birth is dangerous-ignoring all the prenatal factors. RONALD ILLINGWORTH).
Archives of Disease in Childhood 1959;34:357.