• 検索結果がありません。

Unintentional Weight Loss in the Elderly at a Nursing Home in Japan : Time of Onset, Changes, and Association with M ortality

N/A
N/A
Protected

Academic year: 2021

シェア "Unintentional Weight Loss in the Elderly at a Nursing Home in Japan : Time of Onset, Changes, and Association with M ortality"

Copied!
8
0
0

読み込み中.... (全文を見る)

全文

(1)

Unintentional Weight Loss in the Elderly

at a Nursing Home in Japan :

Time of Onset, Changes, and Association with M ortality

Eiko Abe,

Kunihiko Hayashi,

Yasuhiro Matsumura

and Yuichi Sugai

Background: Although unintentional weight loss is a frequently encountered problem in care settings, little is known about when it starts. The authors observed body weight longitudinally in an elderly population and examined its association with mortality. M ethods: Body weight was monitored in residents who lived in a nursing home for the elderly in Tokyo between fiscal years 2002 and 2004,with the final observation date set at October 2008 for survivors and at the date of death for those who died. The 3-year period before the final date was divided into six periods. A linear regression coefficient was calculated as the rate of weight change for each period and compared between survivors and those who died. Results: In the mortality group, significant weight loss was seen from 24 months before death : −0.42kg, −0.62kg, −0.90kg, and −1.78kg in 19-24 months, 13-18 months, 7-12 months, and 1-6 months before death, respectively. In the survival group, there was no significant change in any period. Logistic regression analysis showed that weight change adjusted by sex,age,dementia,and BMI was significantly associated with mortality. The weight loss was large in residents with dementia. Conclusions: Weight loss began 2 years before death. The findings suggest the importance of daily weight measurements to detect changes associated with mortality.(Kitakanto Med J 2011;61:471∼478)

Key words: unintentional weight loss, mortality, elderly people, dementia, nursing home

Introduction

Unintentional weight loss in elderly people,which refers to a reduction in total body weight that occurs involuntarily over time, is a frequently encountered problem in care settings. It is clearly distinguished from intentional weight loss,which refers to the loss of total body mass in an effort to improve fitness and health.

Unintentional weight loss is an important health risk factor among elderly persons and a strong predic-tor of functional limitations, admission to hospital or a nursing home, and mortality.

In clinical practice, unintentional weight loss is seen in 8% of adult outpatients and 27% of frail elderly people aged 65 years, while in most healthy elderly people, body weight is maintained at a stable level with some gains and losses, with unintentional

weight gain evident in around 10-30%. Over the age of 70-75 years, even in healthy individuals, appe-tite declines, energy intake is reduced, and subcutane-ous fat and muscle decrease(sarcopenia) as a result of aging, meaning that average weight decreases. This is known as anorexia of aging. Anorexic factors include: biological changes related to food intake,such as changes in digestive function and taste, as well as decreased metabolism and hormone secre-tion ; somatosensory changes, such as digestive func-tion and taste; psychological factors, such as depres-sion and dementia; social factors,such as poverty and provision of preferred foods; and medical factors, such as diseases and medication. These factors interact with each other and are the main causes of unintentional weight loss.

Unintentional weight loss becomes a health issue in the case of weight loss of around 1% of body

1 Gunma University Graduate School of Health Sciences, 3-39-22 Showa-machi, Maebashi, Gunma 371-8514, Japan 2 Faculty of Health Care,Kiryu University,606-7 Azami Kasakake-cho,Midori,Gunma 379-2392,Japan 3 Tokyo Dementia Care Research and Training Center, Takaido-nishi 1-12-1, Suginami-ku, Tokyo 168-0071, Japan

Received : August 19, 2011 Accepted : October 6, 2011

Address: EIKO ABE : c/o K.Hayashi, Gunma University Graduate School of Health Sciences, 3-39-22 Showa-machi, Maebashi, Gunma 371-8514, Japan

(2)

weight in a normal elderly person, and of 5% over a 6-to 12-month period or 1-4.5kg (10 lb)in frail elderly people following admission to or discharge from hospital,in those living in elder-care facilities,or in dementia sufferers. Unintentional weight loss that increases in association with disease progression is regarded as a particular problem. Despite the fact that unintentional weight loss is thus evident in many elderly people and carries a higher risk of mortality compared with intentional weight loss, meaning that its early detection and treatment are important, little is known about it in Japan. As of 2011, Japanese society is aging rapidly,with 23.1% of the population currently aged 65 years. Since unintentional weight loss may well become an impor-tant health issue for elderly people in the future, this paper attempts to elucidate when it starts and how it changes in an elderly Japanese population.

M ethods

Subjects and Informed Consent

The subjects were 311 individuals aged between 60 and 103 years who lived as residents to the Yokufuukai Dai-San Nanyoen nursing home for the elderly in Tokyo during the 3-year period between April 1,2002 and March 31,2005,and were monitored until October 1, 2008. The average monitoring period was 50.2± 25.3 months (maximum 82 months). For the purpose of analysis,the subjects were divided into survival and mortality groups by outcome on October 1,2008. The final observation date was set at October 1, 2008 in survivors and the date of death for those who died. A total of 57 residents was excluded from the analysis because physical measurements were not performed, body weight was not recorded during long-term hospi-talization of 1 year, or the outcome was unknown.

Five residents who were monitored for< 3 months after admission were also excluded on account of their psychological and physical instability due to the change in environment. After these 62 individuals had been excluded, the data of 249 individuals (51 males, 198 females) were used for statistical analysis.

Informed consent was obtained from residents or their families in accordance with the nursing homes regulations. The research protocol was considered and approved by the Ethics Committee of the Gunma University Faculty of Medicine. Use of residents records was permitted by the Ethics Committee of the Yokufuukai Geriatric Hospital.

Data Collection and M easurements

Basic data for the investigation, including name, date of birth, sex, date of admission and/or leaving, and lifestyle information, as well as medical

informa-tion such as height, weight, and disorders, were obtained from the data management system at Yo-kufuukai Dai-San Nanyoen and from medical and long-term care records. Dementia was diagnosed according to the DSM-IV criteria and cognitive screen-ing, with reference to X-ray, CT, and MRI images of the head and the Mini-Mental State Examination (MMSE). If residents had been transferred to the Yokufuukai Geriatric Hospital and died,medical data were obtained from hospital records. The residents height (cm) was measured by nurses on admission in accordance with nursing home regulations. For weight, values measured on admission and monthly during the day on the third or fourth Sunday of every month were used. Measurements were carried out by care workers or nurses using a digital scale located in the bathroom, avoiding times before baths or directly after meals and with residents lightly dressed,and they were recorded in 0.1kg increments.

Division of M onitoring Periods

The subjects were monitored from admission until the final monitoring date, but since elderly peoples weight may fluctuate from day to day, it is difficult to discern changes over a short period, and analysis was therefore performed by dividing the 3 years before the final monitoring date into six 6-month periods, work-ing backward from the date of final monitorwork-ing : period 1, 1-6 months before; period 2, 7-12 months before; period 3, 13-18 months before; period 4,19-24 months before; period 5, 25-30 months before; and period 6, 31-36 months before the final date. Calculation of Weight Indices and Data Handling

(1) BM I and Strata

Body mass index (BMI) was calculated in kg/m as body weight divided by the square of height on admission. All BMI values were calculated using height on admission. When necessary for analysis, average BMI values during a single period (6 months) were partitioned at 18.5kg/m , with values <18.5kg/ m regarded as low BMI.

(2) Rate of Weight Change

In this study, the rate of weight change (in kg/6 months) was defined as six times the regression coeffi-cient of the regression line(x ,y),where y was individ-ual weight, and x was the period before the final monitoring date(or date of death for individuals who died) (1 x 6). For example, 1.5kg/6months indicates weight loss at an average of 1.5kg over a 6 month period. The weight loss rate is capable of conveying the trend of change,even if data are missing or dramatic change occurs within a single period,since it minimizes their effects. In this study,if weight loss

(3)

occurred, the rate of weight change was expressed as rate of weight loss.

(3) Weight Data Imputation

Many residents were transferred to hospital, with an average of 1.3 hospital admissions (for an average period of 116 days) per person during the monitoring period. During hospitalization, weight was not mea-sured due to symptoms or treatment in many cases, meaning that records were frequently irregular or missing. Even for residents who were not transferred to hospital, weight records were frequently missing because measurements could not be made regularly for reasons such as symptoms of dementia or poor health. Accordingly, for the purpose of analysis, missing records were imputed by the Last Observation Carried Forward (LOCF) method for up to five months (1 n 5).

Statistical Analysis

Univariate analysis was first carried out with respect to sex, age on admission, average monitoring period, weight, BMI, and the presence or absence of dementia as characteristics on admission. To investi-gate sex differences, a t-test was used for continuous variables, and a χ test was used for categorical vari-ables. Rates of weight change in all periods were investigated using the 2-way t-test described below.

First,a t-test was performed to investigate whether the average rate of change was 0, to confirm that the rate of change was significant. Second, a t-test was performed to investigate whether there was any differ-ence between the rates of change in the survival group and mortality group.

In addition,logistic regression analysis was perfor-med with survival or death as the objective variable and rate of weight change, the unit of change, as the explanatory variable, in order to investigate whether there was any association between rate of weight change and prognosis,and the mortality odds ratio was calculated.

Univariate analysis was also carried out for factors anticipated to constitute confounding factors for mortality. In this study, four such factors were identified : sex (M/F), age ( 85 years, <85 years), BMI on admission (mortality per -1kg/m ), and dementia (dementia/non-dementia).

Multivariate logistic regression analysis was per-formed with these factors as adjustment factors,stratifi-cation was carried out by factors for which there was an association,and stratified analysis using a t-test was performed to investigate whether or not there were any differences in rate of weight change between the sur-vival and mortality groups in each period. During multivariate logistic regression analysis, adjustment

factors for category variables were converted to dummy variables. All analyses were performed using JMP software (version 8.02 SAS Institute, Tokyo, Japan).

Results

The residents basic characteristics on admission are shown in Table 1. Women were older than men on admission (p<0.001),their period of residence was longer (p<0.01), and their mean age on the final monitoring date was older (p<0.001). The average age on the final monitoring date was 84.8 years for men and 89.8 years for women,with both men and women showing longevity. All subjects had chronic disease and required both medical and long-term care. In particular,approximately 80% suffered from dementia and,therefore,required specialized care by staff due to its characteristic symptoms.

Overall,29% of individuals at the end stage in the mortality group and 21% of individuals at the final observation in the survival group used tube feeding or a gastrostomy at the end stage because of a swallowing disorder, while 51% of individuals at the end stage in the mortality group and 49% of individuals at the final observation in the survival group were on a soft diet because of a swallowing disorder. The most frequent cause of death was pneumonia (including acute and aspiration pneumonia) in 35%, followed by heart failure in 10%, acute respiratory failure (including respiratory arrest) in 7%, cancer/malignant tumor in 6%,and cerebral infarction in 6%. In the mortality group, half of the residents had died of pneumonia, heart failure, or acute respiratory failure (respiratory arrest) (table was not shown).

Rate of Weight loss in the M ortality Group

Figure 1 and Table 2 shows the rates of weight change in the mortality and survival groups.

In the mortality group,significant progress in rates of weight loss was seen continuously from period 4 to period 1 (p<0.05−p<0.001) (Figure 1, Table 2). The rate of weight loss was the largest in period 1, followed by periods 2,3,and 4,meaning that the closer individuals came to death, the more weight they lost. In the survival group, the rate of weight change was positive from period 4 to period 3 but weight loss was evident during period 2 (7-12 months before), and period 1 (1-6 months before). There were no signifi-cant differences between any periods (Figure1, Table 2).

Logistic regression analysis showed that rates of weight loss from period 4 to period 1 were significant and associated with mortality. The odds ratio per −1.0kg/6 months(95% confidence interval[CI])was shown in Table 3.

(4)

Univariate analysis for factors anticipated to con-stitute confounding factors or covariates (sex, age on admission, BMI on admission[per −1kg/m ], and dementia), with survival or mortality as the objective variable, showed that all of these were significant and associated with mortality(Table 3).

The same results were obtained using the stepwise method, and modeling using sex, age on admission, dementia, and BMI on admission as the four

adjust-ment factors showed that the rate of weight loss was significantly associated with mortality in periods 4 to 1 (Table 3).

In light of these results, stratification by age on admission, dementia, and BMI on admission was performed, and rates of weight loss between the sur-vival and mortality groups were compared. Stratifica-tion analysis by sex was not performed because there were only nine male survivors on the final monitoring

Fig.1 Comparison of weight change in the mortality and suvivar groups in each period. A t-test was used to compare the two groups. Rate of weight change in each group were investigated by performing a t-test with the hypothesis that the average was 0 (results in bars on graph)

Table 1 Patients characteristics on admission and status on final observation date Men :

51 (20.5%) 198 (79.5%)Women : 249 (100%)Total: Characteristics on admission Mean±SD Mean±SD p value Mean±SD

Age (years) 81.6±7.0 85.6±7.4 <0.001 84.7±7.5

85 years (n) 17 ( 6.8%) 115 (46.2%) <0.01 132 (53.0%) Average monitoring period

(residence in nursing home) (months) 40.1±24.9 52.8±24.8 <0.01 50.2±25.3

Height (cm) 156.4±7.8 144.0±6.7 <0.001 147±9.0 Weight (kg) 47.5±7.8 39.9±7.4 <0.001 41.4±8.0 BMI (kg/㎡) 19.3±2.7 19.3±3.7 0.87 19.3±3.5 BMI <18.5kg/㎡ (n) 17 ( 7.0%) 86 (35.2%) 0.29 103 (42.2%) Dementia sufferers (n) 45 (18.0%) 171 (68.7%) 0.73 216 (86.7%) Alzheimer 23 96 119 cerebrovascular dementia 12 19 31 senile dementia 4 29 33 mixed dementia 3 4 7 unknown 3 23 26

Other mental disorders (n) 1 5 6

Status on final monitoring

Survivors (n) 9 ( 3.6%) 65 (26.1%) 74 (29.7%)

Deathes (n) 42 (16.9%) 133 (53.4%) 175 (70.3%)

(5)

date, making it difficult to obtain accurate analytical results.

Comparison of Rates of Weight Loss Stratified by Age, Dementia, and BM I on Admission

Excluding sex,stratification was performed by age on admission ( 85 years, <85 years), dementia (dementia, non-dementia), and BMI on admission ( 18.5kg/m , <18.5kg/m ), and rates of weight change were compared between the mortality and survival groups. The results are shown in Table 2.

(1) Stratification Analysis by Age

Significant weight loss was evident among those aged 85 years in the mortality group from period 3 to period 1(p<0.05−p<0.001). Significant weight loss was also evident among those aged 85 years in the survival group in periods 2 and 1 (p<0.05). There was a significant difference between the mortality and survival groups for those aged 85 years in period 3 only(p<0.05),with a tendency for weight loss evident in both groups. In periods 2 and 1 in particular, the

rate of weight loss in the survival group approached that of the mortality group, with no significant differ-ence between the two groups,and a test of whether or not the hypothetical average was 0 showed that signifi-cant weight loss occurred (p<0.05). Signifisignifi-cant weight loss was observed among those aged <85 years in the mortality group, in periods 2 and 1 (p<0.05− p<0.001),but in the survival group,no weight changes were evident in any period.

(2) Stratification Analysis by Dementia

Significant weight loss was evident among those with dementia in the mortality group from period 4 to period 1 (p<0.05−p<0.001).

There was no significant weight loss in any period among those with dementia in the survival group. Significant weight loss was evident among those with no dementia in the survival group (in period 1; p< 0.05). There was no significant weight loss in any period among those within the non-dementia survival group as well.

(3) Stratification Analysis by BM I on Admission Significant weight loss was evident among those

Table 2 Rates of weight change during each period after stratification (Unit: kg/6 months) Total Age on admission Dementia BMI on admission

Period MortalityGroup (n=175)

Survival Group (n=74)

85y <85y Dementia non-domentia 18.5 <18.5 Mortality Survival Mortality Survival Mortality Survival Mortality Survival Mortality Survival Mortality Survival Period 1 n 170 73 99 30 71 43 153 57 17 16 94 42 74 28 Rate of Weight change -1.777 -0.246 -1.389 -1.236 -2.318 0.445 -1.799 -0.383 -1.580 0.241 -2.141 -0.289 -1.367 -0.230

p ※ 1 n.s.

p ※ 2 n.s. n.s. n.s. n.s. n.s. n.s.

Period 2 n 165 73 95 30 70 43 149 57 16 16 92 42 71 28 Rate of Weight change -0.897 -0.021 -0.831 -0.622 -0.986 0.397 -0.913 -0.140 -0.750 0.403 -1.231 -0.149 -0.477 0.067

p ※ 1 n.s. n.s. n.s.

p ※ 2 n.s. n.s. n.s. n.s. n.s. n.s. n.s.

Period 3 n 143 73 82 30 61 43 130 57 13 16 78 42 63 28 Rate of Weight change -0.619 0.140 -0.848 0.446 -0.312 -0.074 -0.735 0.383 0.540 -0.725 -0.894 0.220 -0.204 -0.012

p ※ 1 n.s. n.s. n.s.

p ※ 2 n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. Period 4 n 129 74 70 30 59 43 116 58 13 16 73 43 54 28 Rate of Weight change -0.415 0.150 -0.433 -0.137 -0.394 0.356 -0.460 0.263 -0.012 -0.260 -0.611 0.183 -0.108 0.088

p ※ 1 n.s. n.s. n.s. n.s.

p ※ 2 n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. Period 5 n 114 74 58 31 56 43 101 58 13 16 68 43 44 28 Rate of Weight change -0.044 0.363 -0.057 0.480 -0.030 0.278 -0.086 0.498 0.284 -0.129 -0.167 0.077 0.261 0.864

p ※ 1 n.s. n.s. n.s. n.s. n.s. n.s. n.s.

p ※ 2 n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. Period 6 n 101 74 50 31 51 43 90 58 11 16 60 43 40 28 Rate of Weight change 0.282 0.404 -0.028 -0.017 0.585 0.707 0.335 0.514 -0.153 0.005 0.192 0.091 0.405 0.920

p ※ 1 n.s. n.s. n.s. n.s. n.s. n.s. n.s.

p ※ 2 n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. p ※ 1 t-test comparing survival and mortality groups p ※ 2 t-test with hypothetical average 0

n.s.: no significant difference p< 0.001 p<0.01 p<0.05 Table 3 Factors affecting survival prognosis ※

Univariate Multivariate

factor Discription Odds ratio CI p value Odds ratio CI p value

Sex M/F 2.28 1.09-5.26 <0.05 ―

Age on admission 85/<85 1.89 1.10-3.30 <0.05 ―

Dementia 2.56 1.21-5.43 <0.05 ―

BMI on admission Per BMI -1kg/㎡ 1.09 1.01-1.18 <0.05 ―

Rate of weight loss Period 1 4.00 1.94-8.55 <0.01 3.64 1.77-8.20 <0.001 Period 2 2.57 1.27-5.53 <0.05 2.57 1.19-5.79 <0.05 Period 3 2.31 1.13-4.83 <0.05 2.44 1.06-5.53 <0.05 Period 4 2.57 1.00-6.61 <0.05 2.70 1.06-7.53 <0.05 Period 5 1.50 0.74-3.14 0.27 1.87 0.83-4.20 0.12 Period 6 1.13 0.53-2.57 0.72 1.42 0.61-3.46 0.65 * Analysis involved logistic regression analysis with survival/mortality as the objective variable

(6)

with BMI 18.5kg/m on admission in the mortality group from period 4 to period 1 (p<0.05−p<0.001). There was no significant weight loss in any period among those with BMI 18.5kg/m on admission in the survival group. Significant weight loss was evi-dent among those with BMI <18.5kg/m on admission in the mortality group (in period 2 and period 1; p< 0.05−p<0.001).

Both weight gain and weight loss occurred among those with BMI <18.5kg/m on admission in the survival group, with significant weight gain of in period 6 (31-36 months before) and period 5 (25-30 months before) (p<0.05). There was no significant weight loss in any period among those with BMI <18.5kg/m on admission in the survival group.

Discussion

Unintentional weight loss in the mortality group began years before death and continued until death, with a gradual increase as death approached. In contrast to the unintentional weight loss seen in the mortality group, almost no change occurred in the survival group. This tendency was unchanged even when the data were stratified,with continuous uninten-tional weight loss evident among all mortality groups and associated with death. In addition, when the changes in body weight were calculated as percentage change, the results were the same. Previous studies have indicated that the amount of weight loss increases the risk of mortality, but in many cases, these studies covered comparatively short monitoring periods of 6 months to 2 years. In addition, in many of these studies, weight measurements were only carried out at long intervals, such as once a year, and weights were often self-reported rather than actually measured. There have been almost no detailed studies addressing the questions of the timing of onset of weight loss that increases the risk of mortality,or of what sort of course it takes. In this study, residents were followed and assessed on the basis of up to 82 months of monitoring by,in principle,actually measuring weight regularly at monthly intervals.

The speed of weight loss was slow at first,at about 400g/6 months at the point when weight loss started, but it increased by around 4.5 -fold to around 1.8kg/ 6 months by immediately before death. Uninten-tional weight loss was found to start years before death. The main causes of such unintentional weight loss are disease, declining physical function due to aging, and reduced appetite and energy intake due to complex psychological, social, and other factors, with undernutrition occurring in many cases. In this study,around 80% of subjects suffered from moderate or severe dementia. Dementia onset and progression

are also associated with weight loss ; this may have affected our results. It has been reported that individuals with dementia eventually develop a swall-owing disorder, and approximately 80% of them lose weight (BMI<18.5kg/m ). Although we cannot exclude the possibility that a swallowing disorder caused the loss of weight in residents with dementia,it is unlikely that a swallowing disorder critically affected the weight loss in this study,since there was no significant difference in the percentage use of tube feeding or gastrostomy, as well as in the use of a soft diet, between patients with and without dementia. Heart failure, which is the second-leading cause of death in residents, often causes weight loss.

In addition, it is also known that patients with heart failure increase their body weight because of edema. Therefore, it is possible that heart failure could have a considerable effect on body weight. However,the present results showed that weight loss in individuals with heart failure in the mortality group was not significantly different from that of individuals without heart failure. Patients with malignant tumor generally lose weight due to cachexia. Therefore, it should be considered whether the residents suffered from cancer. Interestingly, the present data showed that only 6% of individuals died of cancer, which is much less than the percentage of Japanese who die of cancer(29.5%). In the present study heart failure and pneumonia were more frequent than cancer as causes of death in the residents in the present study,which is consistent with that in typical Japanese in 2010. There have also been reports that potential factors, including cognitive decline, reduced food intake, rest-lessness,wandering,impaired senses of smell and taste due to brain contraction, and elevated rate of energy expenditure, may also contribute to weight loss in dementia sufferers, putting them at high risk of losing weight. Weight gain and loss are mainly due to increases and decreases in fat and muscle,and aging is a major factor in elderly people. In the vast majority of people, aging unavoidably results in a decline in digestive function,sense of taste,metabolism,hormone secretion,food intake function,cognitive function,and psychological and intellectual function, as well as muscle loss due to anorexia of aging caused by their interaction.

The social (poverty,eating alone,physical disabil-ity) and medical (disease and medication) changes associated with aging mesh like gears with the biology of aging to reduce energy intake, causing fat and muscle loss. Muscle loss due to aging is theor-ized to occur more easily in men than in women as a result of age-related declines in sex hormones. In the present study,men were at higher risk of mortality,

(7)

indicating that the effect of rate of weight loss on mortality may differ between the sexes, but as there were few surviving men, this could not be verified in detail. This issue requires further study. When the relationship between weight loss and mortality was viewed in terms of age strata,unintentional weight loss began 2 years before death in both the groups aged 85 years and <85 years, suggesting that uninten-tional weight loss is a mortality risk regardless of age. Significant weight loss was evident in the survival group aged 85 years from around 1 year before the final monitoring date, and the possibility that this marked the start of unintentional weight loss cannot be ruled out, suggesting that there is a tendency for unintentional weight loss to occur in individuals of advanced age. When the relationship between weight loss and mortality was viewed in terms of BMI strata on admission, unintentional weight loss began in the mortality group in both BMI groups (BMI 18.5kg/ m and <18.5kg/m ), suggesting that unintentional weight loss is a mortality risk, regardless of BMI. However, weight loss began earlier in the group with BMI 18.5kg/m compared with the group with BMI <18.5kg/m , and the rate of weight loss was higher. The mechanism of this is unknown, but Newman et al. reported that more muscle is lost during weight loss when body fat mass is low and muscle mass is high, and that muscle mass is lost at a faster speed during weight loss than it is gained during weight gain,suggesting that weight loss may accelerate sarcopenia. Therefore, it is possible that the group with BMI 18.5kg/m had greater muscle mass compared with the group with BMI <18.5kg/m ,and sarcopenia was accelerated, suggesting that body type may affect the course of the rate of weight loss. In this study, however, 80% of the subjects suffered from dementia, and their symptoms meant that, in many cases, they refused to undergo body fat and muscle measurements,making this impossible to verify. This remains a topic for future study. The results of this study clearly show that unintentional weight loss increases mortality risk, and a discussion of how to deal with this is required. Keller et al. have reported that weight gain of 5% reduces both mortality and morbidity. Faxen-Irving et al. also reported that long-term mortality is higher for BMI <23kg/m , whereas survival is good for BMI 23kg/m . This suggests that nursing home residents in whom uninten-tional weight loss is considered to have started should first undergo swallowing assessments, after which attempts should be made to prevent further weight loss by means such as changing the content or texture of their diet and increasing the frequency of meals, in-cluding snacks,to increase their food intake. Because

low body weight results in a decline in psychological state and reduced physical activity, preventing low body weight by preventing further weight loss will at least help maintain the quality of life(QOL)of elderly people. In the present study, weight change was monitored retrospectively. However, in order to improve frequency and accuracy, we should also con-sider to monitor weight change prospectively in the future study.

Unintentional weight loss associated with mortal-ity starts years before death and continues long-term. To prevent the progression of unintentional weight loss, it is important to measure weight daily to detect its onset at an early stage and to conduct a multifaceted approach to weight gain.

Conclusions

Unintentional weight loss started years before death and continued until death. Unintentional weight loss was slow at first,but its speed increased to around 4.5-fold immediately before death, and it was associated with mortality. These results were the same when tendencies in pre-death unintentional weight loss were adjusted for related factors (sex, age, BMI on admission, and dementia). Sex, aging, dementia, and BMI may also have affected the ten-dency for unintentional weight loss to occur and its course.

To prevent the progression of unintentional weight loss,it is important to measure weight daily to detect its onset at an early stage and to conduct a multifaceted approach to weight gain.

CONFLICT OF INTEREST : The authors have no conflicts of interest.

FOUNDING/SUPPORT : This study was supported in part by funds from a grant-in-aid scientific research

KAKENHI> (C) No.19659615

ACKNOWLEDGEMENTS : The authors thank the staff at the Yokufuukai Dai-San Nanyoen nursing home.

References

1. Alibhai SM,Greenwood C,Payette H. An approach to the management of unintentional weight loss in elderly people. CMAJ 2005; 172: 73-780.

2. Wernette CM, White BD, Zizza CA. Signaling Proteins that Influence Energy Intake may Affect Unintentional Weight Loss in Elderly Persons. J Am Diet Assoc 2011; 111: 864-873.

3. Lee JS,Kritchevsky SB,Harris TB et al. Short-term weight changes in community-dwelling older adults: the Health, Aging, and Body Composition Weight Change Substudy. Am J Clin Nutr 2005; 82: 644-650.

(8)

body mass index and weight change with all-cause mortality in the elderly. Am J Epidemiol 2006; 163: 938-949. 5. Dey DK, Rothenberg E,Sundh V et al. Body mass index,

weight change and mortality in the elderly. A 15 y longitu-dinal population study of 70y olds. Eur J Clin Nutr 2001; 55: 482-492.

6. Fine JT, Colditz GA, Coakley EH et al. A prospective study of weight change and health-related quality of life in women. JAMA 1999 ; 282: 2136-2142.

7. Sullivan DH, Liu L, Roberson PK et al. Body weight change and mortality in a cohort of elderly patients recently discharged from the hospital. J Am Geriatr Soc 2004; 52: 1696-1701.

8. Keller HH. Weight gain impacts morbidity and mortality in institutionalized older persons. J Am Geriatr Soc 1995; 43: 165-169.

9. Wang SY, Fukagawa N, Hossain M et al. Longitudinal weight changes,length of survival,and energy requirements of long-term care residents with dementia. J Am Geriatr Soc 1997; 45: 1189-1195.

10. Knoops KT, Slump E, de Groot LC et al. Body weight changes in elderly psychogeriatric nursing home residents. J Gerontol A Biol Sci Med Sci 2005; 60: 536-539. 11. Locher JL, Roth DL, Ritchie CS et al. Body mass index,

weight loss, and mortality in community-dwelling older adults. J Gerontol A Biol Sci Med Sci 2007; 62: 1389-1392.

12. Donini LM, Savina C, Cannella C. Eating habits and appetite control in the elderly: the anorexia of aging. Int Psychogeriatr 2003; 15: 73-87.

13. Newman AB, Lee JS, Visser M et al. Weight change and the conservation of lean mass in old age: the Health,Aging and Body Composition Study. Am J Clin Nutr 2005; 82: 872-878, quiz 915-6.

14. Cruz -Jentoft AJ,Baeyens JP,Bauer JM et al. Sarcopenia: European consensus on definition and diagnosis: Report of the European Working Group on Sarcopenia in Older People. Age Ageing 2010; 39 : 412-423.

15. Hays NP, Roberts SB. The anorexia of aging in humans. Physiol Behav 2006; 88: 257-266.

16. Buchman AS, Wilson RS, Bienias JL et al. Change in body mass index and risk of incident Alzheimer disease. Neurology 2005; 65: 892-897.

17. Gillette -Guyonnet S, Nourhashemi F, Andrieu S et al. Weight loss in Alzheimer disease. Am J Clin Nutr 2000; 71: 637S 642S.

18. Wang PN, Yang CL, Lin KN et al. Weight loss, nutri-tional status and physical activity in patients with Alz-heimers disease. A controlled study. J Neurol 2004; 251: 314-320.

19. Faxen-Irving G,Basun H,Cederholm T. Nutritional and cognitive relationships and long-term mortality in patients with various dementia disorders. Age Ageing 2005; 34: 136-141.

20. Abbasi AA, Rudman D. Undernutrition in the nursing home: prevalence, consequences, causes and prevention. Nutr Rev 1994; 52: 113-122.

21. Sullivan DH, Bopp MM, Roberson PK. Protein-energy undernutrition and life-threatening complications among the hospitalized elderly. J Gen Intern Med 2002; 17: 923-932.

22. Nakamura H,Fukushima H,Miwa Y et al. A longitudinal study on the nutritional state of elderly women at a nursing home in Japan. Intern Med 2006; 45: 1113-1120. 23. Lou MF,Dai YT,Huang GS et al. Nutritional status and

health outcomes for older people with dementia living in institutions. J Adv Nurs 2007; 60: 470-477.

Table 1  Patientʼ s characteristics on admission and status on final observation date Men :  
Table 2  Rates of weight change during each period after stratification (Unit: kg/6 months)

参照

関連したドキュメント

Analogs of this theorem were proved by Roitberg for nonregular elliptic boundary- value problems and for general elliptic systems of differential equations, the mod- ified scale of

Then it follows immediately from a suitable version of “Hensel’s Lemma” [cf., e.g., the argument of [4], Lemma 2.1] that S may be obtained, as the notation suggests, as the m A

Definition An embeddable tiled surface is a tiled surface which is actually achieved as the graph of singular leaves of some embedded orientable surface with closed braid

We give a Dehn–Nielsen type theorem for the homology cobordism group of homol- ogy cylinders by considering its action on the acyclic closure, which was defined by Levine in [12]

Correspondingly, the limiting sequence of metric spaces has a surpris- ingly simple description as a collection of random real trees (given below) in which certain pairs of

[Mag3] , Painlev´ e-type differential equations for the recurrence coefficients of semi- classical orthogonal polynomials, J. Zaslavsky , Asymptotic expansions of ratios of

Hence, for these classes of orthogonal polynomials analogous results to those reported above hold, namely an additional three-term recursion relation involving shifts in the

“Indian Camp” has been generally sought in the author’s experience in the Greco- Turkish War: Nick Adams, the implied author and the semi-autobiographical pro- tagonist of the series