Comparison of Nutritional Risk Between Urban and Rural Elderly

    Elderly people encompass a large range of ages, from 65 to well over 100 years of age. The US census for 2000 reported 35 million people over the age of 65, representing approximately 12.4% of the total US population. People over age 85 were the fastest growing segment of the American population, increasing by 38% during the 1990s.¹ In Montana, multiple counties report that more than 20% of their population is older than age 65.¹

    Adequate nutrition is an important component of health maintenance for elderly people. Among the elderly, physical, social, and environmental factors may contribute to nutritional problems. These, in turn, often lead to prolonged hospitalization and less-than-optimal treatment outcomes, including poor wound healing.^2,3 Often, poor nutritional status is compounded by the presence of chronic illness, with up to 67% of "well" elderly reporting the presence of at least one chronic illness.⁴ In poorly nourished hospitalized elderly, morbidity and mortality increase even when no overt clinical signs of malnutrition are present.^5,6 This may be because weight loss, usually considered the benchmark of nutritional decline, is actually a late sign of malnutrition. By the time a significant amount of weight has been lost, the elderly person's nutritional status is severely compromised. When indicators such as screening questionnaires and laboratory values are used, 52% to 85% of institutionalized elderly, and more than 55% of hospitalized elderly, can be considered malnourished.^3,7,8

    The number of people residing in nursing homes decreased from 5.1% in 1990 to 4.5% in 2000.¹ This means more people over age 65 are living in community settings and may or may not be receiving home health care services or participating in senior meal programs.⁹ An evaluation of community-based programs under the Elderly Nutrition Program of the Older Americans Act found 67% to 88% of the participants to be at moderate to high risk for malnutrition.¹⁰ Of elderly receiving home-delivered meals, 43% reported a hospital or nursing home stay in the previous year.¹⁰ Among applicants to the Meals-on-Wheels program, 98% were found to be at risk nutritionally using the Nutrition Screening Initiative (NSI) self-assessment tool.¹¹ Additionally, participation in congregate-site meal programs did not significantly affect the nutritional status of the participants, with 26% of the population consuming diets that placed them at risk for malnutrition.¹²

    The availability of needed community services varies considerably between rural and urban areas. Rural elderly rely on family members for help with cooking and shopping more than their urban counterparts.^13,14 Many rural towns are extremely small, so access to shopping and healthcare often requires travel over long distances. Weather and road conditions may further influence accessibility, and public transportation is often lacking. Specifically, low population densities make availability of needed health, dental, social, and nutritional services questionable in geographically isolated rural Montana, where the average population density is six persons per square mile, and 44 out of 56 counties are designated as rural or frontier.^1,14,15

    Urban elderly have more health and nutrition services available in the larger cities of Montana than their rural counterparts. However, barriers still exist. Public transportation to urban centers from small adjoining urban/rural interface communities or townships is often lacking. Some areas designated as urban do not have local meal programs or senior centers^14,15; the availability of taxi service is limited, and grocery and drug stores are sparse in urban/rural interface areas.^14,16

    Both urban and rural elderly face a multitude of barriers to maintaining adequate nutrition. Physical changes and social and environmental concerns affect elderly persons regardless of community size.¹⁷ As the population of elderly multiplies and healthcare dollars decrease, addressing nutritional issues before a problem develops will become increasingly important. Looking at commonalties and differences between urban and rural elderly that are associated with nutritional risk is an important first step in the identification and timely treatment of nutritional issues. However, a comparative picture of nutritional status in the elderly population living in rural and urban areas is lacking. The objectives of this research were to: 1) examine health behaviors and nutrition information associated with nutritional risk in urban and rural areas, and 2) compile a comprehensive profile of nutritional risk for rural and urban elderly.

Methods

    Data for this analysis were compiled from four research projects, two urban and two rural, conducted on people age 65 and older residing in Montana. Urban participants for Studies 1 and 2 were recruited from a Montana county with a population of 100,000 or more. Rural participants were recruited from counties that were 25 miles outside of a population center of 12,000 or more. Multiple rural/frontier counties across Montana were used for Studies 3 and 4.
Data for these studies were collected between 1997 and 2001.

    Instruments. Data for the four studies were collected using a variety of formats including chart reviews, laboratory data, questionnaires, and interviews. In Study 1, activities of daily living (amount of assistance necessary to dress, feed, bath, toilet, and ambulate), cognitive status, and nutritional risk were obtained from acute care hospital medical records. Nurses assessed nutritional risk for patients on admission to the acute care facility as present or absent. If present, risk was determined to be from disease, physical problems, or inability to eat. In Study 2, participants were asked to answer specific questions about eating habits. These questions were similar but not identical to the nutritional questions used in Studies 3 and 4. The Nutritional Screening Initiative instrument, used extensively in elderly nutritional assessment¹⁸ (eg, to screen all Title-III participants on entry into a senior meal program) has demonstrated consistent reliability and validity. This tool was used to collect data on all rural persons for Studies 3 and 4.

    Laboratory tests that measure nutritional status including serum albumin were available in all four studies. Other laboratory values (hemoglobin, hematocrit, and total serum protein) were available for rural and urban independently living elderly in Studies 2, 3 and 4. When possible, all data were summarized for each group.
The Quality of Life Rating Scale (QOLRS) was used in studies 2, 3, and 4. The QOLRS is a two-item self-report scale that asks people to rate their satisfaction with health and quality of life as global concepts.²¹ Data on health habits, including questions on use of healthcare services (medical and dental visits, blood pressure check), smoking, exercise, and sleep habits, were collected in all four studies. All studies received prior Internal Review Board approval.

    Study descriptions. Study 1:Urban elderly admitted to an acute care facility. This descriptive study used retrospective chart review methods to determine patient characteristics that identified increased nutritional risk on admission to an acute care facility.⁷ Study information extracted from the physician and nurse's admission notes included activities of daily living, use of dietary supplements, and nurse's determination of nutritional risk on admission. Laboratory values for serum albumin were obtained within 48 hours of facility admission. Data were collected for 137 participants.

    Study 2: Urban independently living elderly. This was a descriptive study designed to examine wellness behaviors and nutritional status in independently living elderly.⁴ Study information was obtained from participant questionnaires and included demographics, health behaviors, and eating habits. Blood was drawn for serum albumin, total serum protein, hemoglobin, and hematocrit values. The 46 study participants were recruited from the volunteer office at an urban area hospital.

    Study 3: Rural elderly. This study used a convenience sample of rural elderly from five of the 11 rural/frontier regions in Montana.¹⁹ Participants age 65 and older were recruited from senior citizens centers and Meal-on-Wheels programs in their community. The total sample consisted of 80 rural elderly - 40 who received Meals-on-Wheels and 40 who did not. Study participants had blood drawn for serum albumin, total serum protein, hemoglobin, and hematocrit levels. They completed a questionnaire and personal interview that included demographic information, health behaviors, QOLRS, and the NSI.

    Study 4: Rural elderly with diabetes. Participants in the main study included all persons age 45 years and older diagnosed as having diabetes and treated by a healthcare provider at one of four rural healthcare clinics between January 1, 1999 and August 1, 2000.²⁰ For the purposes of this study, only the subset of participants that were 65 years of age or older (n = 87) were included. A letter and questionnaire booklet that explained the research study were mailed to all possible participants. The questionnaire developed for this study included demographic variables, general health information, QOLRS, and the NSI.

    Statistical analysis. Data were entered into an SPSS 11.5 database for analysis. The results presented are from multiple studies that were designed to answer different questions; consequently, not all information was available for each group. Individual study results are presented here, as well as aggregate results for urban and rural elderly. Statistical comparisons have been made between 1) total rural versus total urban, and 2) urban independently living versus rural independently living elderly.
Nutritional risk was dichotomized as present or absent. Risk was defined as anyone having a serum albumin of < 3.5 gm/dL, scoring moderate or high risk on the NSI tool, or being judged at risk by the hospital nursing staff on admission. Results of the studies were combined into table format to examine descriptive statistics from each study. Items that correlated with nutritional risk were examined from each study for similarities. Finally, common items were placed into a forward, step-wise, logistic regression to look at predictors of nutritional risk for each population.

Results

    Demographics. One hundred, eighty-three (183) urban-living and 167 rural-living people participated. No statistical difference between groups for age or gender was found. The mean age of urban elderly was 77 (+6) and rural was 77 (+7) years of age. In the urban sample, 39% were male and 61% female. In the rural sample, 36% were male and 64% female. The gender percentages in the two samples are consistent with the overall US census information for the 75- to 84-year-old percentages of 39% male and 61% female, respectively1 (see Table 1).

    Study 1. Among the 137 participants, the mean serum albumin of 3.0gm/dL (normal 3.5 to 5.0 gm/dL) was below normal, although the majority of subjects - 113 (82.5%) - were not recorded as experiencing weight loss in the past 3 months. Seventy-one subjects (71, 52%) were assessed as having poor nutritional status because of disease, physical problems, or an inability to eat. However, 46 participants (65%) assessed by acute care staff as having nutritional problems on admission were not prescribed nutritional supplements or nutritional consults.

    Study 2. Among the 46 participants, the mean serum albumin was 3.7gm/dL (SD = .20), serum protein 6.89gm/dL (SD = .47), hemoglobin 14.21gm/dL (SD =. 96), and hematocrit 41.7% (SD = 2.68). The majority of the participants (99%) did not experience any difficulty eating nor were they on a special diet, and 39 (93%) had not lost weight in the past 3 months. However, 20 (49%) ate their meals alone, 30 (71%) considered themselves to be overweight, and no one used nutritional supplements.

    Study 3. Among the 80 participants, NSI scores indicated that 37 (46.3%) were at high risk for malnutrition and 27 (33.8%) were at moderate risk. Twenty-three participants (23, 29%) had a body mass index (BMI) of 30.0 or greater and 22 (27.5%) had a BMI of 23 or less. More than half (48 or 60.8%) of the sample reported they ate three meals a day, 18 (22.8%) ate two meals, and nine (11.4%) ate one meal. Mean laboratory values for serum albumin, protein, hemoglobin, and hematocrit were all within normal limits.

    Study 4. The majority of the 87 participants (90%) felt their health was about the same or better than others their age, although 88.5% said it had worried them some in the past year. They were positive regarding how healthy they were today and rated their overall quality of life as a 7 (+2) on a 10-point Likert scale. One hundred, seventy-six (176, 89%) of the respondents had a yearly medical checkup, 192 (98%) had their blood pressure checked in the past 6 months, and 71 (37%) had a yearly dental examination. The NSI scores ranked 61 (73%) at moderate risk and 17 (20%) at high risk for malnutrition (see Table 2).

    Health status and behaviors. Urban elders rated how healthy they felt significantly higher than their rural counterparts (P <0.05). Urban elderly also rated quality of life higher than rural elderly, but this difference was not statistically significant. Routine healthcare results were mixed. Rural dwellers had a higher percentage of yearly medical examinations and blood pressure checks but fewer dental visits compared to urban residents. Among male participants, 89% had yearly prostate examinations. Yearly mammograms were obtained for 55% of all woman, and 53% performed regular breast self-examinations. No significant difference in rate of breast or prostate screening in rural versus urban areas was noted (P >0.05). The majority of the total sample (>90%) did not smoke (see Table 2).

Nutrition status.
    Laboratory values/BMI. Urban elderly had significantly lower serum albumin levels than their rural counterparts, with elderly newly admitted to an acute care facility having the lowest (3.0 gm/dL, P <0.01). Overall, 60% of urban elderly and 4% of rural elderly had serum albumin levels below 3.5gm/dL. When urban independently living elderly are examined separately from those newly admitted to acute care, 22% had albumin levels of 3.5 gm/dL or less. Total serum protein values trended higher for rural elderly as well (P < 0.01). Hemoglobin and hematocrit values were slightly lower for rural over urban elderly (13.8/14.2gm/dL and 41.3/41.7%, respectively). This was not statistically significant (P >0.05). Body mass index was highest for rural diabetic patients (BMI = 30.4), with rural elderly generally higher than their urban counterparts (28.3/25.7); 54% of the urban elderly had BMI levels within normal limits, compared to 29% of rural elderly (see Table 3).

    Nutrition screening tools. The NSI rated 85% of the rural elderly at some degree of nutritional risk; 53% were found to be at moderate risk and 32% at high risk. Based on NSI score and serum albumin levels, 86.5% of the rural elderly were dichotomized as at nutritional risk (NSI scores were not available for urban elderly). Nurses rated 53% of elderly newly admitted (on admission forms) to an acute care setting as having nutritional risk. Among independently living elderly, 46% were judged to be at some risk using a nutritional questionnaire. When albumin levels <3.5gm/dL are included, 85% of urban elderly could be considered at risk nutritionally. No statistical difference was found between urban and rural elderly for nutritional risk.

     Correlation between patient variables and nutritional risk also were examined. For elderly in an acute care facility needing assistance with activities of daily living, incontinence, weight loss, decreased cognitive status and serum albumin were correlated with the admitting nurse's assessment of "poor" (P <0. 05). Similar patient variables were associated with inadequate nutrition in the independently living urban and rural participants. These included needing assistance with shopping, feeding, or cooking; not having enough money for food; having tooth and mouth problems or poor eating habits; using of three or more medications per day; and having health concerns in the past 3 months (P <0. 01).

Nutritional Risk Profile

    Variables from each study were entered into a forward logistic regression to develop models of risk predictive of poor nutrition. Model development allowed for examination of common predictors of risk across groups. Rural elderly were dichotomized as nutritionally at risk based on their NSI scores. Urban independently living elderly were dichotomized as having nutritional risk if their serum albumin was less than 3.5 and their nutrition questions assessed risk. Elderly persons newly admitted to acute care were dichotomized as nutritionally at risk if the nurse rated them as being at nutritional risk from disease or physical conditions on the admission assessment.

    For both rural samples, money problems, and eating few fruits and vegetables were common themes in the logistic regression analysis. Taking more than three drugs a day, changing the type or amount of food eaten because of a medical condition, and eating few fruits and vegetables were included in the logistic regression model for overall nutritional risk. Tooth or mouth problems, changing the types or amount of food eaten and eating few fruits and vegetables for high nutritional risk. Eating few fruits or vegetables, eating alone, and taking three or more medications were indicative of any level of nutritional risk in the rural diabetic patients (see Table 4).

    For the elderly newly admitted to an acute care setting, needing assistance with activities of daily living (especially dressing and feeding) and a recent weight loss were most predictive of nutritional risk. For independently living urban elderly, having a poor appetite or eating few bread and grain products were included in the final logistic regression model.

Discussion

    Results of these four studies indicate a significant number of older residents in Montana are at increased risk for malnutrition and decreased quality of life. Results from these studies do not support the general belief that rural seniors are at higher risk of malnutrition than their urban counterparts. When screened, more than 80% of urban and rural seniors have moderate or severe nutritional risk. These figures are consistent with previous research findings.⁹ Results further suggest that physical, social, and environmental factors affect the nutritional status of seniors regardless of where they reside.

    Physical problems that influence nutrition include poor nutritional habits, chronic conditions and the associated polypharmacy, and dental and mouth problems. Logistic regression results suggest loss of activities of daily living (ADLs), independent activities of daily living (IADLs), and taking multiple medications influence nutritional status in both rural and urban seniors. The literature suggests that the increase in chronic illness associated with aging has a significant impact on cognitive and physical functioning.^22-24 This was supported in these studies because the average person had 3.7 chronic health conditions and took more than three medications per day regardless of where he/she lived.

    Serum albumin has been widely used as a marker of poor nutritional status. It is an inexpensive test to perform and reflects a half-life of approximately 28 days.^25,26 Urban elderly newly admitted to acute care had the lowest albumin levels, followed by independent urban elderly and rural patients with diabetes. The rural seniors, either independent or receiving Meals-on-Wheels, had the highest albumin levels. Seniors admitted to acute care could be expected to have lower albumin levels as a result of illness. However, when independently living urban seniors are examined separately, their albumin levels are still significantly lower than their rural counterparts (P <0.01).

    Rural residents did not have regular dental examinations; 76% of urban elderly had yearly dental care as compared to only 41% of rural residents. Poor dentition is associated with decreased nutritional status.^19,27,28 Among rural independently living elderly, only 35% of denture wearers had been to a dentist in the past year and 47% had not been to a dentist for 5 or more years. Despite the fact that 94% of the rural residents said a dentist was available in their community, some people indicated that they had to travel more than 60 miles for dental care.²⁹ This may be viewed as a disconnect between the rural versus urban resident's perception of provider availability.

    Lack of a regular exercise program and above normal BMI were areas of concern for both rural and urban elderly. Half of the urban elderly and 65% of the rural elderly did not exercise on a regular basis. A BMI >30, indicative of significant obesity, was noted in 18% of urban and 31% of rural subjects. Although a loss of body weight and under-nutrition in the elderly is a usual concern, obesity is also indicative of malnutrition.³⁰ Rural persons had a BMI indicative of obesity at higher levels than their urban counterparts. This may be due to a lack of exercise facilities and senior exercise programs in the rural areas. Both urban and rural elderly in this study had a higher percentage of obesity than previous research that used the Behavioral Risk Factor Surveillance System (BRFSS) data,³¹ which showed that 17.6% of urban and 18.9% or rural had a BMI > 30.³¹

    Social and environmental factors are important determinates of nutritional status that are frequently overlooked in the assessment process. Social factors, including a change in living arrangement, death of a spouse, reduced income due to retirement, and a continuation of poor life-long eating habits, are important regardless of rural or urban location.²⁵ Logistic regression results showed eating alone and poor dietary choices are important predictors of nutritional risk. Social factors were a concern for everyone, especially the rural Meals-on-Wheels participants: 22% of the rural elderly enrolled in the Meals-on-Wheels program ate fewer than two meals a day, 68% ate alone, and 17% reported they did not have enough money for food every day. More than half of this group was widowed and 60% earned less than $10,000 a year. Forty-three percent of both independent rural and urban seniors not enrolled in meals programs and 26% of rural diabetic seniors ate alone.

    Environmental concerns include distance to services, difficulty with IADLs, and cooking facilities that are unable to accommodate physical limitations. Urban elderly rated their quality of life slightly higher than rural dwellers. However, as people became dependent for activities of daily living, a corresponding decrease occurred in their self-reported quality of life. A significant difference also was noted between rural elderly who were independent and homebound persons who received Meals-on-Wheels (see Figure 1). Quality-of-life self-rating was not available for the acute care group. Of those seniors questioned, the lowest quality of life and state of health was reported by the Meals-on-Wheels recipients, followed by rural patients with diabetes. Rural individuals had the greatest distance to travel for services and medical care. Low quality-of-life scores could be more reflective of their geographic isolation than their perception of service availability.

    These studies also looked at health habits of urban and rural seniors. Chronic diseases may have a negative impact on nutritional status. However, people who obtain regular medical care are more likely to have their nutritional deficiencies diagnosed and treated earlier by their healthcare practitioner. This includes those receiving prostate examination and mammograms. Overall, men had below-optimal rates of testicular self-examination (31.5%), and woman had low rates of mammograms (55%) and breast self-examinations (48%). This is consistent with the findings of Kumar et al,³¹ who reported approximately 66% of rural and 73% of urban woman had not had mammograms in the past 12 months. This is worrisome because senior women are at increased risk for breast cancer and mammogram facilities are readily available in all areas. Although the risk for testicular cancer decreases with age, preventive services are still needed.

Conclusion

    Nutritional risk is difficult to study in seniors. Physical, social, and environmental factors that affect nutrition are all highly related. Even though standardized assessment tools are used, they might not provide a complete picture. Demographic information, chronic illness that affects function, and selected laboratory studies such as albumin need to be examined when determining nutritional risk in the elderly. Commonalities such as living alone, low income, and chronic disease were found for the elderly regardless of where they resided. Future research needs to look at both these commonalities and differences to help provide the needed nutritional services and programs.

    Primary and specialized medical providers and services are clustered in the urban areas of Montana. Some rural residents reported that they needed to travel as far as 115 miles for primary healthcare and 150 miles for specialized care. Dentists were as far as 67 miles away, and physical and occupational therapists, registered dieticians, and nurse educators were more than 100 miles away. Unlike their rural counterparts, urban residents had exercise programs available, as well as many fast food restaurants that allowed access to low cost meals they do not have to prepare themselves. However, this situation places urban residents at risk for consuming empty calories high in fat content. This may be one reason rural elderly had higher albumin levels; the connection between restaurant food consumption should be examined further.

    Older persons often expect to have medical problems and aches and pains. Practitioners need to recognize that the elderly might consider some things to be signs of "normal" aging when in fact they are symptoms of underlying medical etiologies that warrant attention. Nutritional assessment needs to be included as part of all yearly examinations. Because the majority of both rural and urban elderly has yearly medical checkups, people at risk could be easily identified and, if necessary, nutritional interventions implemented more efficaciously. Finally, longitudinal studies with samples drawn from diverse geographic locations that look at changes in nutritional status and the relationship to changes in medical status are needed. Simple, inexpensive, easy-to-follow nutritional support programs should be initiated to see what is most effective in maintaining adequate nutrition.

Acknowledgments

    The authors wish to thank Brenda Hellyer, BSN, RN; Dorothy Kindsfater, MSN, RN; Lynn Paul, EdD, RD; and Clarann Weinert, SC, PhD, RN, FAAN, for their participation in this project.

1. United States Census Bureau: US Government; 2002. Available at: www.census.gov. Accessed October 21,2003.

2. Reuben DB, Moore AA, Damesyn M, Keeler E, Harrison GG, Grendale GA. Correlates of hypoalbuminemia in community-dwelling older persons. Am J Clin Nutr. 1997;66:38-45.

3. Axen KV, Schnoll R. Nutritional issues in the frail older person. Topics in Geriatric Rehabilitation. 1995;11:1-10.

4. Zulkowski K. Nutritional status of independently living elderly. Ostomy/Wound Management. 2000;46(2);56-60.

5. Gloth MF, Gundberg CM, Hollis BW, Haddad JG, Tobin JD., Vitamin D deficiency in homebound elderly persons. JAMA. 1995;274:1683 -1686.

6. Zulkowski K. A conceptual model of pressure ulcer prevalence: MDS+ items and nutrition. Ostomy/Wound Management. 1999;45(2):36-44.

7. Zulkowski K, Kindsfater, D. Nutritional status of elderly persons newly admitted to an acute care setting. Ostomy/Wound Management. 2000;46:32-38.

8. American Dietetic Association. Position of The American Dietetic Association: cost effectiveness of medical nutrition therapy. J Am Diet Assoc. 1995;95:88-91.

9. McCormack P. Undernutrition in the elderly population living at home in the community: a review of the literature. J Adv Nur. 1997;26:856-863.

10. Ponza M, Ohls JC, Millen BE., Serving elderly at risk: the Older Americans Act Nutrition Programs. National Evaluation of the Elderly Nutrition Program 1993-1995. Washington, DC: Mathematica on Policy Research, Inc.;1996.

11. Coulston AM, Craig L, Voss AC. Meals-on-wheels applicants are a population at risk for poor nutritional status. J Am Dietetic Assoc. 1996;96:570 -573.

12. Neyman MR, Zidenberg-Cherr S, McDonald RB. Effect of participation in congregate-site meal programs on nutritional status of the healthy elderly. J Am Dietetic Assoc. 1996;96:475 -483.

13. Stevens DA, Grivetti LE, McDonald RB. Nutrient intake of urban and rural elderly receiving home-delivered meals. J Am Dietetic Assoc. 1992;92:714 -718.

14. Montana Department of Public Health and Human Services. 2002 Montana County Health Profiles. Available at: www.dphhs.state.mt.us/hpsd/pubheal/healplan/profiles/index.htm. Accessed October 21, 2003.

15. State of Montana 2002. Available at: www.dphhs.state.mt.us/hpsd/pubheal/healplan/profiles/index.htm. Accessed October 21, 2003.

16. Riley A. Personal communication. Lockwood, Mont. Fire Department Chief;2002.

17. Salva A, Pera. G. Nutrition and aging. Public Health Nutrition. 2001;4:1375 -1378.

18. American Academy of Family Physicians, Nutrition Screening Initiative 2002. Available at: http://www.aafp.org/nsi/. Accessed October 21, 2003.

19. Zulkowski K, Paul L, Weinert, C. Promoting optimal health through nutrition for rural elderly in Montana 2002. USDA Grant data.

20. Coon P, Zulkowski K., Patient and provider compliance with ADA guidelines. AOA Grant data; 2002.

21. Graham K, Longman AJ. Quality of life and persons with melanoma. Cancer Nursing. 1987;10:338 -346.

22. Dwyer JT. An assessment lexicon: assessment of dietary trends, physical activity patterns and nutritional status in the elderly. Journal of Nutrition Health & Aging. 2001;5:108 -112.

23. Brach JS, VanSwearingen JM. Physical impairment and disability: relationship to performance of activities of daily living in community-dwelling older men. Physical Therapy. 2002;82:752-761.

24. Zulkowski K. MDS items associated with pressure ulcer prevalence in newly institutionalized elderly. Dissertation: State University of New York-UB;1998.

25. Niedert K. (ed). Nutrition Care of the Older Adult. Chicago, Ill. American Dietetic Association; 1998.

26. Clochest JM, Davidson LJ, Piper-Caulkins E, Carno M, Bauldoff GS. Use of serum albumin level in studying clinical outcomes. Outcomes Management in Nursing Practice. 1999;8:61 -66.

27. Ritchie CS, Joshipura K, Hung H, Douglass CW. Nutrition as a mediator in the relation between oral and systemic disease: associations between specific measures of adult oral health and nutritional outcomes. Critical Review of Oral Biological Medicine. 2002;13:291 -300.

28. Sheiham A, Steele JG, Marcenes W, et al. The relationship among dental status, nutrition intake, and nutritional status in older people. Journal of Dental Research. 2000;80:408 -413.

29. Zulkowski K, Albrecht D. Dental status of rural elderly. USDA data2002.

30. Tufts University. Weight loss a "slippery slope" for seniors. 2002. Abvailable at: http://www.healthand age.com/Phome/gid2=1692. Accessed October 21, 2003.

31. Kumar V, Acanfora M, Hennessy CH, Kalache A. Health status of the rural elderly. Journal of Rural Health. 2001;17:328--331.