Adult physical performance is recognized as a marker of both current and future health. Increased cardiorespiratory fitness is independently associated with risk reductions in all-cause mortality (6,9,31), cardiovascular disease (17,32,37), type 2 diabetes (10,36), and stroke (16) in a dose-response manner. Other measures of adult physical performance such as muscle strength have also been associated with reduced disease risk, including sarcopenia (28) and all-cause mortality (23,24).
Whereas physical performance measures, such as muscle strength and aerobic fitness, are likely to be influenced by both genetic and environmental factors, including regular exercise and the total amount of physical activity, there is evidence to suggest that factors in early life may act as a biological determinant of various elements of adult physical performance.
For example, lower birth weight has been associated with impaired aerobic fitness (2,11,25) and higher resting pulse rate in adults independently of adult body size (21). Higher birth weight has been consistently associated with improved handgrip strength in adolescents (20), in young adult females (8), and in older adults (8,13,15,27,39), suggesting that these early life factors may have long-lasting influences on physical performance throughout the life course.
In a small number of studies, infant motor development has also been linked to future physical performance. Age at first walking was associated with adult physical performance in terms of chair rising and standing balance at age 53 yr (14). Delayed age at first walking and standing unaided has also been associated with reduced handgrip strength in both males and females at age 53 yr, although this did not seem to be independent of birth weight (15). These studies suggest that the timing of infant motor development may influence physical performance in later life; however, further research is needed to investigate the influence of infant motor development for other physical performance outcomes, such as aerobic fitness, and whether these associations are independent of birth weight and growth in infancy.
The aim of the present study was to examine the independent associations among birth weight, infant weight gain, and infant motor development with differing adult physical performance outcomes, in terms of muscular strength, muscular endurance, and aerobic fitness, in a large prospective population-based birth cohort. Outcomes measures collected at age 31 yr provides information on physical performance in adulthood before the age-related decline in physical performance, which may have been the case in some previous studies.
The study population was composed of individuals from The Northern Finland Birth Cohort of 1966 (NFBC 1966) who were recruited from all pregnancies with a birth expected between January 1 and December 31, 1966, within Finland's two most northerly provinces, Oulu and Lapland. This cohort has been described previously (22). Briefly, the cohort consisted of 12,058 births, an estimated 96.3% of all live births during the qualifying period. At 1 yr, 10,322 of these children were assessed for standard anthropometric measures and parental report of age at reaching developmental milestones.
At age 31 yr, all of those still alive (n = 11,541) were invited to a follow-up study involving questionnaires and those living in the capital area of Helsinki or within the local area in Northern Finland (n = 8463) were also invited to a clinical examination, of whom 6033 (71%) attended a health center. The clinical examination was carried out by trained research nurses and involved three measures of physical performance: muscle strength, muscle endurance, and aerobic fitness (34). The most common reasons for incomplete performance data were ill health and pregnancy.
Analysis was restricted to singleton infants born at ≥36 wk of gestation. Infants measured at >500 d (n = 22) or <300 d (n = 3) at the year 1 measurement were excluded. Three adult women with extreme short stature (height <140 cm) were excluded. A final data set of 4304 with complete data on birth weight and at least one measure of physical performance was used for subsequent analysis.
The ethics committee of the University of Oulu and The Finish Institute of Occupational Health approved the study, and all participants gave written informed consent.
Early life variables
Birth length and weight were measured by midwives, with 99% of births being in hospitals. Infant weight and length were measured at regional child welfare centers, with supplementary data being collected during a special examination by public health nurses. Information on infant motor development was collected on 91% of all infants, with 95% of these being at least 11.5 months old at the time of the visit. Age at first "walking supported" and age at first "standing unaided" was collected using parental report at the time of the 1 yr visit.
Performance variables at age 31 yr
Muscle strength was measured by recording maximal isometric handgrip strength in the dominant hand using a hand dynamometer (Newtest, Oulu, Finland) while participants were in standing position with the elbow extended and the arm hanging next to, but not touching, their trunk or hip. Allowance was made for hand size by adjusting the width of the grip as appropriate. The maximum of the three measurements was used to allow for improvements of technique with repeated tests.
Muscle endurance was measured by a trunk endurance test (1) where the participant was lying prone on a stand and asked to maintain the upper body level with the hips, whereas only the lower body was supported below the anterior superior iliac spine. This position was maintained for as long as possible or up to a maximum duration of 240 s.
Aerobic fitness was estimated from the postexercise heart rate measured using a heart rate monitor (Fitwatch; Polar, Finland) immediately after a standard 4-min submaximal step test at the stepping rate of 23 steps per minute, using a single step (step height of 33 cm for females and 40 cm for males). Predicted aerobic fitness assessed by the step test has previously been validated against directly measured maximal oxygen consumption in a subset of the population (33).
Potential confounding variables
Social class data of the childhood family were collected by self-reported questionnaire, as routinely defined in Finland into four categories: 1) I and II skilled professionals, 2) III skilled workers, 3) IV unskilled workers, and 4) farmers (with any farmer of large ranches being classified as class III). The father's social class category in 1966 was used to adjust for social class in early life in subsequent multiple regression models.
At age 31 yr, participants were asked to report their educational status. This was classified into four categories: 1) university degree, 2) higher education (polytechnic or vocational college), 3) medium education (vocational school or course), and 4) no vocational education. This variable was used as a proxy measure of adult socioeconomic status and was used as a confounder in subsequent multiple regression models.
Calculations and statistics
Mean and SD for size at birth were very similar to the current UK growth reference (3). At age 1 yr, mean weight and height were slightly higher than the 1990 UK growth reference (mean SD age- and sex-adjusted scores were +0.34 for weight and +0.29 for height). Using data on birth weight and body weight at age 1 yr, we generated internal SD scores by subtracting the sample mean from the individual mean and then by dividing by the sample SD for weight at birth and at 1 yr. Rate of infant weight gain was calculated as the change between SD score at birth and SD score at 1 yr in each individual. A change in weight SD score of 0.67 SD represents the distance between each displayed centile line on a standard growth chart (i.e., 2nd, 9th, 25th, 50th, 75th, 91st, and 98th centile lines).
Positively skewed variables (weight at age 31 yr) were log-transformed to give an approximately normal distribution. Birth weight (SD score) and weight gain (change in SD score between 0 and 1 yr) were analyzed as continuous variables. Age at standing unaided was categorized into quintiles for illustrative purposes for the figures (<9, 9, 10, 11, or ≥12 months).
Initially, correlations between early life variables and adult performance variables were investigated using simple bivariate correlations, including investigating correlations with adult body size and potential confounding variables.
For the trunk endurance test, 27.4% of all participants (18.8% of men and 35.8% of women) achieved the maximum time of 240 s, so were, in effect, censored rather than recording their true maximum performance. To assess the impact of this censoring, the analyses for trunk endurance were repeated using Cox regression. The results and significance levels were similar for both models (data not shown), and we present only results from the linear regression models for consistency across all three outcome variables.
The models were initially investigated by sex separately; however, there was no substantial difference in the regression coefficients between the sexes and formal statistical tests for sex interaction were not significant, so the remaining analyses were performed in the whole data set adjusting for sex. To allow comparisons with other studies, data for sex-specific regression models have also been provided.
All associations between early life variables and adult physical performance variables were analyzed adjusting for sex, gestational age, father's social class in 1966, and adult educational status as potential confounders. The regression models were also investigated using a quadratic term for each of the exposure variables; however, there was no evidence for any nonlinear associations.
Any significant associations were thereafter further adjusted for adult size (height and log-transformed weight) to investigate whether the associations between early life variables and performance outcomes at age 31 yr might be mediated by adult body size. Adult body size was entered as separate weight and height variables to more closely adjust for separate height and weight effects. However, repeating the models adjusting for a single variable of adult body mass index (BMI = weight/height2) did not alter the results (data not shown).
Finally, significant exposure variables were entered into a stepwise regression model to investigate the independent contributions of birth weight, rate of infant growth, and infant motor development on each of the physical performance measures. These models were adjusted for sex, gestational age, father's social class, and adult educational status by forced inclusion of those variables. All statistical analyses were carried out using SPSS version 14.0 (SPSS, Inc., Chicago, IL).
A summary of the main anthropometric measures at birth, at 1 yr, and at age 31 yr along with outcome measures of physical performance is presented in Table 1. Women showed lower physical performance than men in terms of handgrip strength (mean value of 28 kg compared with 50 kg in men) and a slightly higher post step test heart rate (150 beats·min−1 compared with 147 beats·min−1 in men), although women outperformed men in the trunk endurance test (180 s compared with 162 s for men).
Correlations between variables
As expected, birth weight was highly correlated to birth length (r = 0.74, P= 0.01) and was negatively correlated to infant weight gain (r = −0.57, P = 0.01). There was also a weak but significant inverse correlation between birth weight and age at walking supported (r = 0.06, P < 0.001) and age at first standing unaided (r = −0.06, P < 0.001), suggesting that higher birth weight was associated with earlier motor development. Infant weight gain during the first year was not associated with motor development.
Gestational age was not correlated with birth weight or birth length; however, it was positively correlated with infant weight gain (r = 0.07, P < 0.001). Greater gestational age was also correlated with earlier infant motor development (walking supported, r = −0.08, P < 0.001; and standing unaided, r = −0.11, P < 0.001). There were no correlations between gestational age and adult physical performance variables.
Handgrip was highly correlated with adult height (r = 0.73, P < 0.001) and with adult weight (r = 0.55, P < 0.001). Adult height was inversely and significantly correlated with post step test heart rate (r = −0.06, P < 0.001), whereas adult weight was positively correlated with post step test heart rate (r = 0.13, P < 0.001), suggesting that height is correlated with improved aerobic fitness, whereas increasing weight was correlated with lower aerobic fitness. Trunk endurance time negatively correlated with height (r = −0.10, P < 0.001) and negatively correlated with weight (−0.41, P < 0.001), possibly because of the nature of this test in that it requires the volunteer to hold their upper body horizontally without support.
Lower father's social class at birth (i.e., in 1966) was weakly but significantly correlated with increased weight gain between 0 and 1 yr (r = 0.04, P = 0.024) and with slower motor development in terms of walking supported (r = 0.06, P < 0.001). Lower father's social class was also correlated with poorer physical performance in terms of higher post step test heart rate (r = 0.05, P = 0.001) and lower trunk endurance (r = −0.3, P = 0.032) at age 31 yr. On the other hand, lower adult educational status at age 31 yr was correlated with increased handgrip (r = 0.10, P < 0.001) and reduced trunk endurance (r = −0.16, P < 0.001).
Birth size and adult muscle strength
Birth weight was positively associated with adult handgrip strength (Table 2; β = 1.42, 95% confidence interval (CI) = 1.19-1.65, P < 0.001), after adjusting for sex and gestational age, and this association appeared to be linear across the range of birth weights (Fig. 1). This equates to a 3.0-kg increase in handgrip per 1-kg increase in birth weight (adjusted for gestational age and sex) and to a 1.6-kg increase per 1-kg increase in birth weight when further adjusted for adult height. Adjustment for adult body size (height and log-transformed weight at age 31 yr) attenuated the association by approximately 50%, but the independent effect of birth weight on adult handgrip strength remained significant (β = 0.68, 95% CI = 0.45-0.92, P < 0.001). When analyzed by sex separately, males showed a slightly greater increase in handgrip with increasing birth weight SD score (Table 3; β = 0.73, 95% CI = 0.35-1.10, P < 0.001) compared with women (β = 0.65, 95% CI = 0.36-0.94, P < 0.001).
Birth length was also positively associated with handgrip strength (Table 2; β = 1.32, 95% CI = 1.09-1.55, P < 0.001), and this association was attenuated after adjusting for adult body size (i.e., height and weight) but still remained significant (β = 0.44, 95% CI = 0.20-0.68, P < 0.001). Again, there was a greater effect size in males (Table 3; β = 0.51, 95% CI = 0.13-0.90, P = 0.009) compared with that in females (β = 0.34, 95% CI = 0.05-0.64, P = 0.024). Infant weight gain between birth and 1 yr was only associated with handgrip strength in females (Table 3) with lower handgrip associated with increased weight gain between 0 and 1 yr (β = −0.34, 95% CI = −0.59 to −0.10, P = 0.006).
Infant motor development and adult muscle strength
Earlier infant motor development was associated with greater handgrip strength at age 31 yr (Table 2; walking supported, β = −0.22, 95% CI = −0.39 to −0.04, P = 0.015; and standing unaided, β = −0.37, 95% CI = −0.58 to −0.15, P = 0.001), after adjusting for gestational age and sex. These associations were slightly attenuated after further adjusting for father's social class, adult educational status and adult body size (β −0.19, 95% CI = −0.36 to −0.02, P = 0.032 and β −0.31, 95% CI = −0.52 to −0.10, P = 0.004; Table 2). The associations between infant motor development and adult muscle strength showed similar effect sizes between males and females (Table 3).
Birth size and adult muscle endurance
Birth weight was borderline associated with trunk muscle endurance at age 31 yr (β = 1.71, 95% CI = 0.00-3.43, P = 0.051) after adjusting for sex and gestational age. After additional adjustment for social class, educational status, and adult body size, this association strengthened (β = 3.82, 95% CI = 2.17-5.48, P ≤ 0.001; Table 2). This was comparable between males and females (Table 3; β = 3.44, 95% CI = 1.15-5.73, P = 0.003 and β = 4.06, 95% CI = 1.67-6.46, P = 0.001, respectively). Infant weight gain was significantly associated with impaired muscle endurance after adjusting for sex and gestational age (β = −1.83, 95% CI = −3.35 to −0.32, P = 0.018; Fig. 2). This association remained significant after further adjusting for father's social class, adult educational status, and adult height but was attenuated by additional adjustment for adult body weight (β = −0.62 95% CI = −2.02 to 0.77, P = 0.380).
Infant motor development and adult muscle endurance
Earlier infant motor development was significantly and inversely associated with muscle endurance at age 31 yr (walking supported β = −1.51, 95% CI = −2.78 to −0.25, P = 0.019; standing unaided β = −2.17, 95% CI= −3.73 to −0.60, P = 0.007) after adjusting for gestational age and sex. These associations remained relatively unchanged after further adjusting for father's social class, adult educational status, and adult height but strengthened after further adjusting for adult weight (walking supported β = −2.22, 95% CI = −3.38 to −1.06, P < 0.001; standing unaided β = −3.31, 95% CI = −4.76 to −1.86, P < 0.001; Table 2). The associations between infant motor development in terms of walking supported and muscle endurance were comparable between males and females (Table 3) with a greater effect size seen in females for standing unaided (β = −4.12, 95% CI = −6.13 to −2.12, P < 0.001) compared with males (β = −2.32, 95% CI = −4.42 to −0.13, P = 0.030).
Birth size and adult aerobic fitness
Higher birth weight was weakly and inversely associated with post step test heart rate, indicative of higher aerobic fitness (Table 2; β = −0.56, 95% CI = −1.10 to −0.02, P = 0.043) when adjusted for sex and gestational age. This association strengthened after adjusting for father's social class, adult educational status, and adult height (β = −0.80, 95% CI = −1.39 to −0.21, P = 0.007) with a further increase when additionally adjusted for adult weight (β = −1.22, 95% CI= −1.79 to −0.64, P < 0.001). These associations were similar between males and females (Table 3).
In contrast, greater infant weight gain was associated with higher post step test heart rate, indicative of lower aerobic fitness (β = 0.70, 95% CI = 0.22-1.17, P = 0.004) after initial adjustment for gestational age and sex. There was little change in this association after adjusting for father's social class, adult educational status, and adult height; however, this association was attenuated and no longer statistically significant after further adjusting for adult weight (β = 0.45, 95% CI = −0.04 to 0.93, P = 0.071).
Infant motor development and adult aerobic fitness
Earlier infant motor development defined as standing unaided was associated with a lower post step test heart rate indicative of higher aerobic fitness (β = 0.87, 95% CI = 0.38-1.36, P = 0.001) after initial adjustment for sex and gestational age. This association appeared to be linear across the range of infant motor development ages (Fig. 3). There was very little change with further adjusting for father's social class, adult educational status, and adult height, with a slight increase in the association after additional adjustment for adult weight (standing unaided β = 1.06, 95% CI = 0.56-1.56, P < 0.001). Age at walking supported was only significant in the fully adjusted model (β = 0.52, 95% CI = 0.12-0.92, P = 0.012). When analyzed separately by sex, the association between walking supported and aerobic fitness ceased to be significant in males, whereas the association between age at standing unaided ceased to be significant in females; however, the directions of the association and coefficients are comparable so they may just reflect the reduced power from stratifying by sex.
We thereafter introduced birth weight, infant weight gain, and infant motor development in a forward stepwise regression model for each adult physical performance outcome (Table 4).
Higher birth weight (β = 0.08, P ≤ 0.001), greater infant weight gain (β = 0.04, P = 0.003), and earlier motor development (β = −0.03, P = 0.006) were independently associated with muscle strength (Table 4) after adjusting for sex, gestational age, father's social class, adult educational status, and adult height. After further adjusting for adult weight, infant weight gain ceased to remain significantly associated with handgrip. The associations with birth weight were attenuated by almost 50% (β = 0.05, P < 0.001), whereas the association between infant motor development and adult handgrip strength remained virtually unchanged (β = −0.02, P = 0.013).
Lower infant weight gain (β = −0.05, P = 0.004) and earlier motor development (β = −0.05, P = 0.006) were associated with greater muscle endurance after adjusting for sex, gestational age, father's social class, adult educational status, and adult height. After further adjusting for adult weight, infant weight gain ceased to be associated with trunk muscle endurance, whereas higher birth weight (β = 0.07, P < 0.001) and earlier motor development (β = −0.07, P < 0.001) remained statistically significantly associated with greater muscle endurance.
Finally, lower infant weight grain (β = 0.06, P = 0.002) and earlier motor development (β = 0.06, P = 0.002) were independently associated with decreased heart rate post step test, indicative of higher aerobic fitness (Table 3) after adjusting for sex, gestational age, father's social class, educational status, and adult height. After further adjusting for adult weight, infant weight gain ceased to be significantly associated with post step test heart rate, whereas birth weight (β = −0.07, P < 0.001) and earlier infant motor development (0.07, P < 0.001) remained significantly associated with improved aerobic fitness.
Our results from the 1966 Northern Finland Birth Cohort suggest that birth weight, infant weight gain, and infant motor development are independent markers of adult physical performance in terms of muscle strength, muscular endurance, and aerobic fitness. We observed that earlier infant motor development and slower infant weight gain were associated with improved adult physical performance in terms of improved muscle endurance and aerobic fitness, whereas higher birth weight, greater infant growth, and earlier motor development were associated with increased muscle strength. These observations were independent of sex, gestational age, adult height, father's social class category in 1966, and adult educational status. When further adjusted for adult weight, the association with infant growth was attenuated, but birth weight and infant motor development were still significantly associated with muscle strength, muscle endurance, and aerobic fitness.
Our findings show several consistencies with previous studies that have shown positive associations between birth weight and later handgrip strength in adolescents (20) and adults (8,13,15,27,39). Birth weight has also been related to later lean tissue mass (13,18,30,39), so the consistent association between higher birth weight and greater muscle strength and muscle endurance may be explained by increased muscle mass. In our study, this association seemed to be only partially mediated by larger adult body size. However, we cannot preclude that this association is entirely mediated by adult lean body mass because a measure of adult body composition was not available in our study. Two previous studies in adolescents (20) and adults (39), where body composition was estimated using skinfold measurements and bioelectrical impedance, respectively, suggested that fat-free mass did mediate the association seen between birth weight and handgrip strength.
There have been a small number of studies examining associations between birth weight and aerobic fitness later in life (2,11,25), only one of which was in normal-weight infants (2). Our findings are consistent with the observed association between lower birth weight with impaired aerobic fitness in 12 year olds, although this association ceased to be significant when followed up 3 yr later at age 15 yr (2). One recent study also in adolescents found no association between birth weight and aerobic fitness (20). In both these studies, the adolescent's fitness was assessed using shuttle run tests, which could penalize heavier individuals as it is a weight bearing test, so may have limited the associations observed as birth weight is also associated with body size. Although the effect sizes in our study are small, both higher birth weight and earlier infant motor development remained significantly associated with aerobic fitness even after adjusting for adult body size. Further studies with more accurate assessment of cardiorespiratory fitness and detailed body composition would be particularly useful to unravel these associations.
In contrast to the positive associations with birth weight, infant weight gain between birth and 1 yr was adversely associated with muscle endurance and aerobic fitness at age 31 yr after adjusting for adult height, but these associations were attenuated after further adjusting for adult weight, suggesting that the effects of infant weight gain may be mediated by subsequent body weight or composition. However, the association between infant weight gain and muscle endurance remained significant in females only, even after adjusting for adult height and weight, suggesting that infant weight gain may be important for muscle strength in females only. Previous work in the Hertfordshire cohort also observed an association between poorer infant growth and reduced grip strength (27), although this association was not independent of adult height, whereas greater growth in childhood and adolescence has been associated with improved grip strength (15). There is growing evidence that rapid weight gain in early life increases the risk of obesity (19) and metabolic risk in later life (5). Taken together, our findings are consistent with the notion that rapid weight gain during infancy may preferentially be in fat mass rather than in lean tissue (4,5). This may mean that rapid infancy weight gain is associated with greater fat mass in adult life, which in turn may impair muscle endurance and aerobic fitness. Future studies are needed to elucidate whether the currently described consequences of infant weight gain on adult physical performance might contribute to or modify the associations between growth and development in early life and subsequent increased risk of obesity and related metabolic disorders.
To our knowledge, two previous studies have suggested an association between the timing of infant developmental milestones and adult physical performance (14,15). One of these studies suggested a quadratic association with optimal adult performance in terms of chair rising and standing balance, observed in those individuals who walked at the modal age of 12 months (14). The other study noted a linear association between birth weight and handgrip strength (15), although this did not persist after adjusting for birth weight and adult size. We found linear associations between earlier infant motor development and improvements in all three measures of adult physical performance: muscle strength, muscle endurance, and aerobic fitness. These associations were independent of birth weight, infant weight gain, and adult body size, indicating that it is not simply larger birth weight infants who reach these milestones earlier. However, it should be noted that, for the previous studies, the measurement period extended beyond 12 months, which may explain the nonlinear association observed (13). Our study was not able to collect motor development information past the 1-yr assessment, so possibly, it may not have captured those individuals with more delayed motor development.
It has been suggested that some infant motor milestones, such as walking supported, are more influenced by "caregivers'" actions, such as encouraging infants to walk with support, and therefore, may be more influenced by sociocultural differences (38). On the other hand, other milestones such as the age at first standing unaided may reflect more of a biological difference in motor development (38). Interestingly, in this study, lower father's social class was associated with later age at walking supported, but there were no associations between social class and standing without support. The associations between standing without support and physical performance do seem to be slightly stronger than those between walking supported and physical performance, suggesting that caregivers and sociocultural factors may have more influence on age at walking supported.
Several other mechanisms might be involved in explaining these associations. Attainment of infant developmental milestones may have a genetic component because there are differences between ethnic groups that seem to be independent of cultural and social factors (12). However, infant motor development may also be influenced by early life exposures; both low birth weight and preterm delivery are associated with slower infant motor development (7). There is also some evidence that early life factors, in terms of birth weight and head circumference, are associated with differences in stress-related motor activity in childhood (29) and with higher resting pulse in adults, suggesting potential differences in the development of the central nervous system (21). Finally, there is evidence that the attainment of infant motor developmental milestones may be modifiable, for example, breast feeding is associated with earlier infant motor development (26,35). Our current findings suggest that any such modification might have long-lasting effects on adult performance.
A potential limitation of our observational study is the possibility of residual confounding. For example, we performed adjustments for social class according to father's occupation in 1966. A direct measure of adult social class was not available, so self-reported educational status at age 31 yr was used as a proxy measure to adjust for socioeconomic status. These classifications were only divided into four broad categories, and further minor differences might have contributed to the associations observed. Other sources of confounding include other parental exposures that could influence early growth and development, as well as potential long-lasting influences on their offspring, such as maternal smoking, alcohol consumption, obesity, education level, and parenting practices.
There is also the potential for residual confounding in that this study did not include an objective measure of body composition at age 31 yr, so although statistical models have been adjusted for adult body size, this will not fully adjust for differences in body composition, in particular, fat-free mass and fat mass. It should be noted that the influence of body composition may vary across the three performance measures of handgrip strength, trunk endurance, and a step test. For example, height was positively associated with handgrip strength, suggesting that greater lean tissue mass may be important. Whereas the trunk endurance test requires an individual to support their own upper body weight, both height and weight were negatively correlated with trunk endurance, suggesting that trunk endurance may be further reduced for those with excess adiposity. This may explain some of the differences observed between the associations.
We were also not able to adjust for potential differences in lifestyle, such as physical activity levels, within this cohort. Data from large cohorts with well-defined physical activity variables would be particularly valuable for elucidating the potential associations between early life variables and physical activity and physical performance in later life.
The 1966 Northern Finland Birth Cohort captured more than 95% of all births in that region during 1966, and the birth weights were comparable to the current UK reference (3). However, a large proportion of the original cohort were excluded from this analysis owing to incomplete follow-up data. Small but significant differences were observed between individuals included in these analyses and the rest of the cohort in gestational age (40.3 vs 39.9 wk, P ≤ 0.001) and in infant weight gain (change in SD score: −0.048 vs −0.005, P = 0.039). These differences are clinically very small, and there was no difference in birth weight between those included in these analyses and the rest of the cohort; however, possibly, other differences exist that might affect the representative nature of our study population.
In conclusion, this study suggests that higher birth weight, slower infant weight gain, and earlier infant motor development are independently associated with benefits in adult physical performance. The mechanisms that underlie these associations cannot be elucidated in the present study. Similar studies in large twin cohorts may help to distinguish whether such long-term links between early growth and development and adult physical performance might be attributed to genetic, intrauterine, or other environmental factors. Regardless of the exact mechanisms involved, our findings suggest that modifications in infant growth and motor development might have long-lasting influences on later adult physical performance and possibly health.
Author contributions: manuscript (C. R.), interpretation of data (C. R., U. E., and K. O.), data analysis (C. R., U. E., and K. O.), statistics (C. R., S. S., U. E., and K. O.), responsible for data collection and study design (M.-R. J. and T. T.), and critical input on all versions and approval of final version of the manuscript (all authors).
Financial disclosure and conflict of interest: The main funding sources for the Northern Finland Birth Cohort 1966 study have been the Academy of Finland, the Ministry of Social and Health Affairs (Finland), Wellcome Trust (UK), and the Oulu University Hospital, Finland.
The authors would like to thank all the staff and volunteers involved with the Northern Finland Birth Cohort of 1966.
No conflict of interest is declared by any of the authors.
The results of the present study do not constitute endorsement by ACSM.