This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

Anthropometric Changes Over 5 Years in Elderly Canadians by Age, Gender, and Cognitive Status

^a Centre de recherche, Institut universitaire de gériatrie de Montréal, Quebec, Canada
^b Department of Specialised Medicine, Institut universitaire de gériatrie de Montréal, Quebec, Canada
^c Département de nutrition, Université de Montréal, Quebec, Canada

Bryna Shatenstein, PDt, Centre de recherche, Institut universitaire de gériatrie de Montréal, 4565 Queen Mary, Montreal, Quebec, Canada H3W 1W5 E-mail: bryna.shatenstein{at}umontreal.ca.

Decision Editor: John E. Morley, MB, BCh

	Abstract

Top Abstract Methods and Materials Results Discussion References

 
Background. Numerous changes in body composition occur with aging. This study reports on secondary analyses of data from a subsample of institutionalized and free-living elderly Canadians taking part in both phases of the Canadian Study of Health and Aging (CSHA-1 and CHSA-2; n = 10,263) to document and examine correlates of the evolution of anthropometric characteristics over a 5-year period.

Methods. In CSHA-1, community-dwelling (n = 1464) and institutionalized (n = 963) participants' height and weight were measured in clinics. Surviving participants were remeasured in CSHA-2; valid data were available for 487 community-dwelling respondents (66.9% of those seen in clinics in CSHA-2) and 140 institutionalized participants (46.9% of those reassessed). Body mass index (BMI = weight [kg]/height [m²]) was calculated. Paired t tests were used to test changes over the interval, and repeated-measures multivariate analysis of variance was used to examine the extent of differences within and across categories.

Results. The average weight loss between study phases in community-dwelling and institutionalized participants was approximately 2 kg (p < .001). In institutions, this was statistically significant in most stratification categories, as was the case in community-dwelling participants (by gender, age, dementia screening score, and cognitive diagnosis). Among those who were cognitively intact, the greatest weight losses occurred in participants under 90 years old and in those aged 70 to 79 years with a diagnosis of dementia (p < .01). Stature decreased more in institutionalized (2 cm) than in community-dwelling participants (1.4 cm). In institutions, this was significant among the oldest men (p < .005), while in the community there were no differences in the extent of height lost in all stratification categories. The average BMI was largely stable.

Conclusions. Body weight and stature declined with aging among elderly Canadian CSHA participants, particularly in the very old and those with dementia. Such longitudinal anthropometric data are needed along with information on dietary intakes, and medical, cognitive, and functional measures to plan interventions geared to maximizing nutritional and overall health in the elderly population, whatever their cognitive status.

NUMEROUS changes in body composition occur after the seventh decade of human life. Decreased bone mass, changes in the size of body organs, decreases in skeletal muscle and body water, and changes in body fat may be singly or jointly responsible for losses in lean and fat body mass. Age-related decline in stature is a common observation, but its physiological or clinical importance to health outcomes is unknown ⁽¹⁾. Although there are few reports on population-based longitudinal studies of anthropometric measures in elderly men and women ⁽²⁾, it has been observed that weight changes after age 50 are generally associated with deterioration in nutritional and overall health ^(3)(4)(5) and mobility ⁽⁵⁾.

Data on the evolution of anthropometric parameters in the elderly population are equivocal. Among healthy elderly subjects recruited for the New Mexico Aging Process Study (APS), which began in 1979, and the Toulouse Aging Process Study, which started in 1991, investigators have reported an inverse cross-sectional relationship between age and health status, functional characteristics, and cognition. Surprisingly, cross-sectional data showed that body weight and body mass index (BMI) increased in these cohorts with decreasing health ⁽⁶⁾⁽⁷⁾. In the New Mexico APS participants, height declined with age, both cross-sectionally and over time, but weight did not change over time in men. Among women, weight decreased at the rate of 0.14 kg/year ⁽⁸⁾. In the EC/Euronut Survey in Europe on Nutrition and the Elderly, a Concerted Action (SENECA) study of healthy elderly men and women ⁽²⁾, average body weight showed only small mean differences over time, but there were substantial individual changes in one quarter of participants. Overall, the APS found that anthropometric measures remained relatively constant over a 9-year period, showing no alteration with age. The authors concluded that nutritional status was not affected by the aging process alone ⁽⁶⁾. Still, because SENECA participants were lost at follow-up due to ill health, study results pertain to relatively healthy elderly people.

Simplistic analyses of weight and morbidity and mortality in an elderly population can be misleading ⁽⁹⁾, as an aged person's current body weight may not be reflective of his or her long-term usual weight, and weight loss (or gain) may be either voluntary or involuntary. In an analysis of the 1989 National Health Interview Survey, risk of unintentional weight loss was linked to greater age, poorer health status, and smoking. In men, widowhood emerged as a significant risk factor. Education and a low usual BMI were found to be protective for both men and women, and mortality was linked to factors associated with unintentional weight loss ⁽¹⁰⁾. Losonczy and colleagues ⁽⁴⁾ found that weight lost between middle and older age increased mortality risk after the age of 70. They proposed that low weight was indicative of poor health and decline in health led to an increased mortality risk. Yaari and Goldbourt ⁽¹¹⁾, looking at weight changes and mortality among Israeli men, found that all-cause mortality linked to weight loss was greatest among the initially leanest individuals. Unlike Meltzer and Everhart ⁽¹⁰⁾, Yaari and Goldbourt's results ⁽¹¹⁾ appear to indicate a possible increase in mortality risk related to attempts to alter weight. Willett ⁽⁹⁾ has suggested that diminished lean body mass (contributing to weight loss) leads to poor health, whereas decreased fat mass (also leading to weight loss) would, in contrast, be associated with improved health. Thus, numerous authors urge investigators to consider weight history and weight change when undertaking analyses of body weight and health risks in those older than 70 years of age ^(3)(4)(9).

To our knowledge, longitudinal data on height and weight have not previously been published in a cohort of elderly Canadians. This study documents and examines correlates of anthropometric characteristics and their evolution over a 5-year period in a subsample of elderly Canadians who had taken part in both phases the Canadian Study of Health and Aging (CSHA) ⁽¹²⁾. Such information can aid in identifying high-risk target groups who could benefit from nutritional intervention. The data were examined separately for community-dwelling and institutionalized participants and explored age, gender, and two measures of cognitive status as correlates of anthropometric changes. Given the literature on body weight and dementia, we expected that decreases in weight and in BMI from the two phases of the CSHA, CSHA-1 to CSHA-2, would be greater in elderly individuals with dementia compared with those indivuals who were cognitively intact.

	Methods and Materials

Top Abstract Methods and Materials Results Discussion References

 
CSHA
The CSHA was conducted in two phases, with a nationwide prevalence study of dementia in 1991–1992 (CSHA-1), and a longitudinal follow-up phase in 1996 (CSHA-2) designed to assess the incidence of dementia in the aging Canadian population ⁽¹²⁾. People aged older than 65 years were randomly selected using recruitment procedures that differed for those living in the community (n = 9008) and in institutions (n = 1255). Briefly, 36 sample areas defined by feasibility criteria (e.g., accessibility to study centers) were designated across Canada. This sample frame covered around 60% of Canadians aged older than 65 years living in cities and their surrounding rural areas. The community-dwelling sample was recruited from provincial health insurance databases, except in Ontario, where electoral lists were used. The institutionalized sample was made up of a stratified representative sample of nursing homes, chronic care facilities, and collective dwellings from which residents were randomly selected and then recruited.

Cognitive Status
Community-dwelling participants were screened by interviewers for cognitive impairment using the modified Mini-Mental State Examination (3MS) ⁽¹³⁾, a clinical tool for grading patients'cognitive status modified from the Mini-Mental State Examination (MMSE) ⁽¹⁴⁾. It has a 16-item grid addressing memory, mental processes, temporal orientation, spatial orientation, word associations, semantic associations, and ability to follow instructions. The maximum score is 100. Patients are scored as normal (78), having mild cognitive impairment (66–77), declining cognitive abilities (50–65), or severe cognitive difficulties (<50). The untestable, those patients testing positive for dementia, and a random sample of control subjects (having a reference person willing to complete study instruments designed for proxy respondents) underwent a clinical examination. All institutionalized participants who met inclusion criteria (spoke English or French, and lived in the study area) took part in the clinical examination. The 3MS was readministered by a nurse during this clinical assessment. The final cognitive diagnosis was reached by consensus of the clinician, another physician, and a neuropsychologist, aided by the nurse who had administered the clinical questionnaire, using criteria from the Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition ⁽¹⁵⁾. Participants who were not consistent with dementia criteria, but who manifested cognitive impairment, were termed "cognitively impaired, not demented" (CIND), by exclusion.

Participants and Measures
Following the clinical exam recruitment procedure described previously, height and weight were measured in the CSHA-1 in 1464 (16.3%) community-based participants and 963 (76.7%) institutionalized participants. In CSHA-2, 728 community-dwelling subjects and 298 institutionalized subjects were reassessed in clinics, and height and weight were remeasured in 521 (71.6%) and 160 (53.7%) participants, respectively. These measurements were carried out during the clinical examination by nurses who were trained to follow procedures described in the nurse's study manual. It should be noted that different nurses might have obtained these measurements, as some study personnel changed in the 5-year interval from CSHA-1 to CSHA-2. Participants were lost to follow-up in phase 2, mainly due to death (38.5%. community; 65.5%, institutions) or refusal/noncontact (11.7%, community; 3.5%, institutions; Fig. 1).

View larger version (23K): [in this window] [in a new window]   Figure 1. Canadian Study of Health and Aging (CSHA) anthropometric measures, phases 1 and 2.

The distribution of independent variables was examined in the subsamples (institution and community). A systematic review by clinicians and researchers verified height and weight data for plausibility. As a result, 54 participants with unacceptable values (increase in height from CSHA-1 to CSHA-2, substantial and implausible divergence in height and/or weight from CSHA-1 to CSHA-2, or unlikely weight for gender or height) were excluded from this set of analyses. BMI (weight [kg]/height [m²]) was calculated. The significance of changes in the dependent variables of height, weight, and BMI from CSHA-1 to CSHA-2 was tested in each sample using paired t tests, and the extent of differences within and interactions across categories (independent variables of gender, age, 3MS score, and diagnosis) was assessed by repeated-measures multivariate analysis of variance (MANOVA), testing separately for 3MS score and cognitive diagnosis. Analyses were conducted using the SPSS V8 software package (SPSS, Inc., Chicago, IL).

	Results

Top Abstract Methods and Materials Results Discussion References

 
Participants in CSHA-2 retained the recruitment status (institution or community) established in CSHA-1, whatever their residence situation at the time of this second phase of the CSHA. Body weight was available for 249 institutionalized and 603 community-dwelling subjects. Of these, 237 (95.2%) in institutions, and 584 (96.8%) community-dwelling participants had plausible data at both phases. Height was measured in fewer subjects, limiting the calculation of BMI to those with both measurements. Approximately 18% of institutionalized subjects and 44% of community-dwelling participants were diagnosed as cognitively intact; 50% and 27%, respectively, were diagnosed with dementia; and the remainder were CIND (data not presented). Means and SDs, along with the 10th, 50th, and 90th percentiles, stratified by age, sex, 3MS score, and cognitive diagnosis, were used to present changes in weight (Table 1 ) and height (Table 2 ) from CSHA-1 to CSHA-2 in community-dwelling participants. Age, screening, and diagnosis variables reflected values in CSHA-2.

Community-dwelling participants measured in both phases of the CSHA (Table 2 ) had an average height loss of 1.4 cm (p < .001). Although statistically significant losses in height were recorded in all stratification categories, there were no differences in the extent of height loss within groups. There was also a decrease in height of 2 cm (p < .001) in the 148 institutionalized participants and statistically significant height losses in most levels of the stratification categories. Repeated-measures MANOVA showed the height loss in institution-dwelling participants to be greatest in men aged 90 years or older (p < .005; data not presented).

The calculation of changes in BMI was limited to those individuals with both height and weight data in CSHA-1 and CSHA-2. Thus, BMI was determined for 140 institutionalized participants and 487 of those in the community. Overall average BMI was 24. Despite BMI changes in some stratification categories, among institutionalized participants, the average BMI did not change from phase 1 to phase 2, while in community-dwelling participants, the small mean decrease in BMI from CSHA-1 to CSHA-2 was without clinical importance (data not shown).

	Discussion

Top Abstract Methods and Materials Results Discussion References

 
The present series of analyses in a cohort of elderly Canadians showed a decline in body weight and stature with aging (more pronounced among the very old and persons diagnosed with dementia), partially confirming the findings of other longitudinal studies in elderly cohorts ⁽²⁾⁽⁸⁾. Our findings thus support the age-related decline in stature reported in the longitudinal phase of the SENECA study of healthy elderly men and women ⁽²⁾. However, they differ from most of the SENECA study sites' ponderal data, as well as the APS observation of relatively stable body weights over time ^(2)(6)(7). It must be remembered that the APS ⁽⁶⁾⁽⁷⁾ and the SENECA ⁽²⁾ cohorts were assembled from healthy community-dwelling volunteers, while the Canadian study design was built around the objective of studying cognitive impairment in community- and institution-dwelling elderly persons. In addition, the CSHA participants were older, on average, than members of the former cohorts. These basic differences render comparisons difficult.

Changes in BMI over time can result from alterations in stature or body weight, or both, and the latter is largely dependent on a number of potential changes in body composition ⁽¹⁾. Statistically significant decreases in BMI, observed in some subgroups of CSHA community-dwelling subjects (data not shown), were too small to have clinical significance. Indeed, they may have resulted from measurement or statistical artifact. Still, an examination of temporal changes in this parameter may provide valuable information about the evolution of nutritional status in elderly persons, in spite of controversy over the usefulness of BMI in elderly individuals, because of the loss of stature and other age-related changes in body composition ⁽²⁾. Loss of height could reflect the consequences of specific conditions or illnesses (e.g., hip or vertebral fracture resulting from osteoporosis) on skeletal integrity, or may result from the elderly person's inability to fully extend hip or knee joints. Other conditions (e.g., sarcopenia resulting from diminished mobility) may be responsible for the weight differences observed in CSHA-2 compared with CSHA-1.

Concerns relating to the data must be addressed, particularly because this study reports on secondary data analyses. Although initial overall sampling was representative of the Canadian elderly population, only those CSHA participants undergoing clinical examination were weighed and measured. In addition, a substantial proportion was lost to follow-up (see Fig. 1), mainly due to death. Still, there is no reason to suspect bias in the selection of those for whom data were available. The present analyses did not consider CSHA-1 decedents. Rather, the data presented here reflect anthropometric values in the surviving members of the cohort and, as such, can be said to represent survivors of the 5-year follow-up period. Since subjects in the oldest age categories were oversampled in the community, the community-dwelling cohort may not be strictly representative of the home-dwelling elderly population. Health status at the time of these analyses was not considered. Therefore, we could not distinguish between age-related decreases in anthropometric parameters and those caused by illness (e.g., cancer-related weight loss or osteoporosis leading to diminished stature), which may be masked since anthropometric data were gathered at only two time points. Finally, it must be cautioned that some statistically significant differences may be attributed to chance, which increases when numerous statistical tests are performed. Despite these concerns, the present analyses provide much-needed and useful data on anthropometric values in elderly Canadians, adding cognitive status to the more commonly reported correlates of age and gender.

To increase our understanding of changes that occur with successful, usual, or pathological aging, elderly populations should be studied longitudinally, and data collected on key anthropometric measurements along with relevant health and functional variables. The CSHA was designed to assess the prevalence and incidence of dementia and related issues in elderly Canadians. Although there is substantial controversy on this issue, it has been suggested that the weight loss typically associated with cognitive disorders, such as Alzheimer's disease, may not be an automatic feature of the disease. Persons afflicted with this type of dementia may therefore benefit from early dietary interventions designed to improve or maintain their nutritional status and slow the onset of malnutrition, which leads to weight loss and attendant complications, including increased mortality ^(18)(19). The present compilation of data on longitudinal changes in basic anthropometric characteristics in a cohort of elderly Canadians suggests the need for expanded nutrition research. This should include information on dietary and nutritional intakes, well-being and environmental predictors, as well as medical, cognitive, and functional status, in order to plan interventions geared to preserving nutritional and overall health in the elderly population, whatever their cognitive status.

	Acknowledgments

 
The data in this article were collected as part of the Canadian Study of Health and Aging. The core study was funded by the Seniors' Independence Research Program through the National Health Research and Development Program (NHRDP) of Health Canada (Project 6606-3954-MC[S]). Additional funding was provided by Pfizer Canada, Inc, through the Medical Research Council/Pharmaceutical Manufacturers Association of Canada Health Activity Program, NHRDP (Project 6603-1417-302[R]), Bayer, Inc, and the British Columbia Health Research Foundation (Projects 38[93-2] and 34[96-1]). The study was coordinated through the University of Ottawa and the Division of Aging and Seniors, Health Canada. Complete data on institutionalized participants may be obtained from the authors upon request.

Received July 1, 1999

Accepted June 1, 2000

	References

Top Abstract Methods and Materials Results Discussion References

 

1. Kuczmarski RJ, 1989. Need for body composition information in elderly subjects. Am J Clin Nutr. 50:1150-1157.
2. de Groot CPGM, Enzi G, Perdigao AL, Deurenberg P, 1996. Longitudinal changes in anthropometric characteristics of elderly Europeans. Eur J Clin Nutr. 50: (suppl 2) S9-S15.
3. Wannamethee G, Shaper AG, 1990. Weight change in middle-aged British men: implications for health. Eur J Clin Nutr. 44:133-142. [Medline]
4. Losonczy KG, Harris TB, Cornoni-Huntley J, et al. 1995. Does weight loss from middle age to old age explain the inverse weight mortality relation in old age?. Am J Epidemiol. 141:312-321. [Abstract/Free Full Text]
5. Harris TB, Savage PJ, Tell GS, Haan M, Kumaniyka S, Lynch JC, 1997. Carrying the burden of cardiovascular risk in old age: associations of weight and weight change with prevalent cardiovascular disease, risk factors, and health status in the Cardiovascular Health Study. Am J Clin Nutr. 66:837-844. [Abstract/Free Full Text]
6. Vellas BJ, Albarede J-L, Garry PJ, 1992. Diseases and aging: patterns of morbidity with age; relationship between aging and age-associated diseases. Am J Clin Nutr. 55:1225S-1230S. [Abstract/Free Full Text]
7. Nicolas AS, Nourhashemi F, Faisant C, et al. 1998. Nutrition et vieillissement réussi: une étude de 520 personnes issues de Toulouse et d'Albuquerque. Age Nutr. 9:82-91.
8. Garry PJ, Hunt WC, Koehler KM, VanderJagt DJ, Vellas BJ, 1992. Longitudinal study of dietary intakes and plasma lipids in healthy elderly men and women. Am J Clin Nutr. 55:682-688. [Abstract/Free Full Text]
9. Willett WC, 1997. Weight loss in the elderly: cause or effect of poor health?. Am J Clin Nutr. 66:737-738. [Free Full Text]
10. Meltzer AA, Everhart JE, 1995. Unintentional weight loss in the United States. Am J Epidemiol. 142:1039-1046. [Abstract/Free Full Text]
11. Yaari S, Goldbourt U, 1998. Voluntary and involuntary weight loss: associations with long term mortality in 9,228 middle-aged and elderly men. Am J Epidemiol. 148:546-555. [Abstract/Free Full Text]
12. Canadian Study of Health and Aging Working Group1994. Canadian Study of Health and Aging: study methods and prevalence of dementia. Can Med Assoc J. 150:899-913. [Abstract]
13. Teng EL, Chui HC, 1987. The modified Mini-Mental State (3MS) Examination. J Clin Psychiatry. 48:314-318. [Medline]
14. Folstein MF, Folstein SE, McHugh PR, 1975. "Mini-Mental State". A practical method for grading the cognitive state of patients for the clinician. J.Psychiatry Res. 12:189-198.
15. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. Rev 3rd ed. Washington, DC: American Psychiatric Association; 1987.
16. Shatenstein B, Kergoat M-J, Nadon S. Anthropometric indices and their correlates in cognitively intact and demented elderly Canadians. Can J Aging. In press.
17. White H, Pieper C, Schmader K, Fillenbaum G, 1996. Weight change in Alzheimer's disease. J Am Geriatr Soc. 44:265-272. [Medline]
18. Ryan C, Bryant E, Eleazer P, Rhodes A, Guest K, 1995. Unintentional weight loss in long-term care: predictor of mortality in the elderly. South Med J. 88:721-724. [Medline]
19. Payette H, 1997. La malnutrition: est-elle un corrélat inévitable de la démence chez les personnes &acirc;gées?. Hébert R, Kouri K, Lacombe G, , ed.Vieillissement cognitif normal et pathologique. Actes du congrès scientifique, Sherbrooke, 1997. Institut universitaire de gériatrie de Sherbrooke Edisem, Boucherville, Quebec.

This article has been cited by other articles:

				D. S. Knopman, S. D. Edland, R. H. Cha, R. C. Petersen, and W. A. Rocca Incident dementia in women is preceded by weight loss by at least a decade Neurology, August 21, 2007; 69(8): 739 - 746. [Abstract] [Full Text] [PDF]

				R. Stewart, K. Masaki, Q.-L. Xue, R. Peila, H. Petrovitch, L. R. White, and L. J. Launer A 32-Year Prospective Study of Change in Body Weight and Incident Dementia: The Honolulu-Asia Aging Study Arch Neurol, January 1, 2005; 62(1): 55 - 60. [Abstract] [Full Text] [PDF]

				R. Peila, B. L. Rodriguez, L. R. White, and L. J. Launer Fasting insulin and incident dementia in an elderly population of Japanese-American men Neurology, July 27, 2004; 63(2): 228 - 233. [Abstract] [Full Text] [PDF]

				Z. S. Tan, S. Seshadri, A. Beiser, P. W. F. Wilson, D. P. Kiel, M. Tocco, R. B. D'Agostino, and P. A. Wolf Plasma Total Cholesterol Level as a Risk Factor for Alzheimer Disease: The Framingham Study Arch Intern Med, May 12, 2003; 163(9): 1053 - 1057. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation