|Title||Anthropometrics and ageing : impact of weight status on health|
|Author(s)||Hollander, E.L. de|
|Source||University. Promotor(en): Lisette de Groot, co-promotor(en): W.J.E. Bemelmans. - S.l. : s.n. - ISBN 9789461736888 - 176|
Chair Nutrition and Health over the Lifecourse
|Publication type||Dissertation, internally prepared|
|Keyword(s)||antropometrie - antropometrische dimensies - lichaamsgewicht - verouderen - ouderen - gezondheid - anthropometry - anthropometric dimensions - body weight - aging - elderly - health|
|Categories||Human Nutrition and Health|
Background: Weight status is one of the factors that influence healthy ageing. It is often assessed with anthropometric measures such as body mass index (BMI) and waist circumference (WC), which indicate underweight or excess fat. Both are associated with adverse health outcomes in adults. The first paper of this thesis investigates whether this association is consistent over calendar time, to check for possible influences of improved healthcare procedures over time. In old age, this association is unclear. Using several anthropometric measures, the subsequent five papers examine the impact of weight status and development of weight status on coronary heart disease (CHD), mortality, and quality of life (QoL) among the elderly and during ageing.
Methods: A meta-regression analysis of 31 international cohort studies (n=389,212) was used to estimate the multivariable adjusted relative risk (RR) of CHD for an increased BMI and whether the RR was different between calendar periods (i.e. studies that started before 1985 and studies that started after 1985) taking account of the age of the population. Associations of BMI and changes in eight anthropometric measures with all-cause and cause-specific mortality in old age were studied by means of multivariable Cox regression analyses using data from the Survey in Europe on Nutrition and the Elderly: a concerted action study including 70–77-year-olds (n=1,061–1,970). Moreover, the association of WC with all-cause and cause-specific mortality was studied by means of a meta-analysis of 29 international cohort studies including 65–74-year-olds (n=58,609). For an ageing population, we used the Doetinchem Cohort Study including 20–70-year-olds (n=3,408–4,135) and three to four repeated measures of weight and height over a period of 10 to 15 years. In this study population, we used a multivariable regression analysis to examine the association of changes in weight and long-term BMI patterns with QoL (measured by the SF-36 questionnaire).
Results: After simultaneous inclusion of calendar period and age of the population in the model, the meta-regression analysis showed no difference in the RR of CHD in the association with a high BMI between calendar periods. However, a 10-year increment in population age lowered the 1.28 [95%confidence interval (CI): 1.22–1.34] RR of CHD for a five-BMI-unit increment by 29% (95%CI: -55 to -5). Among the elderly, BMI was associated with all-cause mortality, cardiovascular disease (CVD) mortality, and mortality due to causes other than CVD, cancer, and respiratory diseases (p<0.05). A BMI below 24 kg/m2and above 30 kg/m2were the thresholds at which risks of cause-specific mortality were increased by 10%. WC was associated with all-cause, CVD, cancer, and respiratory disease mortality (p<0.05). At the levels for abdominal obesity (102 cm, men; 88 cm, women), the risk of all-cause and CVD mortality was not significantly increased, or only modestly. A risk of 2.0 (clinically relevant) for all-cause and CVD mortality was associated with a WC of 132 and 123 cm in men, and 116 and 105 cm in women, respectively. By using a combination of WC and BMI categories with the combination of a small WC (94 cm, men; 80 cm, women) and a healthy weight (20.0–24.9 kg/m2) as the reference, we observed the highest all-cause and CVD mortality risk of approximately 2.0 for underweight (<20.0 kg/m2; in combination with a small WC), and abdominal obesity within healthy ranges of BMI. Changes in BMI and WC were not associated with all-cause and CVD mortality, except for a decrease in WC ≥3.1 cm in the association with all-cause mortality (1.52, 95%CI: 1.01–2.31). Similarly, a decrease in weight ≥3.2 kg was associated with a 1.48 (95%CI: 0.99–2.20) increased all-cause mortality risk. Moreover, both a decrease and an increase in mid-upper arm circumference (MUAC) were associated with all-cause mortality and CVD mortality. A decrease of ≥1.6 cm and 0.6–1.6 cm in MUAC was associated with a 1.81 (95%CI: 1.17–2.79) and a 1.66 (95%CI: 1.10–2.49) all-cause mortality risk. An increase of ≥1.3 cm in MUAC was associated with a 1.52 (95%CI: 1.00–2.31) all-cause mortality risk and a 1.94 (95%CI: 1.00–3.75) CVD mortality risk. In an ageing population, we found that weight gain, especially weight gain of >6 kg, resulted in a decline in QoL. Weight loss (>2 kg) did not result in large changes in QoL. However, both weight gain and weight loss were adversely associated with changes in QoL as compared to a stable weight (changes ≤2 kg). From examination of long-term BMI patterns, the lowest QoL was observed for the ‘persistent obesity (≥30 kg/ m2)’ pattern. The BMI patterns, ‘persistent obesity’, ‘developing overweight (25.0–29.9 kg/m2’, ‘developing obesity’, and ‘switching between BMI categories’ scored 1.8–11.6 points (p<0.05) lower on QoL than the ‘persistent healthy weight (18.5–24.9 kg/m2)’ pattern. The BMI pattern ‘persistent overweight’ generally did not differ from the ‘persistent healthy weight’ pattern. These findings were consistent among age groups.
Conclusions: Although the risk of CHD in the association with BMI attenuated with increasing age, we found associations of BMI and WC with all-cause and cause-specific mortality among the elderly. These anthropometric measures can be used as single predictors of mortality for the elderly, but higher cut-off points for BMI and WC to indicate underweight and excess fat should be considered. Moreover, a combination of these two anthropometric measures can be recommended, as that would provide more information of the body composition than one anthropometric alone. With regard to assessing changes in body composition, MUAC might be recommended for the elderly. Furthermore, a stable weight is best for health maintenance among all ages, provided this stable weight does not fall within the extreme values of weight, i.e. too light or too heavy. In all, our results underscore the value of anthropometric measures in the management of weight and the importance of the maintenance of a stable weight during ageing.