Size at birth and its relation to muscle strength in young adult women
Hazel Inskip, MRC Epidemiology Resource Centre, Southampton General Hospital, University of Southampton, Southampton SO16 6YD, UK.
(fax: 023 8070 4021; e-mail: firstname.lastname@example.org).
Objective. To assess the relationship between development in utero, assessed by birth weight, and muscle strength in young adult women as assessed by grip strength.
Methods. A total of 1563 participants aged 20–40 years in the Southampton Women's Survey had their grip strength measured during pregnancy. At recruitment to the survey the women had been asked to recall their birth weight or obtain it from their parents. For 536 women born in Southampton, birth weight was obtained from hospital records. Grip strength was related to birth weight using multiple linear regression analysis, adjusting for age, height, weight and reported physical activity.
Results. Grip strength increased with age, height, weight, physical activity and birth weight. In the mutually-adjusted model, grip strength increased by 1.10 kg per kilogram of birth weight (95% CI: 0.58–1.61 kg). In women with hospital birth weight data the relationship strengthened to 1.44 kg per kilogram of birth weight (95% CI: 0.50–2.38 kg).
Conclusions. Grip strength in women in their twenties and thirties is at or approaching its peak. The association between grip strength and birth weight was remarkably similar to findings from other studies of women at younger and older ages. This indicates that in utero development has consequences for muscle strength throughout the life course, even allowing for the increase to peak muscle strength and then its decline as a woman ages.
Grip strength is an important predictor of current and future health. A number of studies have shown that lower grip strength in men and women is associated with an increased risk of disability, morbidity and mortality in later life [1–6], as well as with impaired quality of life particularly in older people [7–10]. Furthermore it has been shown to have an impact on receipt of health care; lower grip strength is associated with longer lengths of stay in hospitalized older people [11, 12].
The findings of consistent associations between grip strength and later health outcomes have led to a search for earlier influences on grip strength. Recent epidemiological studies have shown that birth weight is associated with grip strength in those over 50 years of age [13–16]. It is, however, unclear whether the association is a result of prenatal development affecting the decline in grip strength in later life or by influencing the development of peak grip strength in early adulthood.
We therefore examined the influences on grip strength in a group of young women who were participating in the Southampton Women's Survey (SWS) . We aimed at relating the women's birth weights to their grip strength adjusting for other known influences on grip strength.
Subjects and methods
Between 1998 and 2002, over 12 500 women were recruited to the SWS. They were interviewed in their homes to assess health, diet, lifestyle, body composition, physical activity and socio-economic circumstances. These women were then followed through their subsequent pregnancies and the offspring are followed through childhood. The aim was to assess the influence of preconceptional and antenatal factors on the growth and development of the fetus and then the child. Full details of the study have been published previously .
The measurements obtained at the interviews that took place before the women were followed through their subsequent pregnancies included height and weight. Height was measured with a portable stadiometer (Harpenden; CMS Weighing Equipment Ltd, London, UK) to the nearest 0.1 cm with the head in the Frankfort plane. Weight was measured with calibrated electronic scales (Seca, Hamburg, Germany) to the nearest 0.1 kg and the women were asked to remove their shoes and any heavy items of clothing or jewellery.
Physical activity was assessed using a structured questionnaire based on a validated questionnaire developed by Godin and Shephard . It included a question about how often the woman had taken strenuous exercise in the past 3 months. The answers to this question were categorized into three groups according to whether the woman never took strenuous exercise, took less than 1 h, or one or more hours of strenuous exercise per week.
At the initial interview, the women were asked whether they knew their own birth weight. Those who were uncertain were asked to contact their parents to see whether the information on birth weight could be obtained more precisely. Women who said that they were born in Southampton had their birth weight abstracted from the original hospital birth records.
During pregnancy, the women visited the SWS Ultrasound Unit at the Princess Anne Hospital at 11, 19 and 34 weeks’ gestation. At the 19-week visit, the women's grip strength was measured using a Jamar handgrip dynamometer (Promedics, Blackburn, UK). Three measurements of each hand were taken and the maximum of all six measurements was used in the analysis . Grip strength measurements were introduced to the 19-week pregnancy assessment part way through the study so were only available for those women whose visits took place between 18 February 2002 and 28 September 2005.
Multiple regression analysis was used to relate the woman's grip strength to her birth weight and information collected at the initial recruitment interview. The particular factors that were considered as being potential influences on grip strength were height, weight, physical activity and age, and these factors were explored in the regression modelling process. In order to assess the strength of the association between birth weight and grip strength the analysis was repeated using standard deviation scores (SDS) for both birth weight and grip strength.
Ethics Committee approval
The study was approved by Southampton and South West Hampshire Local Research Ethics Committee, and written consent was obtained from the participants.
Grip strength measurements were available for 1563 of the 1687 women who attended for a 19-week scan between 18 February 2002 and 28 September 2005 (93%). Data were excluded for one woman with rheumatoid arthritis whose grip strength measurements were all less than 5 kg, leaving 1562 measurements available for analysis. Of these women, 1370 were able to recall their birth weight or obtain it from their parents, and 536 were born in Southampton and had records of their birth weight available in hospital records; 447 women had both recalled and recorded birth weights.
The women who participated in the grip strength measurement were 20–40 years old at the time of measurement. Characteristics of the women are presented in Table 1 along with those from the entire cohort of 12 583 women who participated in the initial recruitment interview for the SWS.
Table 1. Characteristics of the women with grip strength measurements and the entire SWS cohort
|Highest educational attainment, n (%)|
| None||38 (2.4)||724 (5.8)|
| GCSE grade D-G||141 (9.1)||1393 (11.1)|
| GCSE grade A*-C||439 (28.2)||3323 (26.6)|
| A levels||489 (31.4)||3735 (29.8)|
| Diploma||94 (6.0)||727 (5.8)|
| Degree or above||356 (22.9)||2616 (20.9)|
|Ethnic group, n (%)|
| White||1512 (96.2)||11,813 (94.0)|
| Nonwhite||50 (3.8)||748 (6.0)|
|Height, cm [mean (SD)]||163 (6.4)||163 (6.4)|
|Weight, kg [mean (SD)]|| 67.6 (13.9)|| 67.4 (14.4)|
|Strenuous exercise in past 3 months, n (%)|
| Never||507 (32)||4534 (36)|
| <1 h week−1||545 (35)||3852 (31)|
| ≥1 h week−1||509 (33)||4125 (33)|
|Birth weight, kg [mean (SD)]|| 3.24 (0.56)|| 3.23 (0.59)|
|Age at initial interview in years [mean (SD)]|| 27.8 (3.9)|| 28.2 (4.2)|
|Age at grip strength measurement in years [mean (SD)]|| 30.6 (3.8)||N/A|
The women with grip strength measurements had on average attained slightly higher levels of educational qualifications and were marginally more likely to be white. Other characteristics were similar between the two groups. The entire cohort of SWS women has previously been shown to be broadly representative of women of this age group in Britain today . Although the percentage in ethnic minority groups is lower than in the country as a whole (4% vs. 12%), 31% of the SWS women were smokers, which is comparable to the figure of 33% for women of a similar age in Britain . Southampton as a city is more deprived than England and Wales as a whole  but 21% of the SWS women had university degrees making the higher educational profile similar to that found in the Labour Force Survey for England in which 22% of women of working age had degrees .
The mean grip strength was 32.2 kg (SD 5.9 kg) and the measurements ranged from 10 to 53 kg. Univariate analysis showed strong relationships with all variables considered (Table 2).
Table 2. Unadjusted and adjusted relationships between grip strength and recalled birth weight, age, height, current weight and taking strenuous exercise
|Birth weight (kg)||2.16||1.62–2.70||1.10||0.58–1.61|
|Takes strenuous exercise|
| <1 h week−1||0.34||−0.37 to 1.06||0.08||−0.62 to 0.77|
| ≥1 h week−1||1.37|| 0.64–2.10||0.91|| 0.21–1.61|
Grip strength increased with age, height, pre-pregnant weight, taking of strenuous exercise and reported birth weight. Notably, grip strength increased by 2.16 kg for every kilogram increase in birth weight (95% CI: 1.62, 2.70 kg). Mutual adjustment attenuated the regression coefficients for all these variables. After adjustment, the coefficient for birth weight was reduced such that grip strength increased by 1.10 kg per kilogram of birth weight (95% CI: 0.58, 1.61 kg). Grip strength was also associated with educational attainment but after adjustment for the other factors this relationship disappeared.
Amongst the 447 women with both reported and recorded birth weight measurements, the correlation between the two assessments was 0.88. On average, the recorded birth weights exceeded the reported values by 0.03 kg, and the standard deviation of the difference between the two measurements was 0.36 kg. Table 3 shows the results for the analysis of the 536 women for whom recorded birth weights were available.
Table 3. Unadjusted and adjusted relationships between grip strength and birth weight, age, height and taking strenuous exercise amongst 536 women for whom hospital birth weight data were available
|Birth weight (kg)||2.92||1.96–3.88||1.44||0.50–2.38|
|Age (years)||0.01||−0.12 to 0.15||−0.02||−0.15 to 0.10|
|Weight (kg)||0.07||0.04–0.11||0.02||−0.01 to 0.05|
|Takes strenuous exercise|
| <1 h week−1||0.29||−0.91 to 1.48||−0.02||−1.11 to 1.07|
| ≥1 h week−1||1.49||0.28–2.70||1.31||0.22–2.40|
Larger regression coefficients are seen for the relationship between grip strength and birth weight when the more accurate recorded birth weight values are used. After adjustment for the confounding variables, grip strength was found to increase by 1.44 kg per kilogram of birth weight. The 95% CI (0.50, 2.38 kg) around this estimate is however wider than before due to the smaller number of women in the analysis. It is noteworthy that in this subset of women, age was not associated with grip strength, and in the mutually-adjusted model the woman's pre-pregnant weight ceased to have a significant relationship with grip strength. Within the 447 women with both reported and recorded birth weight measurements, the relationship between grip strength and reported birth weight was very similar to that in the whole group (adjusted regression coefficient 1.11 kg per kilogram of birth weight, 95% CI: 0.22, 2.01 kg).
The fully adjusted models explained approximately 20% of the variation in grip strength measurements. Recalled birth weight in the unadjusted analysis explained only 4% of the variation but this increased to 6% when the analysis focused only on those with hospital birth records of their birth weight. This compares with the 17% explained by height, which is known to be the major adult determinant of grip strength.
In the analyses using SDS, grip strength increased by 0.10 SDS per SDS of birth weight in the analysis of all women and by 0.12 SDS per SDS of birth weight in those women for whom hospital records of birth weight were available.
In this study, we have shown that grip strength in young adult women is associated with their birth weight independent of their current age, height, weight and physical activity. On average, grip strength increased by 1.1 kg per kilogram increase in reported birth weight, with a larger effect size being observed when using the birth weight as recorded at the time of birth.
Grip strength increased with the size of the woman. Thus taller, heavier women tended to have higher grip strength. Within this population of young women, grip strength also increased with age, though the effect size was small, and in the subset of women with a hospital record of birth weight there was no relationship between age and grip strength. Grip strength appears to be at its greatest in early adulthood, though estimates vary as to the age at which the peak is achieved [22, 23]. Clearly it declines at older ages [22–24], but our data suggest that it may increase slightly in women through their twenties and early thirties, though we do not have longitudinal measurements on these women. Nonetheless, the lack of an association between grip strength and age in the subset of women for whom hospital birth records were available does raise the possibility that the association found in the larger group is an artefact.
Comparison with other studies
These results are similar to findings in the National Survey of Health and Development and the Hertfordshire Cohort Study [13–16]. For example, in the National Survey grip strength increased by 1.3 kg per kilogram of birth weight in adults aged 50 years, a finding comparable to those seen here. The similarity of these findings suggests that developmental influences affect peak grip strength attained in early adulthood rather than simply having an effect on decline in muscle function in later life.
Three studies to date have considered the influence of birth weight on grip strength in children and adolescents, though these compared groups of low birth weight babies with those of normal birth weight (NBW). Ford et al.  assessed 5-year-old children, 24 of whom were very low birth weight (VLBW) children and 18 were NBW. The left and right hands were considered separately and a difference of about 20 N (which converts to approximately 2 kg) was seen between the VLBW children (mean birth weight 1.2 kg) and NBW children (mean birth weight 3.5 kg).
More recently, Rogers et al.  compared grip strength at age 17 years between a group of 53 extremely low birth weight children with mean birth weight 0.72 kg and a group of 31 term-born controls whose mean birth weight was 3.5 kg. Differences of 10 and 11 kg for right and left hand grip strength, respectively, were seen between the two groups for boys but smaller differences of 2 and 1 kg, respectively, were observed in girls.
In both the above reports, groups with particularly low birth weights were studied and it is possible that other factors detrimental to development are operating when babies are born so small. Nonetheless, it does appear that the differences between those born small and those born of normal weight are comparable to the results seen within the NBW spectrum in our study and in those of older adults; in all studies grip strength increased by approximately 1 kg per kilogram of birth weight.
Martorell et al. , however, examined 169 male and 162 female adolescents grouped according to their weight at birth into those with intrauterine growth retardation (IUGR), or of NBW divided into two equal-sized groups to form middle and upper birth weight groups; mean birth weights of the three groups were 2.36, 2.81 and 3.38 kg respectively. Those in the IUGR group were not of such low birth weight as in the two studies described above, but the 19 girls in this group had grip strengths approximately 3 kg lower than the other two groups. There was, however, little difference between the middle and upper birth weight groups. These results do not give as consistent a message about the relationship between birth weight and grip strength as in the studies described above, but it is clear that those with IUGR in this study had lower grip strength in adolescence than those of NBW.
Strengths and weaknesses of the study
We have studied a group of women who were broadly representative of the general population. The range of our reported birth weights was wide (0.9–5.4 kg), though the subset with birth weights obtained by records compiled at the women's births did not include any women with birth weights lower than 1.5 kg (range 1.5–5.6 kg). In such a sample we have found results that are similar to the other studies to date that have explored the relationship between birth weight and grip strength at various ages.
There are however some limitations in our study. We focused on young women in a limited age range. Additionally, these women were assessed in mid-pregnancy. No studies have reported on the changes in grip strength associated with pregnancy and it is possible that muscle strength alters at this time. Mathiowetz et al.  reported normative data for adult women in 5-year age groups for the right and left hand separately. Between 20 and 34 years the mean grip strengths of the right hand varied between 70.4 and 78.7 lb (31.9 and 35.7 kg). These are close to our mean value of 32.2 kg indicating that if grip strength alters in pregnancy then the change is small. However, even if grip strength is affected in this way, it seems unlikely that the association between grip strength and the woman's own birth weight would be distorted to any great extent by effects on grip strength induced by pregnancy.
Interpretation and implications
Although the relationship between birth weight and grip strength is consistent across a number of studies, birth weight only explains a small percentage of the variation in grip strength in the population and an increase in birth weight of 1 SD is only associated with a increase of approximately 0.1 SD in grip strength. However, birth weight is a poor summary measure of fetal growth and development and we are using it as a proxy for the early development of skeletal muscle in the fetus. Focusing on the women who had more accurate birth weights led to a strengthening of the association between grip strength and birth weight and a greater percentage of the variation explained. Identifying a more direct marker of prenatal muscle growth and development would be likely to strengthen the association further, though this is not inevitable . We nonetheless suggest that the influence of early life factors on the development of the muscle strength later in life is likely to be greater than observed in this and other studies. Thus, improving the intrauterine environment should prove a novel means of optimizing muscle strength in later life.
In conclusion, we believe that the consistency of these findings indicates that factors operating in utero influence early human muscle growth and development. This has long-term consequences for muscle strength throughout the life course, even allowing for the increase to peak muscle strength and then its decline as a woman ages.
Conflict of interest statement
We thank the SWS staff who collected and processed the data. The study was funded by the Medical Research Council, the University of Southampton and the Dunhill Medical Trust. The funding agencies had no role in the conduct or reporting of this research.