SEARCH

SEARCH BY CITATION

Keywords:

  • BMI;
  • geriatric;
  • function;
  • homebound;
  • screening

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Objective: To test the a priori hypothesis that obesity is a predictor of risk for reporting homebound status.

Research Methods and Procedures: A longitudinal cohort study was conducted with 21, 645 community-dwelling men and women 65 to 97 years old. A nutrition risk screen was administered baseline between 1994 and 1999 and again 3 to 4 years later. Univariate analyses identified baseline variables associated with subsequent reporting of homebound status. Multivariable logistic regression models were created to identify baseline variables that were significant independent predictors of reporting homebound status.

Results: At baseline, 24% of the cohort had BMI ≥ 30. There were 12, 834 (45% men) respondents at follow-up (68% response). Non-responders at follow-up differed little from responders except for greater baseline age (72.2 ± 6.2 vs. 71.4 ± 5.6 years, p < 0.001) and reporting of any functional limitations (9.2% vs. 4.9%, p < 0.001). At follow-up, those who reported homebound status (n = 169) were significantly (p < 0.001) older (80.3 ± 7.3 vs. 75.1 ± 5.5 years) and more likely to report functional limitations (83.4% vs. 10.8%). Univariate analyses identified 16 baseline variables that were eliminated stepwise until five significant independent predictors remained: age ≥ 75 years (2.21, 1.55 to 3.15/odds ratio, 95% confidence interval), BMI ≥ 35 (1.75, 1.04 to 2.96), poor appetite (2.50, 1.29 to 4.86), low income (1.59, 1.00 to 2.56), and any functional limitation (10.67, 7.36 to 15.46).

Discussion: Obesity remained a significant independent predictor for reporting homebound status and should be considered in screening of older populations and in the planning, implementation, and evaluation of services for homebound older persons.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

Obesity is increasing among older Americans (1, 2, 3) and is associated with serious medical comorbidities that include hypertension, type 2 diabetes, dyslipidemia, metabolic syndrome, coronary artery disease, and osteoarthritis of the knees (4, 5, 6, 7, 8, 9, 10). Annual medical costs for an obese Medicare recipient are $1486 higher than for one of desirable weight (11). Among older persons, high BMI is also associated with increased self-reported functional limitations, decreased measured physical performance, and elevated risk of functional decline (12, 13, 14, 15, 16, 17, 18). Recent studies of homebound older persons have observed a one-third or greater prevalence of obesity (17, 18, 19, 20, 21). Previously identified risk factors for homebound status include age, gender, education, financial resources, transportation availability, social support, environmental hazards, disease burden, and functional limitations. We, therefore, anticipated that obesity might be useful as a practical screening indicator of risk for homebound status. This longitudinal cohort investigation was designed to test the hypothesis that obesity is now a significant independent predictor for risk for reporting homebound status.

Research Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

This investigation was approved by the Geisinger Health System Institutional Research Review Board with a data- sharing agreement with the Vanderbilt Center for Human Nutrition. The Geisinger service region has one of the largest concentrations of rural older persons in the U.S. (22). Many persons reside in small unincorporated towns with populations of 2499 or fewer, and only 1% have access to a hospital in their community (23). The Geisinger Rural Aging Study (GRAS)1 is a cohort of 21, 645 community-dwelling Pennsylvanians ≥65 years old who completed a baseline nutrition risk screening at one of the over 100 Geisinger clinic sites between 1994 and 1999 (24). Participants were enrollees in a regional managed healthcare plan for Medicare beneficiaries. The cohort is 99% non-Hispanic white (self-reported with options defined by investigator). Educational level is high school or higher for 80% of the cohort. The GRAS has been characterized by large-scale mailing of questionnaires (14, 25), telephone interviews (26, 27, 28), and random subsampling targeting smaller representative groups for home visits or clinic-based encounters (28, 29, 30, 31, 32). The overall aim of the GRAS is to relate longitudinal health and functional outcomes to baseline nutritional risk status.

A modified Nutrition Screening Initiative Level II (LII) questionnaire (33, 34) was included in the Health Plan enrollment packet and was completed at a required medical history/physical examination. This LII has been extensively tested and described in previous reports (14, 24, 25, 26, 27, 28, 30, 32). The questionnaire includes self-reported height and weight, weight change over past 6 months (volitional or not), living and eating habits, alcohol and medication use, depression, dentition, functional limitations, and housebound status. Members’ responses were reviewed with the clinic nurse. Assistance was provided as necessary in completion of the screening questionnaire, and proxies were used if necessary. Typical completion times are 5 to 10 minutes.

Serum albumin and cholesterol levels were obtained from routine surveillance measurements taken at the discretion of primary physicians within 12 months before and 2 weeks after completion of the baseline LII. All laboratory analyses were completed at the accredited laboratory facilities of the Geisinger Healthcare System using a Hitachi 717 analyzer (Hitachi, Indianapolis, IN) and Boehringer Mannheim (Indianapolis, IN) reagents and manufacturer-provided guidelines. The Boehringer Mannheim reagent system uses an enzymatic cholesterol method (cholesterol/HP; catalog no. 1039033) and a bromcresol green dye-binding albumin method (albumin BCG; catalog no. 1039025).

Functional status was assessed with the questions usually or always need assistance with (check all that apply): bathing, dressing, grooming, toileting, eating, walking or moving about, traveling (outside the home), preparing food, and shopping for food or other necessities. The number of checkmarks at baseline and rescreening were tallied. Limitations were considered in three groups: any activities of daily living (ADL; bathing, dressing, grooming, toileting, and eating) (35), any instrumental ADL (IADL; walking, traveling, shopping, and preparing food) (36), and as any functional limitation (any ADL or IADL).

Subjects who completed baseline screening were rescreened in rolling fashion with mailed questionnaires at 3- to 4-year intervals. The mailing included instructions and a self-addressed postage-paid return envelope. Non-responders received two subsequent mailings over 2 months. If the subject was unavailable, a proxy was queried in regard to disposition, including mortality or skilled nursing facility (SNF) admission. Individuals were excluded from analysis if they were homebound at baseline or were found to be expired or in an SNF at follow-up. A clinic encounter was not part of the follow-up assessment, so blood tests were not available for analyses.

Data Analysis

The dependent variable, homebound status, was dichotomized as those individuals who checked being housebound at follow-up screening vs. those who did not. Calculations were performed using SPSS for Windows (version 12.0; SPSS, Inc., Chicago, IL). Descriptive statistics were conducted for the items from the LII screen. χ2 Statistics were used for categorical variables and Student's t statistics for continuous variables to compare baseline differences between responders and non-responders at follow-up and differences between responders who reported homebound status at follow-up compared with those who did not. BMI was calculated from self-reported weight and height as BMI = weight (kilograms)/height (meters) squared. Spearman's correlations coefficients were obtained to identify significant bivariate associations (p < 0.05) between reporting being homebound at follow-up and each LII item at baseline. Unadjusted odds ratios (ORs) and confidence intervals were calculated for the subset of 31 significant correlates using logistic regression. Cut-off points for age, BMI (37), albumin (38), and cholesterol (33, 34) were selected on the basis of established standards and findings from the previous GRAS investigations. For the multivariate analyses, the 31 correlates were reduced to 27 after examination for multicollinearity using Spearman's correlation coefficients, interactions, the presence of zero or low cell counts using three-way contingency tables, outliers, and clinical significance. Additional reduction from 27 to 16 correlates occurred with selecting any ADL or IADL to represent any functional limitations.

Multivariable logistic regression models were performed on the remaining 16 variables to examine the association with reporting being homebound. In the first model, the 16 variables were simultaneously entered. Variables that made little or no contribution to the model fit were eliminated using a backward stepwise regression program. Only items that were significant at the 5% level were retained in the final model. Pair-wise interaction terms were entered into both models to look for effect modifiers. The Hosmer and Lemeshow goodness of fit test was performed, and sensitivity and specificity of the two models were compared. Adjusted ORs and confidence intervals were calculated.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

There were 21, 645 rural community-dwelling Pennsylvanians ≥65 years old who completed the baseline LII screen. Individuals were excluded from follow-up analysis who were homebound at baseline (n = 143) or at the time of follow-up were in a SNF (n = 269) or expired (n = 2282). Sixty-eight percent of the eligible sample for rescreening (n = 18, 951) returned a completed survey (n = 12, 834). Baseline variables of non-responders at follow-up did not differ appreciably from those of respondents with the exception of modestly greater age (72.2 ± 6.2 vs. 71.4 ± 5.6 years, p < 0.05) and reported functional limitations (9.2% vs. 4.9% reported at least one ADL or IADL, p < 0.001). Mean BMI did not differ significantly between responders and non-responders, but there were slightly more severely obese individuals among the non-responders (BMI ≥ 35, 8.4% vs. 7.4%, p = 0.022).

Table 1 presents the baseline variables for those reporting homebound status (n = 169) at follow-up vs. not (n = 12, 665). The prevalence of overweight (BMI ≥ 25.0) and obesity (BMI ≥ 30) at baseline among those reporting non-homebound status at follow-up was 69.5% and 26.1%, respectively. The prevalence of overweight and obesity at baseline among those reporting homebound status at follow-up was 66.9% and 28.4%, respectively. At baseline, those reporting homebound status at follow-up had significantly (p < 0.05) greater age and prevalence of low serum albumin. They also had significantly (p < 0.05) greater prevalence of reporting depression, polypharmacy, weight loss, weight gain, eating alone, poor appetite, special diet, subthreshold intake of breads, cereals, pasta, rice, or other grains, difficulty chewing/swallowing, low income, living alone, home security concerns, and ADL/IADL limitations.

Table 1. . Baseline and follow-up LII nutrition risk screening items by subjects reporting homebound status at follow-up (n = 12, 834)
 BaselineFollow-up
ItemNot homebound at follow-up (N = 12, 665)Homebound at follow-up (N = 169)Not homebound at follow-up (N = 12, 665)Homebound at follow-up (N = 169)
  • LII, Level II; SD, standard deviation; ADL, activity(ies) of daily living; IADL, instrumental ADL.

  • *

    Significant difference between those who were not homebound and those who were homebound (p < 0.05).

  • Significant difference between baseline and follow-up (p < 0.05).

  • BMI based on 11, 656 adults at baseline and follow-up: 11, 508 non-homebound and 148 homebound.

  • §

    Albumin collected from 2904 adults at baseline: 2859 non-homebound and 45 homebound.

  • Cholesterol collected from 4395 adults at baseline: 4343 non-homebound and 52 homebound.

Male, n (%)5662 (44.7)68 (40.7)5662 (44.7)68 (40.7)
Age (years), mean (SD)71.4 (5.6)76.6 (7.3)*75.1 (5.5)80.3 (7.3)*,
Age range (years)65 to 9765 to 9769 to 10169 to 101
Age ≥ 75, n (%)3295 (26.0)94 (55.6)*5378 (42.5)126 (74.6)*,
BMI (kg/m2), mean (SD)27.6 (4.9)27.8 (5.7)27.6 (4.9)27.6 (6.6)
BMI < 18.5, n (%)116 (1.0)1 (0.7)176 (1.5)4 (2.6)
BMI 18.5 to 24.9, n (%)3377 (29.4)48 (32.4)3552 (29.3)56 (36.1)
BMI 25.0 to 29.9, n (%)4995 (43.4)57 (38.5)5187 (42.8)45 (29.0)*
BMI 30.0 to 34.9, n (%)2168 (18.9)24 (16.2)2279 (18.8)33 (21.3)
BMI ≥ 35, n (%)843 (7.2)18 (12.2)*931 (7.7)17 (11.0)
Albumin, mean (SD)§4.3 (0.3)4.2 (0.4)N/AN/A
Albumin ≤ 3.8, n (%)163 (5.7)6 (13.3)*N/AN/A
Cholesterol, mean (SD)215.6 (41.1)222.6 (42.4)N/AN/A
Cholesterol ≤ 160, n (%)306 (7.0)3 (5.8)N/AN/A
Have lost 10 pounds or more in the past 6 months, n (%)923 (7.3)21 (12.4)*1447 (11.4)42 (24.9)*,
Have gained 10 pounds or more in the past 6 months, n (%)943 (7.4)22 (13.0)*1035 (8.2)16 (9.5)
Feel depressed, n (%)371 (2.9)13 (7.7)*927 (7.3)52 (30.8)*,
Use three or more prescription drugs daily, n (%)5921 (46.8)98 (58.0)*8511 (67.2)133 (78.7)*,
Do not have enough food to eat each day, n (%)826 (6.5)10 (5.9)1217 (9.6)20 (11.8)
Usually eat alone, n (%)2308 (18.2)56 (33.1)*2903 (22.9)58 (34.3)*,
Do not eat anything on 1 or more days per month, n (%)75 (0.6)3 (1.8)120 (0.9)3 (1.8)
Have a poor appetite, n (%)198 (1.6)14 (8.3)*451 (3.6)39 (23.1)*,
Am on a special diet, n (%)914 (7.2)20 (11.8)*909 (7.2)35 (20.7)*,
Eat vegetable two or fewer times daily, n (%)7131 (56.3)108 (63.9)7659 (60.5)112 (66.3)
Eat milk or milk products once or not at all daily, n (%)7086 (55.9)105 (62.1)8045 (63.5)117 (68.2)
Eat fruit or drink fruit juice once or not at all daily, n (%)7215 (56.9)101 (59.8)8020 (63.3)120 (71.0)*,
Eat breads, cereals, pasta, rice, or other grains five or fewer times daily, n (%)8196 (64.7)129 (76.3)*8761 (69.2)119 (70.4)
Have difficulty chewing or swallowing, n (%)348 (2.7)16 (9.5)*480 (3.8)41 (24.3)*,
Have more than one alcoholic drink per day, n (%)442 (3.5)5 (3.0)477 (3.8)7 (4.1)
Have pain in mouth, teeth, or gums, n (%)234 (1.8)4 (2.4)271 (2.1)10 (5.9)*
Live on an income of less than $6000 per year, n (%)863 (6.8)27 (16.0)*963 (7.6)29 (17.2)*
Live alone, n (%)2629 (20.7)49 (29.0)*3312 (26.2)58 (34.3)*
Am concerned about home security, n (%)171 (1.3)7 (4.1)*296 (2.3)18 (10.7)*,
Live in a home with inadequate heating or cooling, n (%)437 (3.4)6 (3.6)736 (5.8)22 (13.0)*,
Do not have a stove and/or a refrigerator, n (%)125 (1.0)2 (1.2)184 (1.5)7 (4.1)*
Am unable or prefer not to spend money on food (<$25 to $30 per person per week), n (%)207 (1.6)6 (3.6)392 (3.1)14 (8.3)*
Usually or always need assistance with    
 Bathing, n (%)83 (0.7)22 (13.0)*320 (2.5)79 (46.7)*,
 Dressing, n (%)63 (0.5)14 (8.3)*203 (1.6)56 (33.1)*,
 Grooming, n (%)37 (0.3)10 (5.9)*117 (0.9)46 (27.2)*,
 Toileting, n (%)27 (0.2)7 (4.1)*79 (0.6)32 (18.9) *,
 Eating, n (%)20 (0.2)3 (1.8)*44 (0.3)22 (13.0) *,
 Walking or moving about, n (%)137 (1.1)23 (13.6)*408 (3.2)72 (42.6) *,
 Traveling, n (%)332 (2.6)58 (34.3)*826 (6.5)116 (68.5) *,
 Preparing food, n (%)133 (1.0)25 (14.8)*379 (3.0)82 (48.5) *,
 Shopping for food or other, n (%)310 (2.4)47 (27.8)*840 (6.6)110 (65.1) *,
 Any ADL, n (%)200 (1.6)36 (21.3)*618 (4.9)99 (58.6) *,
 Any IADL, n (%)484 (3.8)70 (41.4)*1191 (9.4)134 (79.3) *,
 Any ADL or IADL, n (%)560 (4.4)73 (43.2)*1366 (10.8)141 (83.4) *,

For the purpose of comparison, descriptive data in Table 1 include the follow-up variables for those reporting homebound status vs. not. Mean BMI and BMI distribution changed little, whereas multiple other variables increased in prevalence significantly over the 3- to 4-year follow-up period. Those reporting homebound status at follow-up in comparison with those who did not had significantly (p < 0.05) greater age and prevalence of depression, polypharmacy, weight loss, eating alone, poor appetite, special diet, subthreshold intakes of fruit or fruit juice, difficulty chewing/swallowing, pain in mouth, teeth, or gums, low income, living alone, home security concerns, inadequate heating or cooling, no stove and/or refrigerator, unable or prefer not to spend money on food, and ADL/IADL limitations.

Presented in Table 2 are the unadjusted ORs and 95% confidence intervals for baseline variables associated with reporting homebound status at follow-up. Thirty-one potential predictor variables were identified including age, BMI, weight loss, weight gain, appetite, depression, polypharmacy, eating alone, difficulty chewing and swallowing, and all functional status items. These 31 variables were reduced to 16 in further univariate analysis. Eleven of the 12 functional status items were eliminated with any ADL or IADL retained as one representative functional status item. The food frequency items (bread and vegetable) were eliminated because they have not withstood reliability testing (24). The item eat alone was eliminated because it strongly correlated with live alone, and it is more likely that eating alone is a consequence of living alone. The variable albumin was eliminated because it produced complete separation between housebound and non-housebound subjects, which could result in numerical problems at the multivariate stage. Additionally, obtaining serum albumin levels is not practical for large-scale population screening.

Table 2. . Univariate analysis of the risks of reporting homebound status associated with baseline LII screening items
VariableOdds ratio95% confidence interval
  1. LII, Level II; ADL, activity(ies) of daily living; IADL, instrumental ADL.

Is a male0.850.62 to 1.16
Age ≥ 753.562.62 to 4.84
BMI ≥ 351.751.07 to 2.88
Albumin ≤ 3.8 g/dL2.551.06 to 6.10
Lost 10 pounds or more in past 6 months1.811.14 to 2.87
Gained 10 pounds or more in past 6 months0.540.34 to 0.84
Feel depressed2.761.56 to 4.91
Use three or more prescription drugs daily0.640.47 to 0.86
Usually eat alone2.231.61 to 3.08
Do not eat anything on 1 or more days each month3.040.95 to 9.73
Have a poor appetite5.693.24 to 10.01
Am on a special diet1.731.08 to 2.77
Eat vegetable two or fewer times daily1.381.00 to 1.89
Eat breads, cereals, pasta, rice, or other grains five or fewer times daily1.761.23 to 2.52
Have difficulty chewing or swallowing3.712.19 to 6.27
Live on an income of less than $6000 annually2.601.72 to 3.95
Live alone1.561.12 to 2.18
Am concerned about home security3.161.46 to 6.84
Am unable to spend money on food2.220.97 to 5.07
Usually or always need assistance with  
 Bathing22.7113.81 to 37.34
 Dressing18.089.92 to 32.96
 Grooming21.4810.50 to 43.95
 Toileting20.248.69 to 47.16
 Eating11.443.37 to 38.86
 Walking or moving about14.429.00 to 23.09
 Traveling19.4313.89 to 27.18
 Preparing food16.3810.36 to 25.88
 Shopping for food or other15.3710.78 to 21.91
 Any ADL16.8911.39 to 25.03
 Any IADL17.8112.94 to 24.51
 Any ADL or IADL16.4512.00 to 22.57

The multivariate model for the 16 remaining predictor variables is presented in Table 3. The full multivariate model had a sensitivity of 71.2% and a specificity of 70.7% for predicting reported homebound status. Backward step-wise regression allowed elimination of 11 variables that made no appreciable improvement to the model fit. The final model, comprised of five variables (Table 4), did not differ in sensitivity or specificity as compared with the full model. The five significant independent predictors of homebound status at follow-up in the final model were ≥75 years old, BMI ≥ 35, poor appetite, low income, and any functional limitation. Models that included interaction terms were also evaluated for these data. These terms did not appreciably improve the model fit to our data and were not used in our final analyses.

Table 3. . Multivariate analysis of the risks of reporting homebound status associated with baseline LII screening items*
VariableOdds ratio95% confidence intervalp
  • LII, Level II; ADL, activity(ies) of daily living; IADL, instrumental ADL.

  • *

    The odds ratios in this table are simultaneously adjusted for all the other variables in the table. Sensitivity, 71.2%; specificity, 70.7%.

Is a male1.100.76 to 1.570.617
Age ≥752.381.65 to 3.430.000
BMI ≥ 351.741.02 to 2.950.042
Has gained 10 or more pounds1.510.88 to 2.570.132
Has lost 10 or more pounds1.170.68 to 2.020.560
Have a poor appetite2.501.17 to 4.710.016
Does not eat anything on 1 or more days monthly1.780.50 to 6.340.373
Is on a special diet1.380.81 to 2.370.239
Has difficulty chewing or swallowing1.220.59 to 2.500.592
Uses three or more prescription drugs daily0.970.68 to 1.390.882
Feels depressed1.070.50 to 2.280.868
Lives on an income of less than $6000 annually1.520.93 to 2.490.092
Lives alone0.880.59 to 1.320.540
Concerned about home security1.450.62 to 3.380.391
Is unable to spend money on food0.960.39 to 2.380.926
Needs assistance with any ADL or IADL10.056.80 to 14.840.000
Constant0.01 0.000
Table 4. . Final multivariate analysis of the risks of reporting homebound status*
VariableOdds ratio95% confidence intervalp
  • ADL, activity (ies) of daily living; IADL, instrumental ADL.

  • *

    The variables included in this analysis all had a significant effect on the risk of becoming homebound. BMI ≥ 35 was associated with a 75% increase in risk of becoming homebound with or without adjustment for potential confounding risk factors. Sensitivity, 71.2%; specificity, 69.9%.

Age ≥ 752.211.55 to 3.150.000
BMI ≥ 351.751.04 to 2.960.036
Have a poor appetite2.501.29 to 4.860.007
Live on an income of less than $6000 annually1.591.00 to 2.560.050
Need assistance with any ADL or IADL10.677.36 to 15.460.000
Constant0.01 0.000

Inclusion of disease burden in analyses was not possible because these data were not available from the cohort at baseline. Because the follow-up survey included questions for six chronic conditions (arthritis, type 2 diabetes, cardiovascular disease, lung disease, cancer, and hypertension), we retested the multivariate models with inclusion of the item heavy disease burden, defined as reporting three or more conditions. Although 37% of the responders reporting homebound status at follow-up also reported heavy disease burden (vs. 23% of those not reporting homebound status, p < 0.05), heavy disease burden did not contribute to the fit of the model (p = 0.319). Thus, assuming that disease burden at follow-up would be similar to or greater than that at baseline, with responders being 3 to 4 years older, the predictors identified would not be any less robust.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

To our knowledge, this is the first longitudinal cohort study of community-dwelling older persons to identify obesity (BMI ≥ 35) as an independent risk factor for reporting homebound status. The odds of reporting subsequent homebound status were 75% greater in obese subjects than in non-obese subjects. Moreover, this OR did not change appreciably in a multivariate analysis that adjusted for age, appetite, low income, or functional limitations. It is likely that this relationship only now becomes apparent with the growing prevalence of obesity among older persons (1, 2, 3). These observations may take on even more importance because of the aging baby boomer population that has a much greater prevalence of obesity in comparison with prior generations of aging adults. It is possible that obesity serves as a proxy indicator for other factors that may be related to risk for homebound status like unrecognized disease or inflammatory processes, mobility restriction, sedentary living, deconditioning, and mental health. It is noteworthy that increased risk for reporting homebound status was identified at BMI ≥ 35 because this is the same threshold at which these authors have identified increased risk for reporting functional limitations (14).

In addition to obesity, this investigation identified four additional significant independent predictors for reporting homebound status: age ≥ 75 years, poor appetite, low income, and any functional limitation. It is noteworthy that reported IADL or ADL limitation remained the most potent risk factor for reporting homebound status. Systematic consideration of interaction terms culminated in this final multiplicative model with sound goodness of fit. The retained variables lend validity to this analysis, as findings by other investigators have suggested risk factors for homebound status that include age, gender, education, financial resources, transportation availability, social support, and functional limitations (39).

Obesity offers practical utility as a screening indicator because it is relatively easy to obtain by simple self-report or observation measures. A highly trained health professional is not required to estimate BMI. It is well established that BMI suffers limitations as a measure of adiposity, but it has value as a screening indicator readily obtained from height and weight (37). At the level of obesity that we have identified as a risk factor, it is not subtle and generally does not require precise measurements or circumferences. In a prior study with the GRAS cohort, it was found that 80% of subjects reported their height to within 2.54 cm and weight to within 2.3 kg (24). Other potential screening risk factors including assessment of medical comorbidity can be much more cumbersome to obtain.

Strengths of this investigation include study of a large well-characterized cohort of aging community-dwelling persons. There was a favorable 68% overall follow-up response rate, and responders appeared to well represent the eligible sample when systematically contrasted with non-responders. The prevalence of obesity is comparable with that described for National Health and Nutrition Examination Study III and other aging cohorts (1, 2, 3). The observed mortality rate (10%) is consistent with the aging subjects that were studied (40). SNF admission rate is likely underestimated because it only represents individuals reported to be in a SNF at follow-up. It is also likely that SNF admission, homebound status, and mortality are over-represented among the non-respondents. Limitations include a lack of regional, ethnic, and racial diversity. Risk for homebound status may be even greater in other ethnic and racial groups like African Americans and Hispanics secondary to higher prevalence of obesity and more limited financial and other resources. Further study is necessary to clarify whether these findings can be generalized to other populations. This study was also limited by inability to distinguish the independent baseline contributions of obesity and associated comorbid conditions, although for screening purposes, obesity has practical utility for risk assessment. This investigation was also dependent on self-reported data; thus, the true prevalence of homebound status may be higher if it could be measured objectively. It is nonetheless evident that reporting homebound status is strongly associated with a burden of poor health and functioning. Those who reported homebound status were much more likely at follow-up to report IADL/ADL limitations.

On the basis of this study, it is not possible to comment whether there is a role for weight reduction among obese older persons who are at risk for becoming homebound. For obese older persons with severe frailty, the emphasis may be better placed on preservation of strength and flexibility (14, 41). The recent joint Technical Review and Position Statement of the American Society for Nutrition and NAASO, the Obesity Society (42), concluded that “the current data show that weight loss therapy improves physical function, quality of life, and the medical complications associated with obesity in older persons. Therefore, weight loss therapy that minimizes muscle and bone loss is recommended for older persons who are obese and who have functional impairments or medical complications that can benefit from weight loss.”

We have previously identified poor diet quality and micronutrient deficiencies as prevalent concerns among homebound persons who are obese (28, 32). Participants in the Older Americans Act Nutrition Program (43) who receive home-delivered meals have been found to have higher nutrient intakes than non-participants (44). Thus, it is imperative that quality food and nutrition services be included in comprehensive home and community-based systems (45). Diet quality and nutrient density should continue to be emphasized, although individualized assessment should take into consideration the weight status of individuals when planning meal, snack, supplement, and/or nutrition support interventions.

In conclusion, obesity remained a significant independent predictor of risk for reporting homebound status even after controlling for variables like age, appetite, low income, or functional limitations. It will be important to consider obesity in screening of older populations and in the planning, implementation, and evaluation of services for homebound older persons. It is likely that becoming homebound is often a hidden outcome of obesity because many of these individuals are effectively isolated and no longer able to routinely access the healthcare system because of mobility/transportation limitations.

Acknowledgement

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References

This work was funded, in part, by the U.S. Department of Agriculture, Agricultural Research Service, under agreement 58-1950-1-137. The assistance of Geisinger Nutrition Center personnel in data acquisition is most appreciated.

Footnotes
  • 1

    Nonstandard abbreviations: GRAS, Geisinger Rural Aging Study; LII, Level II; ADL, activity(ies) of daily living; IADL, instrumental ADL; SNF, skilled nursing facility; OR, odds ratio.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Research Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References
  • 1
    Flegal, K. M., Carroll, M. D., Ogden, C. L., Johnson, CL. (2002) Prevalence and trends in obesity among US adults, 1999–2000. JAMA 288: 17231727.
  • 2
    Arterburn, D. E., Crane, P. K., Sullivan, SD. (2004) The coming epidemic of obesity in elderly Americans. J Am Geriatr Soc. 52: 19071912.
  • 3
    Jensen, GL. (2005) Obesity and functional decline: epidemiology and geriatric consequences. Clin Geriatr Med. 21: 677687.
  • 4
    Huang, Z., Willet, W. C., Manson, J. E., et al (1998) Body weight, weight change, and risk for hypertension in women. Ann of Int Med. 128: 8188.
  • 5
    Stevens, J., Cai, J., Pamuk, E. R., Williamson, D. F., Thun, M. J., Wood, JL. (1998) The effect of age on the association between body-mass index and mortality. N Eng J Med. 338: 17.
  • 6
    Hochberg, M. C., Lethbridge-Cejku, M., Scott, W. W., Jr, Reichle, R., Plato, C. C., Tobin, J. D. (1995) The association of body weight, body fatness and body fat distribution with osteoarthritis of the knee: data from the Baltimore Longitudinal Study of Aging. J Rheum. 22: 488493.
  • 7
    Jordan, J. M., Luta, G., Renner, J. B., et al (1996) Self-reported functional status in osteoarthritis of the knee in a rural southern community: the role of sociodemographic factors, obesity, and knee pain. Arthritis Care Res. 9: 272278.
  • 8
    Colditz, G. A., Willet, W. C., Rotnitzky, A., Manson, JE. (1995) Weight gain as a risk factor for clinical diabetes mellitus in women. Ann Int Med. 122: 481486.
  • 9
    Holvoet, P., Kritchevsky, S. B., Tracy, R. P., et al (2004) The metabolic syndrome, circulating oxidized LDL, and risk of myocardial infarction in well-functioning elderly people in the health, aging, and body composition cohort. Diabetes 53: 10681073.
  • 10
    Zamboni, M., Zoico, E., Fantin, F., Panourgia, M. P., Di Francesco, V., Tosoni, P. (2004) Relation between leptin and the metabolic syndrome in elderly women. J Gerontol A Biol Sci Med Sci. 59: 396400.
  • 11
    Finkelstein, E. A., Fiebelkorn, J. C., Wang, G. (2003) National medical spending attributable to overweight and obesity: how much and who's paying? Health Aff (Millwood) (suppl)W3-219W3-226.
  • 12
    Galanos, A. N., Pieper, C. F., Cornoni-Huntley, J. C., Bales, C. W., Fillenbaum, G. (1994) Nutrition and function: is there a relationship between body mass index and the functional capabilities of community-dwelling elderly? J Am Geriatr Soc. 42: 368373.
  • 13
    Apovian, C. M., Frey, C. M., Wood, C. G., Rogers, J. Z., Jensen, GL. (2002) Body mass index (BMI) and physical function in older women. Obes Res. 10: 740747.
  • 14
    Jensen, G. L., Friedmann, JM. (2002) Obesity is associated with functional decline in community-dwelling rural older persons. J Am Geriatr Soc. 50: 918923.
  • 15
    Miller, G. D., Rejeski, W. J., Williamson, J. D., et al (2003) The Arthritis, Diet and Activity Promotion Trial (ADAPT): design, rationale, and baseline results. Control Clin Trials 24: 462480.
  • 16
    Zoico, E., Di Francesco, V., Guralnik, J. M., et al (2004) Physical disability and muscular strength in relation to obesity and different body composition indexes in a sample of healthy elderly women. Int J Obes Relat Metab Disord. 28: 234241.
  • 17
    Sharkey, J. R., Branch, L. G., Guiliania, C., Zohoori, M., Haines, PS. (2004) Nutrient intake and BMI as predictors of severity of ADL disability over 1 year in homebound elders. J Nutr Health Aging 8: 131139.
  • 18
    Sharkey, J. R., Giuliani, C., Haines, P. S., Branch, L. G., Busby-Whitehead, J., Zohoori, N. (2003) Summary measure of dietary musculoskeletal nutrient (calcium, vitamin D, magnesium, and phosphorous) intakes is associated with lower-extremity physical performance in homebound elderly men and women. Am J Clin Nutr. 77: 847856.
  • 19
    Millen, B. E., Silliman, R. A., Cantey-Kiser, J., et al (2001) Nutritional risk in an urban homebound older population: the nutrition and healthy aging project. J Nutr Health Aging 5: 269277.
  • 20
    Sharkey, J. R., Branch, L. G., Zohoori, N., Giuliania, C., Busbly-Whitehead, J., Haines, PS. (2002) Inadequate nutrient intakes among homebound elderly and their correlation with individual characteristics and health-related factors. Am J Clin Nutr. 76: 14351445.
  • 21
    Sharkey, J.R. (2003) Risk and presence of food insufficiency are associated with low nutrient intakes and multimorbidity among homebound older women who receive home-delivered meals. J Nutr. 133: 34853491.
  • 22
    Pennsylvania Department of Health, Pennsylvania Vital Statistics (1997) Annual Report, 1995 Pennsylvania Department of Health, Division of Health Statistics and Research Harrisburg, PA.
  • 23
    Klein, G. L., Kita, K., Fish, J., Sinkus, B., Jensen, GL. (1997) Nutrition and health services for older persons in rural America: a managed care model. J Am Diet Assoc. 97: 885888.
  • 24
    Jensen, G. L., Kita, K., Fish, J., Heydt, D., Frey, C. (1997) Nutrition risk screening characteristics of rural older persons: relation to functional limitations and health care charges. Am J Clin Nutr. 66: 819828.
  • 25
    Jensen, G. L., Friedmann, J., Coleman, C. D., Wright, HS. (2001) Screening for hospitalization and nutrition risks among community dwelling older persons. Am J Clin Nutr. 74: 201205.
  • 26
    Mitchell, D. C., Wright, H. S., Friedmann, J. M., Jensen, GL. (2002) Food frequency questions in the Level II Screen are generally sensitive but not specific indicators of nutrition risk in older adults. J Am Diet Assoc. 102: 842844.
  • 27
    Taylor-Davis, S., Smiciklas-Wright, H., Davis, A. C., Jensen, G. L., Mitchell, DC. (1998) Time and cost for recruiting older adults. J Am Geriatr Soc. 46: 753757.
  • 28
    Ledikwe, J. H., Wright, H. S., Mitchell, D. C., Jensen, G. L., Friedmann, J. M., Still, CD. (2003) Nutrition risk assessment and obesity in rural older adults: a gender difference. Am J Clin Nutr. 77: 551558.
  • 29
    Apovian, C. M., Frey, C. M., Rogers, J. Z., McDermott, B., Jensen, GL. (1996) Body mass index and physical function in obese older women. J Am Geriatr Soc. 44: 14871488.
  • 30
    Friedmann, J., Elasy, T., Jensen, GL. (2001) The relationship between body mass index and self-reported functional limitation among older adults: a gender difference. J Am Geriatr Soc. 49: 398403.
  • 31
    Seale, J. L., Klein, G., Friedmann, J., Jensen, G. L., Mitchell, D., Wright, HS. (2002) Energy expenditure by doubly labeled water, activity recall, and diet records in the rural elderly. Nutrition 18: 568573.
  • 32
    Ledikwe, J. H., Smiciklas-Wright, H., Mitchell, D. C., Miller, C. K., Jensen, GL. (2004) Dietary patterns of rural older adults are associated with weight status and nutritional status. J Am Geriatr Soc. 52: 589595.
  • 33
    Blackburn G. L., Dwyer J. T., Wellman NS., eds. Nutrition Interventions Manual for Professionals Caring for Older Americans The Nutrition Screening Initiative Washington, DC (1992).
  • 34
    Wellman, NS. (1994) The Nutrition Screening Initiative. Nutr Rev. 52: S44S47.
  • 35
    Katz, S., Downs, T. D., Cash, H. R., Grotz, RC. (1970) Progress in development in the index of ADL. J Gerontologist 10: 2030.
  • 36
    Lawton, MP. (1988) Scales to measure competence in everyday activities. Psychpharmacol Bull 24: 609615.
  • 37
    NHLBI Obesity Education Initiative Expert Panel (September 1998) Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults: The Evidence Report National Institutes of Health, National Heart, Lung and Blood Institute Bethesda, MD. XIV, publication no. 98-4083.
  • 38
    Reuben, D. B., Ix, J. H., Greendale, G. A., Seeman, TE. (1999) The predictive value of combined hypoalbuminemia and hypocholesterolemia in high functioning community-dwelling older persons. J Am Geriatr Soc. 47: 402406.
  • 39
    Baker, P. S., Bodner, E. V., Allman, RM. (2003) Measuring life-space mobility in community-dwelling older adults. J Am Geriatr Soc. 51: 16101614.
  • 40
    Arias, E., Smith, BL. (2003) Deaths: preliminary data for 2001. Natl Vital Stat Rep. 51: 144.
  • 41
    Jensen, G. L., Roy, M. A., Buchanan, A. E., Berg, MB. (2004) Weight loss intervention for obese older women: improvements in performance and function. Obes Res. 12: 18141820.
  • 42
    Villareal, D. T., Apovian, C. M., Kushner, R. F., Klein, S. (2005) Obesity in older adults: technical review and position statement of the American Society for Nutrition and NAASO, the Obesity Society. Am J Clin Nutr. 82: 923934.
  • 43
    Wellman, N. S., Rosenzweig, L. Y., Lloyd, JL. (2002) Thirty years of Older Americans Nutrition Program. J Am Diet Assoc. 102: 348350.
  • 44
    Mathematica Policy Research Inc. (1996) Serving Elders at Risk, Older Americans Act Nutrition Programs: National Evaluation of the Elderly Nutrition Program 1993–1995, Volume I: Title III Evaluation Findings U.S. Department of Health and Human Services Washington, DC.
  • 45
    Finelli-Kuczmarski, M., Weddle, DO. (2005) American Dietetic Association Position Statement: nutrition across the spectrum of aging. J Am Diet Assoc. 105: 616633.