Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
The population of China is aging. By the year 2036, 20% of the population will be older than 65 years, double that of 10% in 2000.1 As the population ages, it is particularly important to understand predisposing factors to debilitating age-related disorders such as cognitive impairment. Certain early life exposures may contribute to adulthood cognitive impairment either by affecting premorbid structure or function (the “cognitive reserve hypothesis”) or by increasing vulnerability to neurodegenerative processes.2
Greater adult height and leg length, but not sitting height, have been reported to be good biomarkers of childhood environment3 and reliable predictors of adulthood cardiovascular disease and cancer in western developed populations.4–8 Conversely, adult height and to a lesser extent leg length have been found to be negatively associated with cognitive impairment in later life.9–12 Maximal growth in leg length occurs before puberty, whereas maximum sitting height occurs during puberty.13 Skeletal growth occurs in parallel with brain development. However, it is unclear whether measures of final skeletal growth are universally associated with cognitive function in populations of all socioeconomic histories and geo ethnic backgrounds, independent of potential confounding factors.11,14–16 There is limited research on childhood antecedents of cognitive impairment in populations with a history of socioeconomic development that is different from that experienced in western Europe and North America. Our sample of southern Chinese people offers a unique opportunity to study these relations in a population that has experienced very recent and rapid socioeconomic development, and in which the typical relation between better childhood conditions and longer leg length and height has been less evident, possibly because of epigenetic constraints on growth.17,18
Here we sought to clarify the association between several anthropometric markers and cognition in older adulthood in a large study of older men and women in Guangzhou, China. Our goal was to examine the etiologic role of early life conditions that might affect growth, such as levels of nutrition. Specifically, we hypothesized that later adulthood cognition varies directly with better childhood conditions as measured by final adulthood leg length, sitting height, and height.
The Guangzhou Biobank Cohort Study is an ongoing collaboration between the Guangzhou People's Number 12 Hospital and the Universities of Hong Kong and Birmingham (described in detail elsewhere19). Subjects are recruited from “The Guangzhou Health and Happiness Association for the Respectable Elders,” a community social and welfare association aligned with the municipal government. Membership is open to older persons for a monthly fee of 4 Yuan (50 US cents). About 7% of permanent Guangzhou residents aged 50 years and older are members of this association. Of these, 11% enrolled each time for phases 1 and 2 and were included if they were capable of consenting, ambulatory, and not receiving treatment modalities that, if omitted, might result in immediate life-threatening risk, such as chemotherapy or radiotherapy for cancer, or dialysis for renal failure. Those with less immediate risk, such as a history of vascular disease or associated risk factors (including diabetes and hypertension) were not excluded. The prevalence of hypertension and diabetes in phase 1 participants was similar to that seen in men and women of the same age in a nationally representative survey.19 Of those eligible, 90% of the men and 99% of the women participated.
Subjects underwent a detailed medical interview, including questions regarding lifestyle and socioeconomic status, and a physical examination in 2003 to 2004 for phase 1 and 2005 to 2006 for phase 2. Standing height was measured without shoes to the nearest 0.1 cm, sitting height was measured with the subjects sitting on a standard stool, leg length was calculated as the difference between height and sitting height. These measurements may be affected by shrinkage with old age, for example from osteoporosis (which mainly affects the trunk). During the interview, a test was administered from a test battery developed for the Consortium to Establish a Registry for Alzheimer disease,20 to assess cognitive function in older people. The Consortium tests cover general cognition, confrontational naming, semantic fluency, constructional ability, and new learning ability. We used the test of new learning ability (10-word list learning task). Four words in this test were taken from the original English language test21: “arm,” “letter,” “ticket,” and “grass.” “Pole,” “shore,” “cabin,” and “engine” were replaced with “corner,” “stone,” “book,” and “stick” as in the adapted Consortium 10-word list learning task,21 and “butter” and “queen” were replaced by “soy sauce” and “chairman” as these are more culturally appropriate. During the learning phase, the 10-word list was read out to the subject who was then asked to recall immediately the words they remembered; this was repeated 3 times giving an immediate recall or learning score of the total number of correct responses out of 30. After a 5-minute period of distraction, during which the interview was continued, the subject was then asked to recall as many of the 10 words as he or she was able, giving a delayed recall score out of 10.
The adapted Consortium 10-word list learning task has been validated as a culturally and educationally sensitive tool for identifying dementia in population-based research in developing countries.21 In addition, the delayed recall component of the 10-word list learning task has been validated as a sensitive tool for identifying mild cognitive impairment,22 which is subclinical cognitive decline distinct from normal ageing. Mild cognitive impairment does not interfere with daily life, and may affect several cognitive domains, including general cognition and executive functioning.23–25 Amnesic-mild cognitive impairment is a subtype characterized by memory deficits, corresponding to poor delayed recall on the 10-word list learning task. Both mild cognitive impairment and the amnesic-subtype are conditions with a high risk of progressing to Alzheimer disease or dementia,24–26 probably because recall activity is centered in the entorhinal cortex and hippocampus, where the earliest neuropathologic changes in Alzheimer disease occur.22 Annual conversion rates from mild cognitive impairment to dementia vary from about 10% to 15% in clinic-based studies to about 5% to 10% in population-based studies, with possibly higher rates for the amnesic-subtype.25,26 The Guangzhou Medical Ethics Committee of the Chinese Medical Association approved the study, and all participants gave written, informed consent before participation.
The outcomes were amnesic-mild cognitive impairment and the delayed recall score (out of 10) on the adapted Consortium 10-word learning task. Amnesic-mild cognitive impairment was defined as a delayed recall score of 3 or less out of 10, corresponding to 1 standard deviation below the mean, whereas a score of 4 to 10 was defined as normal cognition. The cutoff is usually set at 1 or 1.5 standard deviations (SDs) below the mean of the delayed 10-word recall, which would correspond in our sample, with mean 5.2 and SD 1.7, to scores of 3.5 and 2.7, respectively.
Height and its components were considered in decimeters, giving estimates per 10 cm change in leg length, sitting height, or height. We also assessed whether nonlinear representations of these exposures (powers, tertiles, quartiles, and quintiles of each dimension) were more appropriate for either outcome, by comparing model fit in parsimonious models using the Akaike Information Criterion (AIC).27 Height and its components may diminish with age because of shrinkage in older people and may increase over generations, so groups based on absolute dimensions may be biased by cohort effects and have very different mean ages. Instead, we grouped by relative size for age. Formulas for adjusting height for age usually use linear, square, and possibly cubic terms,28 but have been developed only in white populations. We fitted sex-specific regression equations to height and its components using linear, square, and cubic terms of age, and then assigned groups (tertiles, quartiles, and quintiles) based on z-scores from these fitted curves. We, thus, obtained groups of relative dimension for age, unbiased by cohort effects. Tertiles generally produced the best-fitting models for leg length, although powers of leg length had a slightly better fit for amnesic-mild cognitive impairment in women. Sitting height as a linear quantity fitted best in all models. Linear height fitted fairly well in men, although quartiles and powers were slightly better for amnesic-mild cognitive impairment; whereas power terms fitted best for women, although quartiles had a similar fit for delayed 10-word recall score. So, for simplicity and comprehensibility, we present all dimensions as linear quantities (ie, per 10-cm change), but we also give leg length in tertiles and height in quartiles.
Multivariable linear regression was used to assess the association of height and its components with delayed 10-word recall score from which we reported adjusted mean differences with 95% confidence intervals (CIs). For dimensions as linear quantities, the mean differences correspond to the number of additional words recalled per 10-cm change in leg length, sitting height, or height. Multivariable logistic regression was used to assess the association of these same exposures with amnesic-mild cognitive impairment. To be consistent with the presentation of delayed 10-word recall score, where a higher score indicates better performance, we used amnesic-mild cognitive impairment as the reference category.
Potential confounders considered, categorized as per Table 1, were age, number of children, educational level, current annual personal income, job type, smoking status, and physical activity in both men and women, and age of menarche in women; these have all been associated with adulthood cognitive outcome.29–32 Apart from age and education, these potential confounders were assessed according to a change-in-estimate criterion (at least ≥5% change)33 for the estimates of the effect of leg length, sitting height, or height on the delayed 10-word recall score in sex-specific linear regressions to create parsimonious models. Most of these potential confounders had fairly small effects on the estimates, although estimates in a full model usually differed by more than 5% from a model adjusted only for age and education. To create parsimonious models, we identified a common subset of confounders that produced estimates similar to the full models. Apart from age and education, these additional confounders were personal income and smoking in men and personal income and age of menarche in women. We present 3 models: model 1 adjusted for age; model 2 additionally adjusted for education, personal income, and in men, smoking, and in women, age of menarche; and model 3 additionally adjusted leg length and sitting height for the other component of height. We also calculated mean delayed 10-word recall score adjusted for age, education, income, plus (in women) age of menarche and (in men) smoking, by sextiles of each dimension.
Education is strongly associated with cognition. Men and women in our sample have similar cognitive scores, but for historical and cultural reasons women have much lower levels of education. Results for men and women are presented separately to avoid artifactual results created by adjustment for education.
We also examined whether the effects of sitting height, leg length, and height were consistent across age groups (<62.5 years and ≥62.5 years) from the heterogeneity-of-effect across strata and model fit using the AIC.27 Evidence of effect-measure modification was assessed from running main effect models with and without the interaction term, and examining which model had the lower AIC value (and hence was the better fitting model). Finally as a sensitivity test, analyses were repeated using a cut off of 1.5 SD below the mean for defining amnesic-mild cognitive impairment. Data analysis was done using STATA v8.2 (STATA Corporation, College Station, Texas).
Of the 20,411 participants there were 14,535 women and 5876 men. Of these, 13,708 women (94%) and 5502 men (94%) had complete data for all variables. Men were aged from 50 to 93 years (mean = 64.7 [SD = 6.3]) years. Women were aged from 50 to 94 years (mean of 61.8 [6.6] years). Almost all the subjects (98%) were Han Chinese. Mean leg lengths were 76.0 (4.0) cm and 70.5 (3.7) cm; mean sitting height 88.4 (3.3) cm and 82.9 (3.3) cm; and mean height 164.3 (5.7) cm and 153.4 5.4 cm in men and women, respectively. Fewer than half of men (41%) and the majority of women (96%) had never smoked. Around 92% of all subjects reported regular physical exercise as defined by the International Physical Activity Questionnaire criteria.26 The men had more education, with 70% of men and 49% of women having attended junior middle school or above.
The prevalence of memory impairment was 15% in men and women. Table 1 shows biologic, socioeconomic, and lifestyle characteristics by amnesic-mild cognitive impairment status. In both men and women, impairment was associated with older age, more children, shorter sitting height and shorter overall height, lower educational attainment, manual occupation, and lower personal annual income. Amnesic-mild cognitive impairment was additionally associated with later age of menarche and ever-smoking in women. The relations of leg length, sitting height, and height with adjusted delayed 10-word recall score were not completely linear (Fig.1).
Longer legs were associated with higher educational status, nonmanual occupation and ever-smoking in men; with older age of menarche in women; and with greater income in both sexes (eTable 1, available with the online version of this article). Greater sitting height was associated with normal cognition, younger age, higher educational attainment, nonmanual occupation, greater income, and fewer children; and with younger age of menarche and never-smoking in women. Greater total height was associated with normal cognition, younger age, higher educational attainment, nonmanual occupation, greater income, and fewer children in both men and women, with ever-smoking in men, and with never smoking in women.
Longer legs, greater sitting height, and greater height were all associated with a larger number of words recalled for both men and women in almost all models (Table 2), even after mutual adjustment of leg length and sitting height—although this was only obvious in the better fitting model for leg length using tertiles. There was a stronger association for sitting height than for leg length, where an extra 10 cm of sitting height was associated with an additional 0.21 (CI = 0.07–0.34) words recalled in men and 0.18 (CI = 0.08–0.27) words in women; the association of leg length with delayed word score was evident only for the tertiles with the longest legs.
Greater sitting height and height were associated with increased odds of normal cognition in men and women in all models, such that after all adjustment (model 3) an extra 10 cm in sitting height was associated with a 31% (CI = 3%–68%) increased odds of normal cognition in men and 56% (increase 32%–83%) in women. Results were essentially similar when using amnesic-mild cognitive impairment defined as a score of 2 or less on the delayed 10-word recall test.
There was some evidence that the relations of leg length, sitting height, and height to delayed 10-word recall varied with age group, specifically for height in both sexes and for leg length in men (eTable 2). Greater height was associated with higher delayed 10-word recall score only in older men and younger women (Table 3).
We examined both leg length and sitting height in relation to cognitive impairment in later adulthood. Consistent with other studies,11,12 height was associated with better cognitive outcomes, with the greater and most consistent contribution from sitting height; height was positively associated with better cognition only in older men and younger women. A relation between leg length and cognition is similar to findings of 2 other studies on smaller samples of first generation African-Caribbean immigrant elders in south London9 and of older Korean women.10 A positive relation between adult cognition and growth in height from 2 to 4 years of age and after 15 years of age has been observed previously.11 Leg growth is mainly prepubertal, whereas growth in sitting height accompanies the pubertal growth spurt in adolescence.13 A stronger effect of sitting height than leg length on cognition suggests that prepubertal growth may be less important than pubertal development.
The association between sitting height and the maintenance of cognitive functioning in late adulthood may have several explanations. Sitting height may be a marker of exposures that affect both growth and neurologic structure. Growth in sitting height seems to be impaired by psychosocial stress.3 Although the exact physiologic pathways remain to be mapped out, stress during growth and development may also detrimentally affect the hippocampus,34 with corresponding consequences for later life cognitive function.34 It is unclear to what extent poor later life cognition in this study is the result of early life psychologic distress, as such information is difficult to collect retrospectively.
Alternatively, sitting height may be a marker of exposures, such as nutrition, that improve both linear growth and neurologic structure. It is well known that malnutrition impairs both growth and cognitive development,35 although this has most commonly been observed in infants and young children rather than adolescents. This association is possibly mediated by the somatotropic axis. Children and adults with both congenital and acquired growth hormone (GH) deficiency have worse adult cognitive outcomes, specifically in memory and attention.36 These impairments improve with GH therapy (which simultaneously, in children, improves somatic growth). In children who are small for gestational age, GH-induced catch-up linear growth during puberty is also associated with cognitive catch-up.37 GH and insulin-like growth factor I (IGF-I) receptors are present in the highest concentration in areas of the brain involved with memory, such as the hippocampus, amygdala, and parahippocampus.38 GH may affect cognition directly (via GH binding sites)38 or indirectly (by increasing IGF levels in the brain, interaction with neurotransmitters, and thyroxin),39 with the GH/IGF-I axis as the key underlying biologic pathway. Growth at puberty is mediated by an interplay of the somatotropic (GH-IGF-I) and gonadotropic (GnRH-LH/FSH-sex steroid) axes. Whether these different physiologic pathways have different effects on brain morphology in adolescence is not known, although the hippocampus is well stocked with estrogen receptors.40 Few studies have examined the effects of sex steroids in adolescence on long-term neurologic structure or cognitive functioning, although adolescence is an important period for neurochemical and neuroanatomic remodeling. Although sex steroids have sexually dimorphic effects on sitting height,41 similar effects of sitting height on cognition in both sexes suggest that the underlying pathway is not testosterone dependent. The areas of the brain implicated in early Alzheimer disease, such as the hippocampus, are also those to take longest to mature during childhood and into adolescence.42 Sitting height may be a better marker of adulthood cognition (specifically memory) than leg length because skeletal growth in the trunk and not the leg coincides better temporally with neurologic development of the areas of the brain involved with memory.
Alternatively, we may be seeing a stronger association between sitting height and cognition than between leg length and cognition because leg length may not be a reliable biomarker of childhood conditions in this and similar populations at the early stages of the epidemiologic transition; epigenetic constraints on growth might have prevented the increase in leg length and height usually seen in response to better childhood conditions.17,18,43 This attenuation of effect might explain the nonlinear relation between leg length and cognition seen here in women and men. Thus, we cannot rule out the possibility that prepubertal environmental exposures are associated with much better cognition, of which we are only seeing an attenuated effect. Nevertheless, we observe an association between growth at puberty and cognition, which may be merely a reflection of generally better conditions throughout early life—conditions that in our population have resulted only in pubertal growth.
Greater sitting height has been found to be associated with metabolic risk in western populations.6–8 In this study, we have previously reported an excess risk of diabetes of about 27% per 10 cm increase in sitting height44 and the association with better cognition presented here. This may imply a trade-off at puberty between cognitive outcome and metabolic risk in later life.
There are a number of potential limitations of the present study. The infrastructure to facilitate fully representative studies in most developing countries, such as China, is not readily available. However, waiting until all the appropriate infrastructure is in place before embarking on studies from the developing world would make it impossible to study a large proportion of the global population during a period of immense upheaval and transition. Although we cannot guarantee to have selected a representative sample, this would invalidate our results only if we have systematically missed men and women with a different relation between the exposure and outcome—for example, if we are missing short men and women with good cognition and tall men and women with poor cognition. It is difficult to see why this should have happened. Our outcome measure is simple and yet has been proven to be one of the best markers of early cognitive impairment and particularly useful in epidemiologic studies.21 Measurement error could affect the anthropometric measures—particularly leg length that is calculated as the difference between 2 measures each with their own variances. Sitting height may be subject to shrinkage with age and ill-health, so we cannot exclude the possibility that there is a common pathway responsible for physical and cognitive decline, regardless of sitting height at the termination of growth and of childhood exposures. Random error would have weakened our results but should have done so equally for leg length and sitting height. We were not able to assess a possible confounding role of depression, which is associated with poor performance on tests of cognitive function. This would have affected our results only if depression were also associated with leg length and sitting height, which seems unlikely. There is no reason to suspect recall bias in age of menarche, as women were unaware of the objectives of the study on interview.
Survival bias is a possibility. However, if this were the case we would have expected differences in association between the older and younger age groups (≤62.5 years or 62.5+ years), of which we found little evidence, except for leg length in women and height in men. Finally, we did not explore the roles of genetics or early sexual maturation. There is no information on the height of the participants’ parents. In our setting it is quite likely that this information was never known, so we cannot exclude the possibility that there is a common genetic pathway underlying both shortness and poor cognition. However, given the limited contribution of genetics to height, this is unlikely. As regards sexual maturation, including age of menarche for women made little difference. We collected no similar information for men, so we cannot elucidate the role of sexual maturation in cognition for men.
Better understanding and, hence, potential modulation of childhood precursors and conditions that result in growth at certain phases (such as longer leg length and sitting height) could reduce or delay the onset of cognitive impairment in later life. Particular environmental conditions that promote better growth in later childhood and adolescence may have long lasting benefits for cognitive development and maintenance, with corresponding implications for healthy aging. These findings also support the hypothesis of a trade-off between growth at certain phases, such as puberty, resulting in better cognition but greater risk of certain adulthood diseases such as diabetes.
We thank R Peto and ZM Chen of the Clinical Trial Service Unit, The University of Oxford for their support. The Guangzhou Cohort Study investigators include: Guangzhou No. 12 Hospital: XQ Lao, WS Zhang, M Cao, T Zhu, B Liu, CQ Jiang (Co-PI); The University of Hong Kong: CM Schooling, SM McGhee, GM Leung, RF Fielding, TH Lam (Co-PI); The University of Birmingham: P Adab, GN Thomas, Y Peng, KK Cheng (Co-PI).
1. Woo J, Kwok T, Sze FK, et al. Ageing in China: health and social consequences and responses. Int J Epidemiol
2. Snowdon DA, Kemper SJ, Mortimer JA, et al. Linguistic ability in early life and cognitive function and Alzheimer's disease in late life. Findings from the Nun Study. JAMA
3. Wadsworth ME, Hardy RJ, Paul AA, et al. Leg and trunk length at 43 years in relation to childhood health, diet and family circumstances; evidence from the 1946 national birth cohort. Int J Epidemiol
4. Gunnell DJ, Davey SG, Frankel S, et al. Childhood leg length and adult mortality: follow up of the Carnegie (Boyd Orr) Survey of Diet and Health in Pre-war Britain. J Epidemiol Community Health
5. Gunnell D, Okasha M, Smith GD, et al. Height, leg length, and cancer risk: a systematic review. Epidemiol Rev
6. Smith GD, Greenwood R, Gunnell D, et al. Leg length, insulin resistance, and coronary heart disease risk: the Caerphilly Study. J Epidemiol Community Health
7. Lawlor DA, Taylor M, Davey SG, et al. Associations of components of adult height with coronary heart disease in postmenopausal women: the British women's heart and health study. Heart
8. Asao K, Kao WH, Baptiste-Roberts K, et al. Short stature and the risk of adiposity, insulin resistance, and type 2 diabetes in middle age: the Third National Health and Nutrition Examination Survey (NHANES III), 1988-1994. Diabetes Care
9. Mak Z, Kim JM, Stewart R. Leg length, cognitive impairment and cognitive decline in an African-Caribbean population. Int J Geriatr Psychiatry
10. Kim JM, Stewart R, Shin IS, et al. Limb length and dementia in an older Korean population. J Neurol Neurosurg Psychiatry
11. Richards M, Hardy R, Kuh D, et al. Birthweight, postnatal growth and cognitive function in a national UK birth cohort. Int J Epidemiol
12. Beeri MS, Davidson M, Silverman JM, et al. Relationship between body height and dementia. Am J Geriatr Psychiatry
13. Duval-Beaupere G. Growth of the trunk and lower limbs after the menarche. Rev Chir Orthop Reparatrice Appar Mot
. 1976;62:501–509[in French].
14. Tanner JM. Galtonian eugenics and the study of growth: the relation of body size, intelligence test score, and social circumstances in children and adults. Eugen Rev
15. Cheung YB, Yip PS, Karlberg JP. Fetal growth, early postnatal growth and motor development in Pakistani infants. Int J Epidemiol
16. Karp R, Martin R, Sewell T, et al. Growth and academic achievement in inner-city kindergarten children. The relationship of height, weight, cognitive ability, and neurodevelopmental level. Clin Pediatr (Phila)
17. Schooling CM, Jiang C, Heys M, et al. Is leg length a biomarker of childhood conditions in older Chinese women? The Guangzhou Biobank Cohort Study. JECH
18. Schooling CM, Jiang CQ, Heys M, et al. Are height and leg length universal markers of childhood conditions? The Guangzhou Biobank Cohort Study. JECH
19. Jiang C, Thomas GN, Lam TH, et al. Cohort profile: The Guangzhou Biobank Cohort Study, a Guangzhou-Hong Kong-Birmingham collaboration. Int J Epidemiol
20. Welsh KA, Hoffman JM, Earl NL, et al. Neural correlates of dementia: regional brain metabolism (FDG-PET) and the CERAD neuropsychological battery. Arch Clin Neuropsychol
21. Prince M, Acosta D, Chiu H, et al. Dementia diagnosis in developing countries: a cross-cultural validation study. Lancet
22. Shankle WR, Romney AK, Hara J, et al. Methods to improve the detection of mild cognitive impairment. Proc Natl Acad Sci USA
23. Ritchie K, Touchon J. Mild cognitive impairment: conceptual basis and current nosological status. Lancet
24. Petersen RC. Mild cognitive impairment: current research and clinical implications. Semin Neurol
25. Gauthier S, Reisberg B, Zaudig M, et al. Mild cognitive impairment. Lancet
26. Mariani E, Monastero R, Mecocci P. Mild cognitive impairment: a systematic review. J Alzheimers Dis
27. Harrell FE J. Regression Modeling Strategies: With Applications to Linear Models, Logistic Regression, and Survival Analysis
. New York, NY: Springer Verlag; 2001.
28. Cline MG, Meredith KE, Boyer JT, et al. Decline of height with age in adults in a general population sample: estimating maximum height and distinguishing birth cohort effects from actual loss of stature with aging. Hum Biol
29. Colucci M, Cammarata S, Assini A, et al. The number of pregnancies is a risk factor for Alzheimer's disease. Eur J Neurol
30. Weuve J, Kang JH, Manson JE, et al. Physical activity, including walking, and cognitive function in older women. JAMA
31. Swan GE, Lessov-Schlaggar CN. The effects of tobacco smoke and nicotine on cognition and the brain. Neuropsychol Rev
32. Geerlings MI, Ruitenberg A, Witteman JC, et al. Reproductive period and risk of dementia in postmenopausal women. JAMA
33. Rothman KJ, Greenland S. Modern Epidemiology.
2nd ed. Philadelphia: Lippincott-Raven; 1998.
34. Fenoglio KA, Brunson KL, Baram TZ. Hippocampal neuroplasticity induced by early-life stress: functional and molecular aspects. Front Neuroendocrinol
35. Jukes M. The long-term impact of preschool health and nutrition on education. Food Nutr Bull
36. van Dam PS, de Winter CF, de VR, et al. Childhood-onset growth hormone deficiency, cognitive function and brain N-acetylaspartate. Psychoneuroendocrinology
37. van Pareren YK, Duivenvoorden HJ, Slijper FS, et al. Intelligence and psychosocial functioning during long-term growth hormone therapy in children born small for gestational age. J Clin Endocrinol Metab
38. Lai Z, Roos P, Zhai O, et al. Age-related reduction of human growth hormone-binding sites in the human brain. Brain Res
39. Lobie PE, Zhu T, Graichen R, et al. Growth hormone, insulin-like growth factor I and the CNS: localization, function and mechanism of action. Growth Horm IGF Res
. 2000;10(suppl B):S51–S56.
40. Liu F, Day M, Muniz LC, et al. Activation of estrogen receptor-beta regulates hippocampal synaptic plasticity and improves memory. Nat Neurosci
41. Tanner JM, Whitehouse RH, Hughes PC, et al. Relative importance of growth hormone and sex steroids for the growth at puberty of trunk length, limb length, and muscle width in growth hormone-deficient children. J Pediatr
42. Moceri VM, Kukull WA, Emanual I, et al. Using census data and birth certificates to reconstruct the early-life socioeconomic environment and the relation to the development of Alzheimer's disease. Epidemiology
43. Hui LL, Schooling CM, Cowling BJ, et al. Are universal standards for optimal infant growth appropriate? Evidence from a Hong Kong Chinese birth cohort. Arch Dis Child
44. Schooling CM, Jiang C, Lam TH, et al. Height, its components and cardiovascular risk in older Chinese: a cross-sectional analysis of The Guangzhou Biobank Cohort Study. Am J Public Health