Abstract
Objectives
Previous studies showed a U-shaped association between BMI and (physical) frailty. We studied the association between BMI and physical, cognitive, psychological, and social frailty. Furthermore, the overlap between and prevalence of these frailty domains was examined.
Design
Cross-sectional study.
Setting
The Doetinchem Cohort Study is a longitudinal population-based study starting in 1987-1991 examining men and women aged 20-59 with follow-up examinations every 5 yrs.
Participants
For the current analyses, we used data from round 5 (2008-2012) with 4019 participants aged 41-81 yrs.
Measurements
Physical frailty was defined as having ≥ 2 of 4 frailty criteria from the Frailty Phenotype (unintentional weight loss, exhaustion, physical activity, handgrip strength). Cognitive frailty was defined as the < 10th percentile on global cognitive functioning (based on memory, speed, flexibility). Psychological frailty was defined as having 2 out of 2 criteria (depression, mental health). Social frailty was defined as having ≥ 2 of 3 criteria (loneliness, social support, social participation). BMI was divided into four classes. Analyses were adjusted for sex, age, level of education, and smoking.
Results
A U-shaped association was observed between BMI and physical frailty, a small linear association for BMI and cognitive frailty and no association between BMI and psychological and social frailty. The four frailty domains showed only a small proportion of overlap. The prevalence of physical, cognitive and social frailty increased with age, whereas psychological frailty did not.
Conclusion
We confirm that not only underweight but also obesity is associated with physical frailty. Obesity also seems to be associated with cognitive frailty. Further, frailty prevention should focus on multiple domains and target individuals at a younger age (<65yrs).
Key words: BMI, frailty domains, cohort study
Introduction
Frailty can be described as the result of the accumulation of deficits (1) in physical (2), psychological or social functioning (3). This accumulation may lead to an increased vulnerability. One event, for example breaking a hip through falling, can have a major impact on frail individuals. The recovery process can take a long time and sometimes these individuals do not recover at all. In other words, frail individuals become relatively easily imbalanced and have difficulty returning to their balanced state (4, 5). Frailty also increases the risk for negative health outcomes (6, 7) including falls, disability, hospitalization, institutionalization, and mortality (2, 8, 9, 10). To prevent the development of these undesirable outcomes a first step is to identify frail individuals, preferably in an early stage. For this purpose, different frailty instruments have been developed over the years. The Frailty Phenotype for example, as first described by Fried (2), refers to physical frailty. The Frailty Index of Rockwood (11) is based on a broader definition of frailty and is built on the principle of the accumulation of deficits; it contains items regarding the presence of diseases, the ability to perform everyday activities, and physical and neurologic signs. Over time, the focus has shifted to other frailty domains, such as psychological and social frailty. In response to this development new frailty instruments were established. The Tilburg Frailty Indicator (12), the Edmonton Frail Scale (13), and the Groningen Frailty Indicator (14) are all instruments that include measurements for multiple frailty domains, which are aggregated into an overall score. In addition, frailty domains are also being studied individually. Currently, in particular the concept of cognitive frailty receives much attention (15, 16, 17, 18).
One important indicator for frailty is underweight or unintentional weight loss. The association between having a low body mass index (BMI) and frailty is acknowledged and is for example included in Fried's definition of the Frailty Phenotype as well as in other definitions. However, recent studies also found an association between obesity and frailty. An important finding, because the prevalence of overweight and obesity among the elderly is increasing (19). In fact, the results seem to indicate that there is a U-shaped association between BMI and (physical) frailty (20, 21, 22), so both ends of the weight continuum are of concern.
Here we report on the association between BMI and physical, cognitive, psychological and social frailty in the Doetinchem Cohort Study. We investigated whether having a low or high BMI is associated with a higher risk of physical frailty only, or if this U-shaped association also applies to cognitive, psychological, and/or social frailty. In addition, to gain more insight in these domains we studied the overlap between and prevalence of the different frailty domains.
Methods
Participants and study design
The Doetinchem Cohort Study is an ongoing longitudinal population-based cohort study, which started in 1987-1991. The study design is described by Verschuren et al (23). Briefly, the Doetinchem Cohort Study was designed to study the influence of lifestyle and biological risk factors on health over the life course. At baseline (round 1) men and women aged 20-59 years and living in Doetinchem, a provincial town in the Netherlands, were examined. From the 12439 participants who participated in the first round, a random sample of 7769 persons was re-invited for a follow-up study, to be reexamined every 5 years for 25 years. Those who were invited in round 2 (1993-1998) were invited again, excluding those who emigrated or actively withdrew from the study. The response rates varied between 75% and 80%, resulting in 4019 participants for round 5. For round 5, the study was approved according to the guidelines of the Helsinki Declaration by the Medical Ethics Committee of the University Medical Center Utrecht. Written informed consent was obtained from all participants during each examination round. For the current analyses, we used data from the fifth examination round (2008-2012) (n=4019). First, we excluded 20 participants with missing data for BMI (n=3999). Next, there were 619 missings regarding cognition data because cognitive function was tested only among participants ≥45 yrs. Finally, there were 8 missings for psychological frailty and 5 missings for social frailty. As a result, the analyses for physical frailty included 3999 participants, for cognitive frailty 3380 participants, for psychological frailty 3991 participants, and for social frailty 3994 participants.
Measurements
The study protocol consisted of questionnaires and physical, functional and biological measurements. Trained personnel performed all measurements in a standardized way.
Socio-demographic factors
Level of education was measured as the highest level reached during follow-up and categorized into low (intermediate secondary education or less), intermediate (intermediate vocational and higher secondary education) and high (higher vocational education or university).
Life-style
Smoking status was categorized into current smoker, former smoker and non-smoker. Being physically active was defined as adherence to the Dutch physical activity guideline, which recommends 30 minutes of moderate to vigorous physical activity per day on at least 5 days per week (24).
Body composition measurements
Body weight was measured to the nearest 100 g on calibrated scales and height to the nearest 0.5 cm. BMI was calculated and categorized into underweight < 20 kg/m2, normal weight 20-24.9 kg/m2, overweight 25-29.9 kg/m2, and obesity ≥ 30 kg/ m2.
Disease
Multimorbidity was defined as having two or more chronic diseases based on self-report. The following diseases were asked via self-report and included: diabetes, cancer, myocardial infarction, cerebrovascular accident, and chronic non-specific lung diseases.
Frailty criteria
An overview of the frailty criteria per domain are listed in Table 1. We used the Frailty Phenotype to define physical frailty (Fried), because this is a validated and widely used instrument. The cognitive, psychological, and social frailty domains are based on the Tilburg Frailty Indicator (TFI) (12) and the corresponding conceptual model of Gobbens (25). In the TFI, cognitive functioning is part of the psychological frailty domain. There is increasing support for the idea of cognitive frailty being a separate frailty domain (15, 16). Consequently, we constructed separate cognitive and psychological frailty domains, which are based on the TFI. A detailed description of the criteria can be found in the Supplementary material. Briefly, participants were considered to be physically frail (Fried) if they fulfilled ≥ 2 of 4 frailty criteria described by Fried (unintentional weight loss, exhaustion, physical activity, and handgrip strength) (2). Participants were considered cognitively frail when scoring <10th percentile on a global cognitive functioning score based on memory, speed, and flexibility. Cognitive scores were adjusted for level of education and number of tests performed during follow-up. Psychological frailty was defined as fulfilling both criteria for depression (26) and for general mental health (27). Social frailty was defined as meeting ≥ 2 of 3 criteria using the Loneliness scale (28), Social Support List-12 (29) and a questionnaire about social participation from the Dutch Municipal Health Services Elderly Monitor (30). Participants were considered to be physically frail (Gobbens) if they fulfilled ≥ 4 of 8 frailty criteria described by Gobbens (12). Physical frailty (Gobbens) was used for the sensitivity analysis.
Table 1.
Overview criteria per frailty domain
| Domains | Criteria | Cut-off | Based on |
|---|---|---|---|
| Physical frailty (Fried) (2) | - unintentional weight loss - exhaustion - low physical activity - reduced handgrip strength | ≥ 2 criteria | - unintentional weight loss: > 5% weight loss between round 4 and 5 and not being on a diet - low physical activity: meeting all of the following three criteria: 1) < 10th percentile of a physical activity score 2) < 25th percentile of walking hours per week during the last 12 months 3) failing to meet the Dutch physical activity guideline - exhaustion: 2 statements of the Center for Epidemiologic Studies Depression scale (26) - handgrip strength: dynamometer |
| Physical frailty (Gobbens) (12) | - unintentional weight loss - exhaustion - reduced handgrip strength - perceived health - limited in walking - disturbed balance - hearing impairment - vision impairment | ≥ 4 criteria | - unintentional weight loss (see physical frailty (Fried)) - exhaustion (see physical frailty (Fried)) - handgrip strength (see physical frailty (Fried)) - perceived health: one question of 36-Item Short-Form Health Survey (27, 36) - 1 question on 100 m walking - Tandem Stand Balance Test - 3 questions regarding hearing - 3 questions regarding vision |
| Cognitive frailty | - reduced global cognitive functioning | < 10th percentile | - global cognitive functioning score based on tests for memory, speed and flexibility: - 15 Words Verbal Learning Test (37) - Stroop Color–Word Test (38) - Word Fluency Test (39) - Letter Digit Substitution Test (40) |
| Psychological frailty | - depressive symptoms - mental health | = 2 criteria | - Center for Epidemiologic Studies Depression scale - Mental Health Inventory 5 (27) |
| Social frailty | - loneliness - little social support - low social participation | ≥ 2 criteria | - Loneliness Scale (28) - Social Support List-12 (29) - Questionnaire Dutch Elderly Monitor (30) |
Note: This table shows the criteria used per frailty domain. Physical frailty (Fried), cognitive frailty, psychological frailty, and social frailty were used to study the association between BMI and frailty. Physical frailty (Fried) and physical frailty (Gobbens) were used for the sensitivity analysis.
Our cut-off point for physical frailty (Fried) and physical frailty (Gobbens) deviate from the cut-off points described by these authors. Due to data unavailability we use 4 instead of 5 criteria for physical frailty (Fried) and adjusted the cut-off point so the prevalence is similar to the prevalence described per age group by Fried et al (2). We adjusted the cut-off point for physical frailty (Gobbens) from ≥ 3 of 8 criteria to ≥ 4 of 8 frailty criteria, so the prevalence is similar to the prevalence described by Fried.
Statistical analyses
Descriptive analyses were carried out for the non-frail population and for the physical, cognitive, psychological and social frail groups separately. For each frailty domain, the association between BMI and frailty was studied using a logistic regression model with classes of BMI as the independent variable. As mentioned above, BMI was categorized into four classes: underweight, normal weight, overweight, and obese. In the logistic regression model, we adjusted for sociodemographic characteristics (sex, age, and level of education) and smoking status, because these are considered to be potential confounders for frailty (1, 21, 31, 32). Age was considered as a continuous variable. Smoking status for the logistic regression model was categorized into smoking and non-smoking. The overlap between the different frailty domains was studied with frequency tables. In order to assess the effect of using different definitions of physical frailty, we performed a sensitivity analysis comparing physical frailty as defined by Fried (2) based on four criteria and physical frailty as defined by Gobbens (12) entailing eight criteria. We calculated the inter-rater agreement regarding the two definitions for physical frailty using Cohen's Kappa coefficient. All analyses were carried out in SAS 9.3 for Windows (SAS Institute Inc., Cary, NC, USA).
Results
Population characteristics
Characteristics of the non-frail population and the frail populations stratified by the different frailty domains are presented in Table 2. The mean age of the psychologically frail population was similar to the mean age of the non-frail population. In the other three frailty domains, the mean age was higher compared to the non-frail population. Psychological frailty was more common among women (68.9%) than men (31.1%). In contrast, cognitive frailty was more common among men (67.4%). Compared to the non-frail population, a low education, and current smoking were more prevalent, mean BMI was higher, and multimorbidity was higher in the frail population for all four domains.
Table 2.
Characteristics table of the non-frail and the frail populations
| Variables | Non-frail(n=2747) | Physically frail (Fried)(n=150) | Cognitively frail(n=310) | Psychologically frail(n=248) | Socially frail(n=162) |
|---|---|---|---|---|---|
| Socio-demographic | |||||
| sex (men) (%) | 1274 (46.4) | 58 (38.7) | 209 (67.4) | 77 (31.1) | 84 (51.9) |
| age (yrs) (SD) | 60.2 (8.8) | 63.8 (11.1) | 70.2 (6.8) | 59.2 (9.8) | 63.0 (10.4) |
| low level of education (%) | 1091 (39.7) | 77 (51.3) | 154 (49.7) | 141 (56.9) | 92 (56.8) |
| Body composition | |||||
| BMI1 (kg/m2) (SD) | 26.7 (4.0) | 28.1 (5.9) | 28.0 (4.3) | 27.4 (5.0) | 27.1 (4.7) |
| underweight (%) | 52 (1.9) | 8 (5.3) | - | 9 (3.6) | 6 (3.7) |
| normal weight (%) | 950 (34.6) | 41 (27.3) | 75 (24.2) | 69 (27.8) | 54 (33.3) |
| overweight (%) | 1247 (45.4) | 53 (35.3) | 156 (50.3) | 114 (46.0) | 71 (43.8) |
| obese (%) | 498 (18.1) | 48 (32.0) | 79 (25.5) | 56 (22.6) | 31 (19.1) |
| Life-style | |||||
| current smoker (%) | 410 (15.0) | 39 (26.4) | 55 (18.0) | 77 (31.2) | 34 (21.3) |
| physical activity2 (%) | 2220 (80.9) | 61 (40.7) | 223 (72.2) | 169 (68.2) | 107 (66.1) |
| Health and disease | |||||
| poor self-reported health (%) | 312 (11.4) | 72 (48.0) | 85 (27.5) | 116 (47.0) | 53 (32.7) |
| multimorbidity3 (%) | 183 (6.7) | 32 (21.3) | 56 (18.2) | 38 (15.3) | 24 (14.8) |
Legend: N (%) or mean (SD);
BMI=body mass index: underweight (< 20 kg/m2), normal weight (20-24.9 kg/m2), overweight (25-29.9 kg/m2), and obesity (≥ 30 kg/m2);
physical activity: meet Dutch physical activity guideline (30 minutes of moderate to vigorous physical activity per day on at least 5 days per week);
multimorbidity: having two or more out of diabetes, cancer, myocardial infarction, cerebrovascular accident, and chronic respiratory symptoms; Note: Summing up the total numbers of participants for the different frailty domains (n=150 for physical, n=310 for cognitive, n=248 for psychological, n=162 for social) leads to an overestimation of the frail population, because participants can be frail for one, or more domains. In total, 703 participants are frail for one or more domains.
BMI and frailty
Unadjusted proportions showed a U-shaped association between BMI and physical and psychological frailty (Figure 1). A linear association was found between BMI (starting at the BMI class of 20-24.9 kg/m2) and cognitive frailty. There was no association between BMI and social frailty.
Figure 1.
The association between BMI and frailty (unadjusted proportions). Physical frailty (Fried) (A), cognitive frailty (B), psychological frailty (C), and social frailty (D). Note: There were no participants in the lowest BMI class (BMI <20) for cognitive frailty
After adjustment for sex, age, level of education and smoking status, the U-shaped association between BMI and physical frailty remained. Prevalence of physical frailty was 8.2% in underweight participants, 2.9% in normal weight participants, 2.6% in overweight participants and 5.0% in obese participants (Figure 2). After adjustment for the confounders mentioned above, the association between BMI and cognitive frailty attenuated, but was still linear. However, after adjustment no association was observed between BMI and psychological and social frailty.
Figure 2.
The association between BMI and frailty (adjusted proportions). Physical frailty (Fried) (A), cognitive frailty (B), psychological frailty (C), and social frailty (D). Note: Proportions were adjusted for sex, age, level of education, and smoking. There were no participants in the lowest BMI class (BMI <20) for cognitive frailty and there were few participants in the lowest BMI class for psychological frailty
Sensitivity analysis
In order to assess the effect of using different definitions of physical frailty, we performed a sensitivity analyses where we compared the association between BMI and physical frailty defined by Fried (2) with physical frailty defined by Gobbens (12). The prevalence of physical frailty (Fried) was 3.8% (n=150) and the prevalence of physical frailty (Gobbens) was 2.7% (n=108). The sensitivity analysis showed a similar U-shaped association between BMI and physical frailty (Gobbens) for both the unadjusted and adjusted proportions (Figure 3 A and B). The Kappa agreement regarding the two physical frailty instruments was moderate with 42% (with n=56 being physically frail according to both instruments).
Figure 3.
The association between BMI and physical frailty (Gobbens). Unadjusted proportions (A) and adjusted proportions for sex, age, level of education, and smoking (B)
Overlap of frailty domains
A total of 150 (3.8%) participants were physically frail (Fried), 310 (9.2%) were cognitively frail, 248 (6.2%) were psychologically frail, and 162 (4.1%) socially frail. Only one participant was frail for all four domains. Limited overlap was observed between the different frailty domains (Figure 4). The percentage overlap between pairs of two domains ranged from 4.9% to 12.0%. The lowest overlap was observed between the cognitive and the social domain with 4.9% participants meeting the criteria for (at least) these two domains. The highest overlap was observed between the psychological and the social domain with 12.0% participants meeting the criteria for (at least) these two domains. The percentages were obtained by dividing the number of participants who were frail for both these two domains by the sum of individuals within these two domains.
Figure 4.
Venn diagram showing the prevalence and overlap between the different frailty domains
Prevalence per age group
When studying the prevalence in five-year age groups, the prevalence increased with age for physical, cognitive, and social frailty (Figure 5). However, psychological frailty was most common in the lowest age group of 41-44 years old.
Figure 5.
Prevalence in five-year age groups per frailty domain
Discussion
In this study, we found a U-shaped association between BMI and physical frailty, a small linear association between BMI and cognitive frailty and no association between BMI and psychological and social frailty. Further, the different frailty domains (physical, cognitive, psychological, and social) showed only a small proportion of overlap. Finally, the prevalence of physical, cognitive and social frailty increased with age, whereas the prevalence of psychological frailty did not.
Our findings regarding the association between BMI and physical frailty (Fried) are in line with previously published results (20, 21). The studies of Blaum, Hubbard, and our study, all showed that there is a U-shaped association between BMI and physical frailty based on the Frailty Phenotype. Physical frailty as defined by Gobbens seems to identify a population that is in general older and less healthy than the physically frail population as defined by Fried (Supplementary Table 1). Nevertheless, the association between BMI and physical frailty (Gobbens) showed a similar U-shaped curve. Thus, the association between BMI and physical frailty as defined by Fried is similar in different populations. Further, the association between BMI and physical frailty is similar when using different instruments for defining physical frailty. It is not yet clear why both underweight and obesity are associated with physical frailty. Possibly these individuals have a shared characteristic which could be the signal or outcome of a similar (underlying) biological mechanism, resulting in a U-shaped association between BMI and physical frailty.
Our results regarding the associations between BMI and psychological and social frailty are in line with the results reported by Gobbens et al (12). They showed that there is a statistically significant association between BMI and physical frailty and no statistically significant association between BMI and psychological and social frailty. The association between BMI and cognitive frailty has, to the best of our knowledge, not yet been studied.
Our study has several strengths and limitations. We studied frailty in the Doetinchem Cohort Study, a large populationbased study with a high response rate, which gave us amongst others a better understanding of the age-distribution of the different frailty domains. The fact that we studied the association between BMI and four separate frailty domains is in our view a strength, because it gives us more insight in the differences between the frailty domains. On the one hand, the operationalization of these domains could be considered a limitation, because they are based on validated instruments, but they are not an exact copy of these instruments. On the other hand, for some domains such as the cognitive frailty domain we used all available data to define frailty (3 cognitive tests) instead of a single question, which was used in the original frailty instrument (TFI). For physical frailty, the criteria from the Frailty Phenotype were used and for psychological, and social frailty the criteria from the TFI were used. In contrast to the TFI, we considered psychological and cognitive frailty as two separate domains because psychological processes differ considerably from cognitive processes. In the literature on frailty, cognitive frailty is also increasingly being recognized as a separate domain (15, 16). For cognitive frailty, we used wellknown and validated cognitive functioning tests. Unfortunately, we did not have data on (self)-reported dementia, which is recommend to take into consideration according to the definition for cognitive frailty that is currently being developed (15, 16). However, due to selection bias (participants with declining cognitive function tend to refuse to participate in the cognitive functioning tests) we assume that the available cognitive data are from participants without dementia.
The associations between BMI and the different frailty domains were studied cross-sectionally, prohibiting causal inference. In an additional analysis, we also studied the association between BMI measured in round 2 (15 years earlier) and frailty in round 5. The associations between BMI (round 2) and the different frailty domains (round 5) are similar to the association we found in the cross-sectional analyses (see Supplementary Figure 1 and Figure 2). In these additional analyses, the number of cases in the lowest BMI class is limited and interpretation of the results for this BMI class is difficult.
A different limitation is the BMI cutoff points used in this study. Due to the age range of our population, we could not use BMI cut-off points that are specific for elderly individuals. Currently, it is suggested that a BMI < 23 would be a suitable cut-off point for underweight in elderly (33). In addition, the cut-off point regarding obesity in elderly is also under discussion and it is suggested that the value should be higher than a BMI of 30. However, the age-range of the participants in the Doetinchem Cohort Study is between 41 and 81 years old. Therefore, we decided to only adjust the cut-off point for underweight from <18.5 to <20, which is in line with other studies (21, 34) and to keep the other cut-off points according to the WHO recommendations.
To the best of our knowledge, there is only one other paper describing the overlap between different frailty domains. Garre- Olmo (35) described the overlap between the physical, mental, and social frailty domains in men and women aged 75 and over, with the mental frailty domain including both psychological and cognitive measurements. In their study, 1.9% of the individuals were found to be frail for all three domains. In our study, 0.1% of the individuals were frail for the physical, cognitive, psychological, and social domain. This difference could be explained by the fact that our population is much younger, or by the fact that they have three frailty domains while we have four frailty domains. Both the study of Garre-Olmo et al and our study seem to show limited overlap between the different frailty domains, which could have important implications regarding frailty research and prevention. Prevention that focusses on a specific frailty domain could be more beneficial than prevention focusing on overall frailty. For example, prevention for physical frailty will most likely focus on physical activity and nutrition, while prevention for social frailty will be directed towards improving someone's social network. Furthermore, the age at which prevention should start could also differ per frailty domain where prevention for psychological frailty should start at an earlier age than prevention for cognitive frailty. Because the scope of the prevention strategies will differ quite a lot, it will be more efficient to develop a prevention strategy per frailty domain.
The Doetinchem Cohort Study is a unique cohort for studying the development of frailty because of the relative young age of the participants and the ability to define multiple frailty domains due to the wide array of collected variables. Compared to other studies, the prevalence of frailty is lower which could indeed be explained by the relatively young population. Frailty is often studied in populations where the participants are ≥65 yrs. Because of the age distribution of our population, we were able to extend the results of previous studies and show that several domains (physical, psychological, social) of frailty were already present from age 41 onwards. This suggests that frailty may already develop at a relatively young age indicating that identification and prevention should start prior to age 65. In our study, psychological frailty was present at a relatively low age (highest prevalence in age group 41-49) while physical, cognitive, and social frailty were present at a higher age (highest prevalence in age group 70-81). In addition, the prevalence for physical, cognitive, and social frailty seems to increase with age, while the prevalence for psychological frailty does not seem to be age related. Whether the prevalence for physical, cognitive, and social frailty will keep increasing with age in the Doetinchem Cohort Study will be revealed in the coming years.
Conclusion
We found a U-shaped association only between BMI and physical frailty. A small linear association was observed between BMI and cognitive frailty, and there was no association between BMI and psychological and social frailty. The prevalence of physical, cognitive and social frailty increased with age, whereas the prevalence of psychological frailty did not.
We confirm that both underweight and obesity are associated with physical frailty. Obesity also seems to be associated with cognitive frailty. Although we cannot draw any causal inferences from this study, we do think that maintenance of a healthy body weight throughout the life course is important. Furthermore, the limited overlap between the different frailty domains is a first indication that the domains entail distinct populations. Therefore, we suggest to target prevention on multiple frailty domains (e.g. physical, cognitive, psychological, and social frailty) rather than on one domain only. Finally, when taking the age-distribution into account, prevention for frailty should start at a younger age (<65yrs) and prevention for psychological frailty may start even earlier.
Acknowledgement
We would like to thank the epidemiologists and fieldworkers of the Municipal Health Service in Doetinchem for their contribution to the data collection for this study.
Funding
This work was supported by the Ministry of Health, Welfare and Sport of the Netherlands, the National Institute for Public Health and the Environment (grant number S132002) and by Biobanking and Biomolecular Resources Research Infrastructure-NL (grant number CP2011-27).
Contributors
All authors were involved in interpreting the data, drafting and approving the manuscript.
Competing interest
None.
Ethical Standards
Round 5 of the Doetinchem Cohort Study was approved according to the guidelines of the Helsinki Declaration by the Medical Ethics Committee of the University Medical Center Utrecht.
Electronic supplementary material
Supplementary material is available for this article at https://doi.org/10.1007/s12603-016-0854-3 and is accessible for authorized users.
Supplementary material
References
- 1.Rockwood K, Mitnitski A. Frailty in relation to the accumulation of deficits. The journals of gerontology Series A, Biological sciences and medical sciences. 2007;62:722–727. doi: 10.1093/gerona/62.7.722. 10.1093/gerona/62.7.722 PubMed PMID: 17634318. [DOI] [PubMed] [Google Scholar]
- 2.Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. The journals of gerontology;A, Biological sciences and medical sciences. 2001;56:M146–156. doi: 10.1093/gerona/56.3.m146. 10.1093/gerona/56.3.M146 [DOI] [PubMed] [Google Scholar]
- 3.Gobbens RJ, Luijkx KG, Wijnen-Sponselee MT, Schols JM. Toward a conceptual definition of frail community dwelling older people. Nursing outlook. 2010;58:76–86. doi: 10.1016/j.outlook.2009.09.005. 10.1016/j.outlook.2009.09.005 PubMed PMID: 20362776. [DOI] [PubMed] [Google Scholar]
- 4.Hamerman D. Toward an understanding of frailty. Annals of internal medicine. 1999;130:945–950. doi: 10.7326/0003-4819-130-11-199906010-00022. 10.7326/0003-4819-130-11-199906010-00022 PubMed PMID: 10375351. [DOI] [PubMed] [Google Scholar]
- 5.Rockwood K, Fox RA, Stolee P, Robertson D, Beattie BL. Frailty in elderly people: an evolving concept. CMAJ: Canadian Medical Association journal = journal de l’Association medicale canadienne. 1994;150:489–495. PubMed PMID: 8313261. [PMC free article] [PubMed] [Google Scholar]
- 6.Morley JE, Vellas B, van Kan GA, et al. Frailty consensus: a call to action. Journal of the American Medical Directors Association. 2013;14:392–397. doi: 10.1016/j.jamda.2013.03.022. 10.1016/j.jamda.2013.03.022 PubMed PMID: 23764209, PMCID 4084863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Walston J, Hadley EC, Ferrucci L, et al. Research agenda for frailty in older adults: toward a better understanding of physiology and etiology: summary from the American Geriatrics Society/National Institute on Aging Research Conference on Frailty in Older Adults. Journal of the American Geriatrics Society. 2006;54:991–1001. doi: 10.1111/j.1532-5415.2006.00745.x. 10.1111/j.1532-5415.2006.00745.x PubMed PMID: 16776798. [DOI] [PubMed] [Google Scholar]
- 8.Rockwood K, Stadnyk K, MacKnight C M, Dowell I, Hebert R, Hogan DB. A brief clinical instrument to classify frailty in elderly people. Lancet (London, England) 1999;353:205–206. doi: 10.1016/S0140-6736(98)04402-X. 10.1016/S0140-6736(98)04402-X [DOI] [PubMed] [Google Scholar]
- 9.Speechley M, Tinetti M. Falls and injuries in frail and vigorous community elderly persons. Journal of the American Geriatrics Society. 1991;39:46–52. doi: 10.1111/j.1532-5415.1991.tb05905.x. 10.1111/j.1532-5415.1991.tb05905.x PubMed PMID: 1987256. [DOI] [PubMed] [Google Scholar]
- 10.Winograd CH, Gerety MB, Chung M, Goldstein MK, Dominguez F Jr, Vallone R. Screening for frailty: criteria and predictors of outcomes. Journal of the American Geriatrics Society. 1991;39:778–784. doi: 10.1111/j.1532-5415.1991.tb02700.x. 10.1111/j.1532-5415.1991.tb02700.x PubMed PMID: 1906492. [DOI] [PubMed] [Google Scholar]
- 11.Rockwood K, Song X, MacKnight C, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ: Canadian Medical Association journal = journal de l’Association medicale canadienne. 2005;173:489–495. doi: 10.1503/cmaj.050051. 10.1503/cmaj.050051 PubMed PMID: 16129869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gobbens RJ, van Assen MA, Luijkx KG, Wijnen-Sponselee MT, Schols JM. The Tilburg Frailty Indicator: psychometric properties. Journal of the American Medical Directors Association. 2010;11:344–355. doi: 10.1016/j.jamda.2009.11.003. 10.1016/j.jamda.2009.11.003 PubMed PMID: 20511102. [DOI] [PubMed] [Google Scholar]
- 13.Rolfson DB, Majumdar SR, Tsuyuki RT, Tahir A, Rockwood K. Validity and reliability of the Edmonton Frail Scale. Age and ageing. 2006;35:526–529. doi: 10.1093/ageing/afl041. 10.1093/ageing/afl041 PubMed PMID: 16757522. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Schuurmans H, Steverink N, Lindenberg S, Frieswijk N, Slaets JP. Old or frail: what tells us more? The journals of gerontology Series A, Biological sciences and medical sciences. 2004;59:M962–965. doi: 10.1093/gerona/59.9.m962. 10.1093/gerona/59.9.M962 PubMed PMID: 15472162. [DOI] [PubMed] [Google Scholar]
- 15.Dartigues JF, Amieva H. Cognitive frailty: rational and definition from an (I.a.N.a./ i.a.g.g.) international consensus group. The journal of nutrition, health & aging. 2014;18:95. doi: 10.1007/s12603-013-0437-5. 10.1007/s12603-013-0437-5 [DOI] [PubMed] [Google Scholar]
- 16.Kelaiditi E, Cesari M, Canevelli M, et al. Cognitive frailty: rational and definition from an (I.A.N.A./I.A.G.G.) international consensus group. The journal of nutrition, health & aging. 2013;17:726–734. doi: 10.1007/s12603-013-0367-2. 10.1007/s12603-013-0367-2 [DOI] [PubMed] [Google Scholar]
- 17.Panza F D, Introno A, Colacicco AM, et al. Current epidemiology of mild cognitive impairment and other predementia syndromes. The American journal of geriatric psychiatry: official journal of the American Association for Geriatric Psychiatry. 2005;13:633–644. doi: 10.1176/appi.ajgp.13.8.633. 10.1097/00019442-200508000-00002 [DOI] [PubMed] [Google Scholar]
- 18.Woods AJ, Cohen RA, Pahor M. Cognitive frailty: frontiers and challenges. The journal of nutrition, health & aging. 2013;17:741–743. doi: 10.1007/s12603-013-0398-8. 10.1007/s12603-013-0398-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Fakhouri TH, Ogden CL, Carroll MD, Kit BK, Flegal KM, Prevalence of obesity among older adults in the United States, 2007-2010, NCHS data brief, 2012, 1–8 [PubMed]
- 20.Blaum CS, Xue QL, Michelon E, Semba RD, Fried LP. The association between obesity and the frailty syndrome in older women: the Women’s Health and Aging Studies. Journal of the American Geriatrics Society. 2005;53:927–934. doi: 10.1111/j.1532-5415.2005.53300.x. 10.1111/j.1532-5415.2005.53300.x PubMed PMID: 15935013. [DOI] [PubMed] [Google Scholar]
- 21.Hubbard RE, Lang IA, Llewellyn DJ, Rockwood K. Frailty, body mass index, and abdominal obesity in older people. The journals of gerontology Series A, Biological sciences and medical sciences. 2010;65:377–381. doi: 10.1093/gerona/glp186. 10.1093/gerona/glp186 PubMed PMID: 19942592. [DOI] [PubMed] [Google Scholar]
- 22.Puts MT, Visser M, Twisk JW, Deeg DJ, Lips P. Endocrine and inflammatory markers as predictors of frailty. Clinical endocrinology. 2005;63:403–411. doi: 10.1111/j.1365-2265.2005.02355.x. 10.1111/j.1365-2265.2005.02355.x PubMed PMID: 16181232. [DOI] [PubMed] [Google Scholar]
- 23.Verschuren WM, Blokstra A, Picavet HS, Smit HA. Cohort profile: the Doetinchem Cohort Study. International journal of epidemiology. 2008;37:1236–1241. doi: 10.1093/ije/dym292. 10.1093/ije/dym292 PubMed PMID: 18238821. [DOI] [PubMed] [Google Scholar]
- 24.Kemper HCG, Ooijendijk WTM, Stiggelbout M. Consensus over de Nederlandse norm voor gezond bewegen [Consensus on the Dutch standard for healthy exercise] Tijdschr Soc Gezondheidsz. 2000;78:180–183. [Google Scholar]
- 25.Gobbens RJ, Luijkx KG, Wijnen-Sponselee MT, Schols JM. Towards an integral conceptual model of frailty. The journal of nutrition, health & aging. 2010;14:175–181. doi: 10.1007/s12603-010-0045-6. 10.1007/s12603-010-0045-6 [DOI] [PubMed] [Google Scholar]
- 26.Radloff LS. The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement. 1977;1:385–401. 10.1177/014662167700100306 [Google Scholar]
- 27.van der Zee KI, Sanderman R. Het Meten van Gezondheidstoestand Met de RAND-36: een Handleiding. Noordelijk Centrum voor Gezondheidsvraagstukken; Groningen, The Netherlands: 1993. [Google Scholar]
- 28.de Jong Gierveld J, Kamphuis FH. The development of a Rasch-type loneliness-scale. Appl Psychol Meas. 1985;9:289–299. 10.1177/014662168500900307 [Google Scholar]
- 29.Kempen GIJM, Van Eijk LM. The psychometric properties of the SSL12-I, a short scale for measuring social support in the elderly. Soc Indic Res. 1995;35:303–312. 10.1007/BF01079163 [Google Scholar]
- 30.Dutch Municipal Health Services Elderly Monitor [Monitor Gezondheid Ouderen].
- 33.Winter JE, MacInnis RJ, Wattanapenpaiboon N, Nowson CA. BMI and all-cause mortality in older adults: a meta-analysis. The American journal of clinical nutrition. 2014;99:875–890. doi: 10.3945/ajcn.113.068122. 10.3945/ajcn.113.068122 PubMed PMID: 24452240. [DOI] [PubMed] [Google Scholar]
- 34.Sergi G, Perissinotto E, Pisent C, et al. An adequate threshold for body mass index to detect underweight condition in elderly persons: the Italian Longitudinal Study on Aging (ILSA) The journals of gerontology Series A, Biological sciences and medical sciences. 2005;60:866–871. doi: 10.1093/gerona/60.7.866. 10.1093/gerona/60.7.866 PubMed PMID: 16079209. [DOI] [PubMed] [Google Scholar]
- 35.Garre-Olmo J, Calvo-Perxas L, Lopez-Pousa S, de Gracia Blanco M, Vilalta-Franch J. Prevalence of frailty phenotypes and risk of mortality in a community-dwelling elderly cohort. Age and ageing. 2013;42:46–51. doi: 10.1093/ageing/afs047. 10.1093/ageing/afs047 PubMed PMID: 22454134. [DOI] [PubMed] [Google Scholar]
Uncited references
- 31.Hubbard RE, Searle SD, Mitnitski A, Rockwood K. Effect of smoking on the accumulation of deficits, frailty and survival in older adults: a secondary analysis from the Canadian Study of Health and Aging. The journal of nutrition, health & aging. 2009;13:468–472. doi: 10.1007/s12603-009-0085-y. 10.1007/s12603-009-0085-y [DOI] [PubMed] [Google Scholar]
- 32.Santos-Eggimann B, Cuenoud P, Spagnoli J, Junod J. Prevalence of frailty in middleaged and older community-dwelling Europeans living in 10 countries. The journals of gerontology Series A, Biological sciences and medical sciences. 2009;64:675–681. doi: 10.1093/gerona/glp012. 10.1093/gerona/glp012 PubMed PMID: 19276189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Medical care. 1992;30:473–483. 10.1097/00005650-199206000-00002 PubMed PMID: 1593914. [PubMed] [Google Scholar]
- 37.van der Elst W, van Boxtel MP, van Breukelen GJ, Jolles J. Rey’s verbal learning test: normative data for 1855 healthy participants aged 24-81 years and the influence of age, sex, education, and mode of presentation. Journal of the International Neuropsychological Society: JINS. 2005;11:290–302. doi: 10.1017/S1355617705050344. PubMed PMID: 15892905. [DOI] [PubMed] [Google Scholar]
- 38.van der Elst W, van Boxtel MP, van Breukelen GJ, Jolles J. The Stroop color-word test: influence of age, sex, and education; and normative data for a large sample across the adult age range. Assessment. 2006;13:62–79. doi: 10.1177/1073191105283427. 10.1177/1073191105283427 PubMed PMID: 16443719. [DOI] [PubMed] [Google Scholar]
- 39.van der Elst W, van Boxtel MP, van Breukelen GJ, Jolles J. The Concept Shifting Test: adult normative data. Psychological assessment. 2006;18:424–432. doi: 10.1037/1040-3590.18.4.424. 10.1037/1040-3590.18.4.424 PubMed PMID: 17154763. [DOI] [PubMed] [Google Scholar]
- 40.van der Elst W, van Boxtel MP, van Breukelen GJ, Jolles J. The Letter Digit Substitution Test: normative data for 1,858 healthy participants aged 24-81 from the Maastricht Aging Study (MAAS): influence of age, education, and sex. Journal of clinical and experimental neuropsychology. 2006;28:998–1009. doi: 10.1080/13803390591004428. 10.1080/13803390591004428 PubMed PMID: 16822738. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary material