The Relationship Between Nutritional Status and Physical Function, Admission Frequency, Length of Hospital Stay, and Mortality in Old People Living in Long-Term Care Facilities Sin-Jie Lin 1 & Shinn-Jang Hwang 2 & Chieh-Yu Liu 3 & Hung-Ru Lin 4 * 1 RN, MSN, Department of Medical Research, National Taiwan University Hospital & 2 MD, Director, Department of Family Medicine, Taipei Veterans General Hospital, and Professor, School of Medicine, National Yang Ming University & 3 PhD, Associate Professor, Department of Nursing, National Taipei University of Nursing and Health Sciences & 4 RN, PhD, Associate Professor and Associate Chair, Department of Nursing, National Taipei University of Nursing and Health Sciences. Introduction Taiwan’s aging population has made care of old people a critical issue. Demand for long-term care facilities (LTCFs) for old people has increased. According to Taiwan’s De- partment of Social Affairs under the Ministry of the In- terior (2011), there were only 183 public and private LTCFs in Taiwan at the end of 1999. At the end of January 2011, this number had increased to 1,050, indicating strong demand. Elderly health is complex, and changes in nutritional sta- tus are critical. The physical, mental, and social deteriora- tion that accompanies aging directly or indirectly restricts diet quantity and quality, which can result in dystrophy or malnutrition (Singh, Bhaldraithe, Bondin, & Goorah, 2010). Russell and Elia (2008) highlighted the risk of malnutri- tion growing with age and noted a higher incidence among those currently receiving care or living in institutions. Studies addressing the situation in various countries found that some 35%Y65% of hospitalized old people (Omran & Morley, 2000; Stechmiller, 2003; Suominen, Sandelin, Soini, & Pitkala, 2009) and 20%Y50% of elderly LTCF residents experienced malnutrition (Chen, 2007; Chiu, Lin, Hsieh, ABSTRACT Background: Nutrition is an important issue for elderly res- idents of long-term care facilities (LTCFs). About 20% of el- derly LTCF residents in Taiwan are malnourished. Purpose: This study investigated correlations between nutri- tional status and physical function, admission frequency, hos- pital stay duration, and mortality in elderly LTCF residents. Methods: Researchers used a retrospective study design and convenient sampling to enroll 174 subjects aged 67 to 105 years (average, 82.5 years) who were living in legally reg- istered LTCFs in Beitou District, Taipei City, Taiwan. A review of LTCF resident files provided data on subjects’ demograph- ics, physical examination laboratory results for the most recent 1-year period, anthropometry, physical function, admission fre- quency, hospital stay duration, and causes of admissions. Sub- jects had lived in their LTCF for more than 1 year before their enrollment date. Subjects who died during and after the study period were also included in analysis. Results: Results showed significant changes over the study year in subjects’ nutritional status, physical function, and calf circumference. Physical function was found significantly cor- related with calf circumference, hospitalization status was found correlated with nasal-gastric tube feeding status, and eating pattern was found correlated with calf circumference and levels of both serum albumin and cholesterol. Nutritional status, calf circumference, albumin level, and cholesterol level also correlated significantly with hospitalization status. In this study, the likelihood of hospitalization increased with age and nasal-gastric tube feeding use. Hospital stay duration for sub- jects receiving nasal-gastric tube feeding was longer than that for those receiving oral feeding. Also, weak nutritional status scores for calf circumference and hemoglobin levels were fac- tors associated with increased mortality risk. Conclusions/Implications for Practice: Findings recom- mend that greater attention should be paid to the nutritional status of elderly persons living in LTCFs to reduce hospital- ization and death risks, cut medical expenses, and improve quality of care. KEY WORDS: long-term care facility, nutritional status, admission frequency, hospital stay duration, mortality. The Journal of Nursing Research h VOL. 20, NO. 2, JUNE 2012 110 Accepted for publication: February 20, 2012 *Address correspondence to: Hung-Ru Lin, No. 365, Ming Te Rd., Beitou District, Taipei 11219, Taiwan, ROC. Tel: +886 (2) 2822-7101 ext. 3197; Fax: +886 (2) 2821-3233; E-mail: [email protected]doi:10.1097/jnr.0b013e318254eac9
12
Embed
The Relationship Between Nutritional Status and Physical ...2)p.110-1… · (2000) used the Mini Nutritional Assessment to investigate conscious elderly living in communities and
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Relationship Between Nutritional Statusand Physical Function, Admission Frequency,Length of Hospital Stay, and Mortality in OldPeople Living in Long-Term Care Facilities
1RN, MSN, Department of Medical Research, National Taiwan University Hospital & 2MD, Director, Departmentof Family Medicine, Taipei Veterans General Hospital, and Professor, School of Medicine, National Yang MingUniversity & 3PhD, Associate Professor, Department of Nursing, National Taipei University of Nursing andHealth Sciences & 4RN, PhD, Associate Professor and Associate Chair, Department of Nursing, National
Taipei University of Nursing and Health Sciences.
IntroductionTaiwan’s aging population has made care of old people acritical issue. Demand for long-term care facilities (LTCFs)for old people has increased. According to Taiwan’s De-partment of Social Affairs under the Ministry of the In-terior (2011), there were only 183 public and private LTCFsin Taiwan at the end of 1999. At the end of January 2011,this number had increased to 1,050, indicating strongdemand.
Elderly health is complex, and changes in nutritional sta-tus are critical. The physical, mental, and social deteriora-tion that accompanies aging directly or indirectly restrictsdiet quantity and quality, which can result in dystrophy ormalnutrition (Singh, Bhaldraithe, Bondin, & Goorah, 2010).Russell and Elia (2008) highlighted the risk of malnutri-tion growing with age and noted a higher incidence amongthose currently receiving care or living in institutions. Studiesaddressing the situation in various countries found that some35%Y65% of hospitalized old people (Omran & Morley,2000; Stechmiller, 2003; Suominen, Sandelin, Soini, &Pitkala, 2009) and 20%Y50% of elderly LTCF residentsexperienced malnutrition (Chen, 2007; Chiu, Lin, Hsieh,
ABSTRACTBackground: Nutrition is an important issue for elderly res-idents of long-term care facilities (LTCFs). About 20% of el-derly LTCF residents in Taiwan are malnourished.
Purpose: This study investigated correlations between nutri-tional status and physical function, admission frequency, hos-pital stay duration, and mortality in elderly LTCF residents.
Methods: Researchers used a retrospective study designand convenient sampling to enroll 174 subjects aged 67 to105 years (average, 82.5 years) who were living in legally reg-istered LTCFs in Beitou District, Taipei City, Taiwan. A reviewof LTCF resident files provided data on subjects’ demograph-ics, physical examination laboratory results for the most recent1-year period, anthropometry, physical function, admission fre-quency, hospital stay duration, and causes of admissions. Sub-jects had lived in their LTCF for more than 1 year before theirenrollment date. Subjects who died during and after the studyperiod were also included in analysis.
Results: Results showed significant changes over the studyyear in subjects’ nutritional status, physical function, and calfcircumference. Physical function was found significantly cor-related with calf circumference, hospitalization status wasfound correlated with nasal-gastric tube feeding status, andeating pattern was found correlated with calf circumferenceand levels of both serum albumin and cholesterol. Nutritionalstatus, calf circumference, albumin level, and cholesterol levelalso correlated significantly with hospitalization status. In thisstudy, the likelihood of hospitalization increased with age andnasal-gastric tube feeding use. Hospital stay duration for sub-jects receiving nasal-gastric tube feeding was longer than thatfor those receiving oral feeding. Also, weak nutritional statusscores for calf circumference and hemoglobin levels were fac-tors associated with increased mortality risk.
Conclusions/Implications for Practice: Findings recom-mend that greater attention should be paid to the nutritionalstatus of elderly persons living in LTCFs to reduce hospital-
ization and death risks, cut medical expenses, and improvequality of care.
The Journal of Nursing Research h VOL. 20, NO. 2, JUNE 2012
110
Accepted for publication: February 20, 2012*Address correspondence to: Hung-Ru Lin, No. 365, Ming TeRd., Beitou District, Taipei 11219, Taiwan, ROC.Tel: +886 (2) 2822-7101 ext. 3197; Fax: +886 (2) 2821-3233;E-mail: [email protected]
doi:10.1097/jnr.0b013e318254eac9
Li, & Chiou, 2005; Visvanathan, Penhall, & Chapman,2004). Guigoz, Lauque, and Vellas (2002) assessed the nu-tritional status of 2,264 elderly residents of 12 LTCFs inSwitzerland and found that 37% experienced malnutri-tion and 44% experienced potential nutritional risk.
Deterioration in physical function is common in oldpeople. Physical function disability refers to the inabilityto execute self-care and participate in community activi-ties such as those included under activities of daily living(ADLs), instrumented ADLs, and mobility and balance(Verbrugge & Jette, 1994). Physical function disability isnot only a predictive factor of health status, eligibility fornursing home admission (Ostir, Markides, Black, &Goodwin, 1998), and mortality in old people but alsomenaces their physical and psychological health (Cresset al., 1995). Taiwan’s Department of Statistics identified9.11% of Taiwan’s old people as having one or moredisabilities in 2009 (Taiwan Association of Long-TermCare Professionals, 2003) and old people with disabilitiesaccounting for the largest population receiving long-termcare. Shyu, Hsiung, Dai, Chen, and Huang (1993) foundin a survey of 143 residents living in 37 elderly care facil-ities that 36.4% were completely dependent and requirednutritional intake assistance.
Previous studies have suggested a close association be-tween nutritional status during hospitalization and longerhospital stays, patient mortality before and after discharge,and other complications (Atalay, Yagmur, Nursal, Atalay, &Noyam, 2008; Kaur, Miller, Halbert, Giles, & Crotty, 2008;Volkert, Pauly, Stehle, & Sieber, 2011). In Taiwan, Chan(2000) used the Mini Nutritional Assessment to investigateconscious elderly living in communities and care facilities.They identified a medical service utilization effect on nutri-tional status and found that old people with poorer nutri-tion used outpatient clinics, emergency rooms, and admissionservices more frequently. Failure to prevent or treat mal-nutrition early on can further restrict patient motility andcause decreased immunity, increased infection risk, pres-sure sores, delayed wound healing, poor prognoses, anddepression. Also, patient mental health and quality of lifemay be affected (Keller, Ostbye, & Goy, 2004; Kvamme,Gronil, Florholmen, & Jacobsen, 2011), and risks of mor-bidity, infection, complication, and death may increase (Keller& Ostbye, 2003; Volkert et al., 2011). The purpose of thisstudy was to investigate the nutritional status of old peo-ple living in LTCFs and analyze nutrition-related predictorsaffecting physical function, admission frequency, hospitalstay duration, and mortality.
Methods
Research DesignThis study used a retrospective design and collected demo-graphic, physical examination (serum albumin, hemoglo-bin, and cholesterol), anthropometry (body mass index
[BMI] and calf circumference), physical function (Barthelindex), admission frequency, hospital stay duration, andcause of death data from subjects’ medical records.
Study Sites and SubjectsThis study used convenience sampling. Subjects were allresidents of legally registered LTCFs located in TaipeiCity’s Beitou District in Taiwan. Subjects’ inclusion cri-teria included (a) aged 65 years or older and (b) enrolledas an LTCF resident for more than a year (including thosewho had passed away within that period). The eight LTCFstargeted for data collection were selected because of theirlong-term cooperation with the researchers. Residents of thoseeight facilities had all completed their health examinationsand laboratory tests at the same medical center. Also, allparticipating LTCFs used the same brand of chair scale tomeasure body weight.
Data CollectionThis study used a retrospective design. Data were collectedfrom subjects’ medical records.
Demographic data
Researchers collected demographic data in two areas:Personal characteristics: Gender, age, feeding pattern,
number of diagnosed chronic diseases, type of chronicdiseases, number of medications used, and type of med-ications used.
Physical function, hospitalization, and survival status:Barthel index score (physical function), admission fre-quency, hospital stay duration, and survival status/cause of death.
The Barthel index is used worldwide to assess func-tional independence and mobility. It is an ordinal scalethat compares 10 ADLs scored on a 5-point gradient scalewith a maximum total score of 100. Barthel index scale re-liability is quite high, earning a Cronbach’s alpha of .90.Validity analysis of the convergent criterion using the Func-tional Independence Measure showed satisfactory correla-tion in most areas. Explained variance was 63.8% in afactor analysis that retained 1 domain and 10 original scaleitems (Minosso, Amendola, Alvarenga, & Oliveira, 2010).
Nutritional status
This data sheet included two parts:Physical examination laboratory tests: Serum albumin,
hemoglobin, and cholesterol.Anthropometry: BMI and calf circumference.BMI was calculated as body weight in kilograms di-
vided by squared height in meters. Height was measuredto the nearest 0.1 cm using a stadiometer. When no validheight measurement could be obtained because of reasonssuch as inability to stand, researchers measured the kneeheight of the left leg using a Mediform sliding caliper with
VOL. 20, NO. 2, JUNE 2012Nutritional Status of Elderly
111
knee and ankle joints fixed at 90- angles (Wijnhoven et al.,2010). Weight was measured to the nearest 0.1 kg using achair scale (BW-130; ISO 9001 and ISO13485 approvedand CE accredited).
Researchers measured calf circumference at the point ofwidest circumference to the nearest 0.1 cm on the left leg(right leg for left-handed persons). Measurements weretaken in a seated position with knee and ankle at rightangles and feet resting on the floor (Debette et al., 2008).
In terms of anthropometric measurements, values forintraobserver technical error of measurement and coeffi-cient of reliability were .73Y.95 and .98Y.99, respectively(Ulijaszek & Kerr, 1999). The first author, a trained nurse,collected anthropometric data using a standardized pro-tocol described above. Anthropometric measures were per-formed using a nonelastic but flexible plastic tape. Dataused in analysis were the computed mean of two repeatedmeasurements.
Univariate analysis showed calf circumference as sig-nificantly correlated with other nutritional anthropometricmakers (r = .706, p G .0001, with BMI; Bonnefoy, Jauffret,Kostka, & Jusot, 2002). Intraclass correlation coefficientswere 9.99, with subjects’ mid-upper arm circumferencemeasurements (Wijnhoven et al., 2010).
ProceduresData were collected between July 2008 and August 2009.The institutional review board of the National TaipeiUniversity of Nursing and Health Sciences gave prior ap-proval to this study, and researchers obtained agreementsto participate from the eight LTCFs in Taipei City. Writteninformation explaining the study was given to LTCF chair-persons and elderly residents or their families before sign-ing the consent form. Subjects were enrolled only afterthey provided informed consent. By law, the LTCF is re-sponsible to schedule all resident health and blood exam-inations and maintain all resident records (Department ofSocial Affairs Ministry of the Interior, 2007). For residentswho died during the study year, the LTCF director calledthe individual’s family members to confirm their consentto participate before researchers contacted them to explainstudy purpose. Researchers gathered study-related datafrom medical record after family member(s) had providedinformed consent.
Data AnalysisCollected data were encoded and edited using MicrosoftExcel, and statistical analyses were performed throughSPSS 17.0 (SPSS, Inc., Chicago, IL, USA) statistical soft-ware. Statistical methods such as frequency distribution,percentage, mean, standard deviation, McNemar’s Test,independent sample t test, one-way ANOVA, logistic re-gression, and generalized linear models were used to ana-
lyze the correlation between subjects’ nutritional status andphysical function and hospitalization status. A chi-squaretest, Pearson’s productYmoment correlation, GoodmanYKruskal Tau coefficient, and KaplanYMeier method ana-lyzed subjects’ nutritional and survival status.
Results
Demographic ProfileThe 174 subjects in this study were between 67 and 105 yearsof age (mean = 82.5 years; SD = 7.3 years). Most weremen, 43.1% were over the age of 85 years, and most re-ceived nasal-gastric tube feeding. Subjects had a mean of2.1 (SD = 1.0) cardiac diseases. Most subjects (68.4%)experienced primarily chronic vascular diseases. Subjectswere on an average of 5.4 different medications (SD = 2.2),with the highest-use subject on 11 medications. Subjects’physical functions were classified based on Barthel indexscores. Mean Barthel index scores at the start and end ofthis 1-year study were 20.5 (SD = 31.7) and 17.1 (SD =28.6), respectively, identifying most of the subjects as com-pletely dependent. A total of 23 subjects (13.2%) died dur-ing the study period and were not included in the follow-updata (Table 1).
Anthropometry and Laboratory TestsAnthropometry and nutritional status data for subjects in-cluded physiological indices at the beginning (N = 174) andend (n = 151; 23 subjects died within a year) of the study.These indices included BMI, calf circumference, hemoglo-bin level, albumin level, and total cholesterol level (Table 2).Most subjects had normal BMI (Q18.5 kg/m2; study start:157 subjects [90.2%] vs. study end: 139 subjects [79.9%])and experienced decreased calf circumference (G31 cm). Meancalf circumference was 27.1 cm (SD = 4.2) and 26.7 cm(SD = 4.1), respectively, at the start and end of the study.Sixteen subjects experienced statistically significant (p = .012)decreases in calf circumference (changing from meeting thestandard to not meeting the standard). The subjects’ hemo-globin levels were normal (male: 12 g/dl, female: 10 g/dl;study begins: 133 subjects [76.4%] vs. study ends: 119 sub-jects [68.4%]). Most subjects had normal albumin levels(study begins: 132 subjects [75.9%] vs. study ends: 112 sub-jects [64.6%]). Mean total cholesterol level was also nor-mal (study begins: 160 subjects [92%] vs. study ends: 141subjects [81%]).
Physical FunctionResults showed most subjects (n = 145) had not experiencedchanges in physical function within a 1-year period, al-though eight subjects did. Four subjects changed from mod-erately dependent to severely dependent, two subjects changedfrom moderately dependent to completely dependent, and
The Journal of Nursing Research Sin-Jie Lin et al.
112
two subjects changed from severely dependent to com-pletely dependent. Statistics showed a significant difference(p = .046) in subject physical function within the year.
Hospitalization and Survival StatusIn this study, between July 1, 2008, and August 31, 2009,61 subjects (35.1%) were hospitalized. Most subjects werehospitalized once (32 subjects, 18.4%), with 19 subjectshospitalized twice and 8 subjects hospitalized thrice. Meandays of hospitalization for all subjects and those who hadbeen hospitalized during the study period were 6.86 and19.72 (SD = 12.75 and 14.73), respectively. In terms ofreasons for hospitalization, most were hospitalized be-cause of infection (57 subjects, 93.4%), and the secondlargest number of subjects were hospitalized for respira-tory diseases (11 subjects, 18%). Among the subjects hos-pitalized twice, five were hospitalized because of infectionand their hospital stays lasted for more than 20 days.Among the 174 subjects enrolled in this study, most hadsurvived the year (151 subjects, 86.8%) whereas 23 ofthem (13.2%) had died. The main cause of death wascardiopulmonary failure (9 subjects, 39.1%), followed byinfection (5 subjects, 21.7%).
Correlation Between Nutritional Status
and SurvivalAn analysis of subject survival showed survival signifi-cantly correlated with gender, physical function status, calfcircumference, and hemoglobin levels. Female mortality washigher than male mortality (15 vs. 8, p = .025). Number andtypes of medications used (r = j.15, p = .043) and physicalfunction status (r = j.15, p = .05) were negatively correlatedwith survival status. Calf circumference (#2 = 7.47, p = .003)and hemoglobin level (#2 = 5.84, p = .031) correlated positivelywith survival status. A GoodmanYKruskal Tau analysisfound that gender (G-K Tau = .03, p = .022), physical func-tion status (G-K Tau =.04, p = .026), calf circumference (G-KTau = .04, p = .006), and hemoglobin level (G-K Tau = .03,p = .016) significantly correlated with survival status (Table 2).
Physical function(start/end of study)Completely independent(ADL 100)
5/3 2.9/1.7
Mild dependent(ADL 91Y99)
1/0 0.6/0
TABLE 1.Subject Demographic Characteristics(N = 174), continued
Item n % Mean SD
Moderately dependent(ADL 61Y90)
24/21 13.8/12.1
Severely dependent(ADL 21Y60)
20/12 11.5/6.9
Completely dependent(ADL 0Y20)
124/115 71.3/66.1
Missing value 0/23 0/13.2
Note. Missing values represent subjects who died during the study period.ADL = activity of daily living.
VOL. 20, NO. 2, JUNE 2012Nutritional Status of Elderly
113
Correlation Between Feeding Patterns and
Hospitalization Status
We categorized the nutritional status of subjects into threegroups (static, increasing, and declining) based on changesover the year. Results showed feeding patterns of subjects
to correlate significantly with number of hospitalizationsduring the period. Subjects who received tube feeding werehospitalized more frequently than subjects who received oralfeeding (t = j2.77, p = .006). In addition, feeding patterncorrelated significantly to total number of hospitalizationdays during the year. Total number of hospitalization days
TABLE 2.Correlations Between Survival Status and Demographic Characteristics andNutritional Status (N = 174)
Note. BMI = body mass index.aSubject who had leg amputated.*p G .05. **p G .01.
The Journal of Nursing Research Sin-Jie Lin et al.
114
for subjects who received tube feeding were more than thatfor subjects who received oral feeding (t = j3.49, p =.001;Table 3).
Predictors of Hospitalization Status
and SurvivalWe further analyzed predictors affecting the hospitaliza-tion status of subjects. Hospitalization was used as thedependent variable whereas gender, age, eating pattern,number of chronic diseases, number, and type of medi-cations, physical function status, and nutritional status atthe start and end of the 1-year study period were used asindependent variables for stepwise logistic regressionanalysis. Among hospitalization status predictors, onlyage and eating pattern had a significant effect. Risk ofhospitalization increased 1.01 times with each 1-year in-crement in subjects’ age (95% CI: 1.06, 1.11). The risk ofhospitalization for tube-fed subjects was 1.5 times that oftheir orally fed peers (95% CI: 2.96, 5.86). Using gener-alized linear models for analysis, we found that tube feed-ing is a predictor of the total number of hospitalizationdays (p = .007, 95% CI: j10.20, j1.59). The mean totalnumber of hospitalization days for tube-fed subjects of5.9 days (SD = 2.2) was significantly more than that fororally fed subjects.
The KaplanYMeier method was used in survival anal-ysis on the variables of age, eating pattern, and nutritionalstatus, and the subjects’ mortality over the study year wascalculated. In terms of stratification variables, mortality insubjects with good nutrition (calf circumference 9 31 cm)was significantly lower than that in malnourished subjects(calf circumference G 31 cm, p = .008). As for hemoglobinlevel, mortality in subjects with good nutrition (hemoglo-bin level: male, 12; female, 10) was also significantly lowerthan that in malnourished subjects (hemoglobin level:male, G12; female, G10; p = .012; Table 4).
DiscussionDespite major improvements in elderly care, preventionof malnutrition remains a major challenge. It is vital that
healthcare institutions are aware of the issue of elderly mal-nutrition and provide sufficient training and education to allhealthcare staff (Singh et al., 2010).
Nutritional Status and Physical FunctionPrevious studies have shown that physical deteriorationaccompanied by aging directly or indirectly restricted thequantity and quality of elderly diet and caused dystrophyor malnutrition (Chen, 2007; Visvanathan et al., 2004). Inthis study, our follow-up on changes in subjects’ Barthelindex scores showed a significant change in subjects’ phys-ical function over the year. Physical function status andnutritional status were further analyzed, and correlationswere found. We identified physical function as significantlycorrelated with calf circumference and cholesterol level.Kikafunda and Lukwago (2005) randomly selected 100 oldpeople and found poorer nutritional status in subjects withpoorer physical function, disabilities, and tube feeding. Chen,Tang, Wang, and Huang (2009) performed a follow-up on306 hospitalized elderly patients and found changes in ADLto indirectly affect their nutritional status. Results of the cur-rent study are consistent with those of previous studies, sug-gesting poor nutritional status in elderly subjects with poorphysical function status.
In terms of the physical functions of subjects enrolled inthis study, most were completely dependent. This suggestedthat subjects living in LTCFs generally needed assistance forphysical activities. As for relevant studies in Taiwan, Lai(2007) conducted a study on a specific veteran’s nursinghome and found 73 of 160 subjects (45.6%) as completelydependent. Shyu et al. (1993) conducted a study of 143elderly residents living in 37 LTCFs and found that 36.4%were completely dependent and reliant on assistance fornutritional intake.
Volkert and colleagues (2011) found 57.7% of tube-fedresidents in nursing home to be malnourished. In this study,the feeding pattern of subjects was primarily by tube feed-ing. Further analysis indicated eating pattern as correlatedto nutritional status and significantly correlated to calf cir-cumference, albumin level, and total cholesterol level. Lin(2000) and Su (2005) indicated in their studies on LTCFs
TABLE 3.
Correlation Between Eating Patterns and Hospitalization Status During the StudyPeriod (N = 174)
Times of Hospitalization Total Number of Hospitalization Days
VOL. 20, NO. 2, JUNE 2012Nutritional Status of Elderly
115
and nursing homes in central and southern Taiwan that oldpeople experienced a high level of malnutrition. In addition,the overall nutritional status of orally fed old people wasbetter. Huang et al. (2005) investigated 1,385 hospitalizedtube-fed patients and found that blood albumin of 72.8%of subjects was lower than 3.5 mg/dl and the proportion ofmalnourished tube-fed patients was higher than the overallhospitalized patient population. Kikafunda and Lukwago(2005) selected 105,100 adults aged 60Y90 years as sub-jects and found that as many as 33% had difficulty eatingand experienced disabilities related to malnutrition.
Portero-Mclellan and colleagues (2010) also found a pos-itive correlation between patient calf circumference and nu-tritional status, indicating this measurement as a potentialcomplementary tool for monitoring elderly inpatient nutri-tional status. Our study found decreased calf circumferenceas the most significant symptom of malnutrition in ourstudy population. Further analysis found eating pattern,physical function status, and calf circumference as signifi-cantly correlated with one another. Cuervo, Garcia, et al.(2009) used a nutritional assessment scale to screen 22,007
old people living in communities and found calf circumfer-ence as an important index of malnutrition. Lai (2007) in-terviewed 160 veteran servicemen and found that as manyas 80% of residents experienced malnutrition or were atrisk of malnutrition, with calf circumference as the primarynutritional status predictor. Chiu et al. (2005) collected in-formation on 280 LTCF residents and found eating patternand physical function to be predictors of nutritional status.In this study, most tube-fed old people exhibited a higherlevel of dependence. Subjects needed to rely on assistancefor feeding, which resulted in unbalanced nutrient intake.Moreover, subjects might experience muscular dystrophy anddegeneration characterized by gradually decreasing calf cir-cumference due to being bedridden and inability to walk in-dependently, which also further enhanced malnutrition risks.
Correlation Between Nutritional and
Hospitalization StatusesVolkert and colleagues (2011) found that malnutrition sig-nificantly related to nausea/vomiting, constipation, pressure
TABLE 4.
Survival Analysis of Subjects’ Demographic Characteristics and NutritionalStatus (N = 174)
Variable Alive Dead n Log-Rank Test p
Age 0.88 .64665Y74 years 27 3 3075Y84 years 61 8 69Q85 years 63 12 75
Note. BMI = body mass index.aSubject who had leg amputated.*p G .05. **p G .01.
The Journal of Nursing Research Sin-Jie Lin et al.
116
ulcers, dehydration, infections, antibiotic use, and hospital-ization. In this study, we also found subjects’ eating pat-terns significantly correlated with hospitalization status overthe year. Admission frequencies and total number of hos-pitalization days were greater for tube-fed than orally fedsubjects. Although this study was unable to verify a directcorrelation between nutritional and hospitalization statuses,eating pattern was correlated with nutritional status as wellas calf circumference, albumin level, and total cholesterollevel. This infers nutritional status, albumin level, and totalcholesterol level as indirectly correlated with hospitalizationstatus over the year. Seiler (2001) found that old people whowere hospitalized, living in nursing homes, or living at hometended to experience malnutrition and that malnutritionincreased admission frequency. Stratton and Elia (2006)screened the nutritional status of 1,000 old people and foundthat as many as 42% faced a high risk of malnutrition dur-ing hospitalization. Malnutrition was significantly correlatedwith prolonged hospital stays. Correia and Waitzberg (2003)found the average number of hospitalized days for adultswith poor nutritional status to be longer than those withgood nutritional status (16.7 vs. 10.1 days).
Previous studies have indicated that tube feeding leadsto aspiration pneumonia (Metheny, 2006; Metheny, Stewart,& McClave, 2011). Most of our subjects were tube fed. Themain reason for hospitalization was pneumonia, and wefound tube feeding as correlated with hospitalization status.Health professionals should further examine this phenome-non and carefully assess nutrition and health statuses whencaring for tube-fed individuals.
Relationship Between Nutritional and
Survival StatusesThis study found gender, physical function status, calf cir-cumference, and hemoglobin level to be significantly cor-related with survival. While female deaths outnumberedmale deaths in our subject population, a relationship be-tween gender and survival could not be verified as malesubjects outnumbered female subjects (99:75). This ratiodiffered significantly from most of other similar studies,which included more female subjects than male subjects(Carriere, Dupuy, Lacrous, Cristol, & Pelcourt, 2008;Harris, Davies, Ward, & Haboubi, 2008; Portero-Mclellanet al., 2010). Number and type of medications used andphysical function status were negatively correlated with sur-vival. The uneven distribution of number and type of med-ications used and physical function status resulted inincreased subjects’ deaths in those taking three medicationsas compared with those who were completely independentor only mildly dependent. As such, correlations among num-ber and type of medications used, physical function status,and survival could not be verified by this study.
Analysis of subjects’ nutritional status and survival re-vealed calf circumference and poor nutritional status as mea-sured by decreased hemoglobin level to correlate positively
with survival. This finding was consistent with relevant stud-ies. Some studies have indicated that calf circumference is animportant predictor of nutritional status (Lai, 2007; Cuervo,Ansorena, et al., 2009). Decreased hemoglobin level impliedinsufficient protein intake, which leads to greater risk ofmalnutrition-related mortality (Fontaine & Raynaud-Simon,2008). Correia and Waitzberg (2003) found mortality amongmalnourished elderly subjects to be 2.63 times that of sub-jects with good nutrition. Drame et al. (2009) conducted astudy on 1,306 elderly patients hospitalized after emergencyroom visits and found that malnutrition correlated with mor-tality within 3 years. In this study, most subjects with de-creased calf circumference were also severely dependent andprimarily tube fed. Decreased hemoglobin level was corre-lated with insufficient protein intake. This infers that severedependence leads to muscular dystrophy and degeneration,which further increases the risk of decreased calf circumfer-ence. Subjects who failed to take food by themselves due tosevere dependence had less nutrient intake, increased mal-nutrition, and increased mortality.
Predictive Power of Nutritional Status on
Hospitalization and MortalityIn this study, the predictors of subjects’ hospitalizationwere primarily age and eating pattern. Every increase of1 year in age increased subjects’ risk of hospitalization by1.01 times. Risk of hospitalization for tube-fed subjectswas 1.5 times higher than that for their orally fed peers.Total number of hospitalization days for tube-fed subjectsduring the study year were higher than that of orally fedsubjects, indicating that tube feeding may be an importantpredictor of hospitalization. In this study, eating patterncorrelated significantly with calf circumference, albuminlevel, and total cholesterol level, inferring that these threevariables were indirect predictors of hospitalization. Sorensenet al. (2008) assessed the nutritional status of 5,051 subjectsenrolled at 26 hospitals in 12 countries and found that el-derly participants aged Q70 years had more days of hos-pitalization. In terms of predictors of survival for elderlysubjects living in LTCFs, the mortality rate of subjects withnormal calf circumference was less than that of subjectswith decreased calf circumference. Mortality in subjects withnormal hemoglobin levels was lower than that in subjectswith decreased levels. Yang’s (2004) investigation of mortal-ity and nutritional status among 276 tube-fed patients foundhemoglobin level as a significant predictor of mortality.Sorensen et al. (2008) also found that both mortality anddays of hospitalization were higher in malnourished thannonmalnourished subjects.
LimitationsData in this study were collected from medical chart re-cords. Although all facilities used the same brand chair
VOL. 20, NO. 2, JUNE 2012Nutritional Status of Elderly
117
scale, the potential for interfacility measurement error re-mains. Also, researchers did not perform the knee lengthestimates of height, calf circumference measurements, andphysical function Barthel index assessments, which mayintroduce inconsistencies in obtained data.
Most subjects in this study were completely dependent.Measuring knee length was complicated because of stifflimbs and joint contraction, which may have affected BMIdata accuracy. This study was designed as a retrospectivestudy, and it was difficult to obtain informed consent toaccess data on subjects who had died. The number of sub-jects who died may be underestimated, which could affectoverall results. In addition, based on survival analysis se-lection, when comparing the nutritional status of subjectswho are still alive after a year with that of those who diedwithin that year, we found that the former had better nu-tritional status at the points in the year immediately beforethe death of subjects in the latter category. Thus, resultsshould only be applicable for predictions related to elderlyLTCF resident alive more than 1 year after enrollment.
ConclusionsThe subjects enrolled in this study were primarily tube fed,experienced multiple diseases, were completely dependentin terms of physical function, and were concurrently tak-ing multiple medications. Physical function status for sub-jects was largely ‘‘completely dependent.’’
Care in applying tube feeding is an issue worth furtherexamination and consideration. Results of the currentstudy show hospitalization over the year to correlate withtube-feeding status. Aspiration pneumonia was found tobe the most common cause for hospitalization of tube-fedsubjects. Calf circumference and hemoglobin level corre-lated positively with survival status. Decreased hemoglo-bin level correlated with insufficient protein intake. Furtherinvestigation into proper nutrition protocols for elderly res-idents of LTCFs is recommended. There is little data on calfcircumferences in Taiwan’s elderly population, and dataused in most Taiwan studies drew primarily on foreignstudy data. We thus recommend a stronger effort in Taiwanto collect representative samples that can be used in furtheranalyses and to establish culture-specific data that enablescalf circumference to be used as a simple and convenient in-dex for measuring nutritional status.
Relevance to Clinical PracticeFor those in the healthcare profession, results of the cur-rent study suggest the following: (a) Subjects with poorerphysical function status exhibited decreased calf circum-ference and malnutrition. Maintenance of physical func-tion also correlated closely with the maintenance of subjects’nutritional status. Therefore, we advise care facilities tostrengthen rehabilitation exercise programs for residents,arrange regular rehabilitation therapist visits, and providerehabilitation assistance. (b) Calf circumference, hemoglo-
bin level, albumin level, and cholesterol level were impor-tant predictors of hospitalization and death. It is advisedthat, in addition to regular follow-up on health examinationhematological reports, care facilities should also increasefollow-up on calf circumference to reduce malnutrition-relatedhospitalization and death risks and medical costs. (c) Mostresidents in the study were tube fed. LTCF staffs are advisedto understand how aggressive intervention using nutritionalsupport and improvement measures can help balance andstabilize residents’ nutritional intake. Moreover, nutrition-related in-service training for relevant staff should be strength-ened to increase the quality of resident care. (d) This studyfound that tube feeding correlated with hospitalization. Werecommend convening discussions involving professionals,family members of elderly residents, and LTCF caregiversto consider the benefits of employing other eating patternssuch as gastrostomy to decrease aspiration pneumonia-related hospitalization incidences. (e) LTCF nursing staffsmust be professional and independent. Nursing educationshould strengthen nursing staff assessment of residents’ nu-tritional statuses to increase sensitivity to identify residentsat high risk of malnutrition and execute timely preventionand treatment. Moreover, strengthening interteam cooper-ation is necessary.
ReferencesAtalay, B. G., Yagmur, C., Nursal, T. Z., Atalay, H., & Noyan, T.
(2008). Use of subjective global assessment and clinical
outcomes in critically ill geriatric patients receiving nutrition
support. Journal of Parenteral and Enteral Nutrition, 32(4),
454Y459.
Bonnefoy, M., Jauffret, M., Kostka, T., & Jusot, J. F. (2002).
Usefulness of calf circumference measurement in assessing
the nutritional state of hospitalized elderly people. Geron-
tology, 48(3), 162Y169.
Carriere, I., Dupuy, A. M., Lacroux, A., Cristol, J. P., & Delcourt, C.
(2008). Biomarkers of inflammation and malnutrition associ-
ated with early death in healthy elderly people. Journal of
American Geriatrics Society, 56(5), 840Y846.
Chan, L. C. (2000). Association between nutritional status and
medical services utilization of the elderly in the northern
Taiwan (Unpublished master’s thesis). National Defense
Medical Center, Taipei City, Taiwan, ROC. (Original work pub-
lished in Chinese)
Chen, C. C., Tang, S. T., Wang, C., & Huang, G. H. (2009).
Trajectory and determinants of nutritional health in older
patients during and six-month post-hospitalisation. Journal
of Clinical Nursing, 18(23), 3299Y3307.
Chen, P. S. (2007). Nutrition status assessment of homebound
and institutionalized elderly in Chiayi area (Unpublished
master’s thesis). Chia Nan University of Pharmacy and
Science, Tainan City, Taiwan, ROC. (Original work published
in Chinese)
Chiu, Y.W., Lin, W. Y., Hsieh, P. C., Li, C. I., & Chiou, C. H. (2005). A
survey to identify effective nutritional assessment indicators
in long-term care facilities. Journal of Evidence-Based Nurs-
ing, 1(2), 112Y122. (Original work published in Chinese)
The Journal of Nursing Research Sin-Jie Lin et al.
118
Correia, M. I., & Waitzberg, D. L. (2003). The impact of
malnutrition on morbidity, mortality, length of hospital stay
and costs evaluated through a multivariate model analysis.