Multivitamin supplementation improves nutritional status and ... - Nature

5 downloads 139 Views 144KB Size Report
Nov 28, 2007 - Ascorbic acid (vitamin C). 75 mg ..... intramuscular injection of a megadose of cholecalciferol for ... Grieger J, Nowson C, Ackland ML (2007).
European Journal of Clinical Nutrition (2009) 63, 558–565

& 2009 Macmillan Publishers Limited All rights reserved 0954-3007/09 $32.00 www.nature.com/ejcn

ORIGINAL ARTICLE

Multivitamin supplementation improves nutritional status and bone quality in aged care residents JA Grieger1, CA Nowson1, HF Jarman1, R Malon2 and LM Ackland3 1 School of Exercise and Nutrition Sciences, Deakin University, Burwood, Victoria, Australia; 2Division of Rehabilitation Services and Aged Care, Barwon Health, Victoria, Australia and 3School of Biological and Chemical Sciences, Deakin University, Burwood, Victoria, Australia

Objectives: To assess the effectiveness of a multivitamin (MV) tablet on nutritional status, quantitative heel ultrasound (QUS), mobility, muscle strength and falls. The design comprised two groups matched on mobility levels, randomized to receive a daily MV or placebo (P) tablet for 6 months. The setting was an Australian residential care facility. Subjects: A total of 92 aged care residents. Serum micronutrients, body weight, QUS, rate of falls, hand grip strength, and the timed up and go test were assessed at baseline and 6 months. Results: A total of 49 participants consumed a MV and 43, a matched P for 6 months. There was a greater increase in the MV vs P group for serum 25(OH)D (mean difference±standard error, 33.4±2.6 nmol l1), folate (13.4±2.8 nmol l1), and vitamin B12 (178.0±40.3 pmol l1) (all Po0.001). Adequate 25(OH)D concentrations (X50 nmol l1) were found among 77% of participants in the MV group vs 10% taking P (Po0.001). Adjusting for baseline levels, the increase in QUS was greater in the MV vs P group (3.0±2.0 dB MHz1 vs 2.9±2.1 dB MHz1, respectively, P ¼ 0.041). There was a trend towards a 63% lower mean number of falls in the MV vs P group (0.3±0.1 falls vs 0.8±0.3 falls, P ¼ 0.078). Conclusions: MV supplementation raised serum vitamin B12 and folate concentrations and increased serum 25(OH)D, which was accompanied by an apparent positive effect on bone density. We also found a trend towards a reduction in falls and this could contribute to a reduction in fractures.

European Journal of Clinical Nutrition (2009) 63, 558–565; doi:10.1038/sj.ejcn.1602963; published online 28 November 2007 Keywords: multivitamin; residential care; 25(OH)D; heel ultrasound

Introduction The elderly population in residential care settings have a poor dietary intake (Barr et al., 1984; Nowson et al., 2003; Grieger and Nowson, 2007). This may be attributed to a low

Correspondence: Professor CA Nowson or Dr JA Grieger, School of Exercise and Nutrition Sciences, Deakin University, 221 Burwood Highway, Burwood, Victoria, 3125, Australia. E-mails: [email protected], [email protected] Contributors: CAN was the chief investigator for this study whose major role was initiating the trial, designing the study objectives and specifying the best method for data collection. CAN also provided support with the statistical analyses, data interpretation and the compilation of the completed manuscript. JAG was the PhD student involved in acquisition of subjects, data collection, statistical analyses, interpreting results and the compilation of this manuscript. HFJ and RM were the co-investigators of the project, contributing to the initial development and design of the study objectives, and reviewing of the manuscript. LMA contributed in the measurement of serum zinc and reviewing the manuscript. Received 21 December 2006; revised 4 April 2007; accepted 20 September 2007; published online 28 November 2007

energy intake, reduced physical activity levels, dementia or depression. Moreover, the elderly have higher daily requirements for nutrients such as protein, calcium and vitamin D compared with younger populations (NHMRC, 2006). These higher nutritional demands, combined with lower total energy intakes, make it difficult for the elderly to meet the recommended micronutrient intakes from dietary sources alone. To promote good health, many people worldwide take various forms of multivitamin (MV) supplements (Bender, 2002; Fawzi and Stampfer, 2003). In elderly community dwellers, daily consumption of a MV/mineral for 2 or 4 months has improved various serum micronutrient concentrations (Mann et al., 1987; McKay et al., 2000a, b); however, following 1-year consumption, no reduction in clinical infections or quality of life was found (Avenell et al., 2005). No studies were found that have documented the impact of a daily MV on functional status (for example, muscle strength and mobility); and to our knowledge, there has been no long-term placebo (P)-controlled, MV supplementation studies among

Multivitamin supplementation in residential aged care JA Grieger et al

559 aged care residents who are likely to be at greater risk for nutritional deficiencies and poor functional status. Some studies however have supplemented elderly groups with one or two nutrients. In a randomized, P-controlled, double-blind trial for 2 years, among Australian aged care residents, vitamin D supplementation (ergocalciferol, initially 250 mg given once weekly and then 25 mg daily) in combination with 600 mg of calcium, reduced the rates of falls (Flicker et al., 2005); and among 84-year-old, ambulatory women, supplementation with 20 mg vitamin D3 and 1200 mg calcium for 18 months, reduced hip fractures by 43% and total non-vertebral fractures by 32% (Chapuy et al., 1992). In a randomized, double blind, controlled trial, supplementation with folate (5 mg) and vitamin B12 (1500 mg) has been found to reduce hip fracture rates in elderly stroke patients (Sato et al., 2005). As the elderly appear to be at risk for many micronutrient deficiencies, our aim was to assess the effectiveness of a daily MV for 6 months on nutritional status, bone quality and muscle strength.

Subjects and methods Study design The study utilized a double blind, randomized, P-controlled study for 6 months. Initially, it was intended to be a 2  2 factorial design, utilizing fortified milk vs usual milk, where subjects were stratified by mobility for milk allocation; and MV vs P supplementation. The fortified milk was specifically developed by Murray Goulburn Co-operative Ltd, and was designed to contain higher quantities of protein, calcium and vitamin D than usual milk, but similar energy and fat content. However, due to problems with the distribution of the milk to subjects within the facility, the use of fortified milk ceased completely at week 16 of the 6-month study. The use of fortified milk over this short-time period resulted in no significant changes in macro or micronutrient intakes. Importantly, there was no difference in nutrient intake between the P and MV group. Therefore, the following results were analysed between the P and MV group. All subjects, irrespective of ward/hostel, were randomized (Random number generator, Microsoft Excel, 2002, by CN) to receive the daily MV, which is a readily available off-theshelf MV supplement ‘Heron Women’s multivitamin’, or a placebo tablet (Sigma Company Ltd, Australia) in a double blind fashion. The composition of the MV reflects the labelled content which was confirmed by the manufacturer’s assayed composition of the batch used (Table 1). MV and P tablets were administered by the nursing staff during the morning medication round (0800 hours) with some subjects in low care facilities choosing to self-medicate. The main components of the P were starch maize, cellulose and magnesium stearate. This project was approved by the Barwon Health Research and Ethics Advisory Committee and Deakin University Human Research Ethics Committee.

Table 1

Composition of the multivitamin supplement Compositiona

Ascorbic acid (vitamin C) b-Carotene (retinol) Biotin (vitamin H) Calcium pantothenate (pantothenic acid, vitamin B5) Cholecalciferol (vitamin D3) Choline bitartrate Cyanocobalamin (vitamin B12) D-a-Tocopheryl acid succinate (vitamin E) Folic acid Inositol Nicotinamide Pyridoxine hydrochloride (vitamin B6) Riboflavin Thiamine hydrochloride (vitamin B1) Calcium carbonate (elemental calcium: 144 mg) Ferrous fumarate (iron: 5 mg) Magnesium oxide (magnesium: 75 mg) Manganese amino acid chelate (manganese: 750 mg) Zinc amino acid chelate (zinc: 6 mg) Potassium sulphate (potassium: 1.5 mg) Bioflavonoids Siberian ginseng a

75 mg 3 mg 100 mg 35 mg 10 mg 7.9 mg 25 mg 12.2 mg 200 mg 8 mg 50 mg 25 mg 10 mg 15 mg 360 mg 15.2 mg 125 mg 7.5 mg 30 mg 3.4 mg 25 mg 10 mg

Confirmed by manufacturer assay.

Subjects Subjects were recruited from Barwon Health, Victoria’s largest, regional health care provider. It provides 758 inpatient beds; with the Aged Care facility providing 260 high-level care (HLC) nursing home beds and 106 low-level care (LLC) hostel beds. Residents in the dementia, rehabilitation and palliative care wards were excluded from the study. All subjects who were able to consume food orally were eligible to participate in the study (LLC: n ¼ 106; HLC: n ¼ 228).

Study measurements Information on dietary requirements, mobility levels, medication usage and medical diagnoses were obtained from the medical records. Body weight was measured (monthly) by the nursing staff.

Dietary intake Dietary intakes were assessed on three separate occasions for a 24 h period (month 2, 4 and 6), where plate waste was visually assessed (Nowson et al., 2003) at all main meals and mid-meals. The aim was to assess each subject once over the study period. Energy and nutrient intakes were calculated using the dietary analysis computer package Food Works, version 4 (Xyris Software; Brisbane, Australia). Values for vitamin D content were added to the database using data European Journal of Clinical Nutrition

Multivitamin supplementation in residential aged care JA Grieger et al

560 from British Food Composition (Paul and Southgate, 1979; Fraser, 1990).

Falls incidence Medical diagnosis of osteoporosis and having a history of falls were collected from the medical records at the beginning of the study. For each subject, falls incidence was documented over the 6-month study period by the nursing staff and recorded in the medical history. Medical histories of all subjects were viewed by research staff at the end of the study and the data on falls was extracted.

Anthropometry, QUS and functional status Stature height was calculated by measuring knee height (cm), quantitative heel ultrasound (QUS) was measured using the broadband ultrasonic attenuation value (dB per MHz), the timed up and go test (TUG) and hand grip strength (HGS) were assessed using standard methodological approaches described previously (Grieger et al., 2007).

Biochemistry After an overnight fast of at least 8 h, blood samples (10 ml) were collected by venipuncture at baseline and 6 months for measurement of serum albumin, zinc, vitamin B12, folate and 25(OH)D concentrations. Serum albumin was assayed using a Randox Daytona automated clinical chemistry analyzer (Antrim, UK, 2002). Radioimmunoassays were used to measure serum 25(OH)D (DiaSorin Inc., Stillwater, Minnesota, USA), vitamin B12 and folate (BioRad Laboratories, NSW, Australia) concentrations. The Intra assay CV for 25(OH)D was 8.8%; for vitamin B12 was 13.5% and for folate was 9.2%. Serum zinc was measured using flame atomic absorption spectrophotometry by direct aspiration (Varian SpectrAA-800). The criteria for defining deficiency levels for albumin (Abbasi and Rudman, 1993; Alhamdan, 2004), 25(OH)D (Diamond et al., 2005), vitamin B12 (Clarke et al., 2004), folate (Clarke et al., 2004) and zinc (Pepersack et al., 2001) have been described previously.

Compliance Following tablet consumption, a confirmatory signature was made by the nursing staff in the medication history. Tablet compliance was assessed by researchers at the end of the study by counting the signatures in each subject’s medical histories. A total of 10 subjects from LLC were selfmedicated.

Statistical analysis Statistical analysis was performed using SPSS for Windows (SPSS Inc., Chicago, IL, USA, Release 11.5.2.1). Student’s t-tests were used to identify differences in continuous European Journal of Clinical Nutrition

variables between groups. w2-tests were used to analyse categorical data. The univariate general linear model was used when comparing absolute and percent changes between the groups over time. Data that was skewed was log 10 transformed (serum folate, 25(OH)D and grip hand strength). Log 10 data was used for all analyses; however, the geometric means were reported in the text.

Results A total of 122 subjects gave informed consent, or if unable, their next of kin gave proxy consent (Figure 1). Two subjects withdrew and five died prior to study commencement. A total of 115 subjects commenced the study. There were no differences in baseline characteristics between the MV and P groups (Table 2).

Tablet supplementation A total of 57 and 58 subjects were randomly assigned the P and MV tablet, respectively; however, eight subjects were excluded as they were already found to be taking MVs, and two subjects refused to take the MV/P tablet. Withdrawal from the study included seven deaths (four in the P group, three in the MV group); three subjects taking the P suffered side effects (rash/vertigo; behavioral issues and indigestion); and three subjects were transferred to an external facility. Therefore, 49 and 43 subjects completed the study from the MV and P group, respectively. At the end of the study, quantitative heel ultrasound (QUS) could not be measured in 28 subjects: 24 subjects were bed bound, too frail, or could not position their foot correctly and three were non compliant. One subject was excluded from QUS analysis due to abnormally low results (o2.5 s.d. below the mean). Three subjects did not provide serum samples. At the beginning and at the end of the study, seven subjects, each in the P and MV group, took a calcium þ vitamin D supplement; 8 and 12 subjects took a calcium supplement; and four and nine subjects took a vitamin D supplement in the P and MV groups, respectively. The range of daily calcium and vitamin D intake from these supplements was 600–1000 mg and 0.25–25 mg, respectively; this supplement use was maintained throughout the study.

Tablet compliance Participants consumed on average 82% of the tablets (range: 79–100%) in the P group and 91% (range 76–100%) in the MV group (P40.05). The tablets were consumed for a minimum of 19, and a maximum of 22 weeks.

Dietary intakes Dietary intake was measured on 39 (80%) subjects in the MV group and 38 (88%) subjects in the P group. There was no

Multivitamin supplementation in residential aged care JA Grieger et al

561 Subjects assessed for eligibility (n=334) High level care residents (n=228) Low level care residents (n=106) Expressed interest in study (n=122) Residents consent (n=55) Next of kin consent (n=67) Withdrew consent (n=2) Deceased before study commencement (n=5)

Randomized (n=115) High level care residents (n=85) Low level care residents (n=30)

Allocated Multivitamin (n=58) Received allocated Multivitamin (n=54)

Allocated Placebo (n=57) Received allocated Placebo (n=51)

Did not receive allocated Multivitamin (n=4) 1 was taking iron, 2 were taking multivitamins; 1 refused

Did not receive allocated Placebo (n=6) 5 were taking multivitamins, 1 refused

Lost to follow-up (n=8) Death (n=4), moved facility (n=1), side effects (n=3)

Lost to follow-up (n=5) Death (n=3), moved facility (n=2)

Completed Placebo (n=43)

Completed Multivitamin (n=49) Figure 1

Flow chart indicating recruitment numbers, randomization and subjects analysed.

Table 2 Frequency of baseline characteristics between the study groups

Males/females Mobility Immobile With assistance Independent Medications Bisphosphonates Calcium þ vitamin D Calcium only Vitamin D only

Placebo n

Multivitamin n

13/30

19/30

19 12 12

24 11 14

2 7 8 4

4 7 12 9

difference in dietary intakes between the MV and P groups assessed during the study. The mean (±s.e.m.) daily energy, calcium and folate intakes from dietary sources for the P (n ¼ 38) and MV (n ¼ 39) sample, respectively, were

6.4±0.3 MJ vs 6.8±0.3 MJ; 804±54 mg vs 931±70 mg; and 295±23 mg vs 237±19 mg, all P40.05). The median (inter-quartile) range for vitamin D intake from dietary sources for the P and MV group was 2.1 (0.3) mg vs 2.6 (0.4) mg, respectively.

Nutritional status At the end of the study, there was a greater rise in the MV group compared to the P group for serum 25(OH)D (33.4.±2.6 nmol l1), folate (13.4±2.8 nmol l1) and vitamin B12 (178.0±40.3 pmol l1) (all Po0.001) (Table 3). The percentage differences in serum micronutrients between the groups are illustrated in Figure 2. Over the study period, the mean (s.d.) daily vitamin D intake was 9.7 (0.6) mg (range: 6.8–10.0 mg). For those consuming a vitamin D supplement in addition to the MV tablet (n ¼ 10, 21%), the mean (±s.e.m.) vitamin D intake was 30.0±3.9 mg. At the start of the study, those consuming a vitamin D supplement had 77% higher serum 25(OH)D European Journal of Clinical Nutrition

Multivitamin supplementation in residential aged care JA Grieger et al

562 Table 3 Mean (±s.e.m.) characteristics at baseline and the change between the multivitamin and placebo groups Placebo (n ¼ 43)

Body weight (kg) BMI (kg m–2) QUS (dB MHz–1) HGS (kg grip force) TUG (s) 25(OH)D (nmol l–1) Albumin (g l–1) Zinc (mmol l–1) Folate (nmol l–1) Vitamin B12 (pmol l–1)

P-valuec

Multivitamin (n ¼ 49)

n

Meana

Changeb

n

Meana

Changeb

43 43 34 32 20 42 42 42 42 42

63.6±2.2 25.4±0.7 42.0±3.5 27.6±3.1 40.0±3.8 35.5±2.5 38.8±0.5 10.6±0.3 19.0±2.0 296.0±25.0

0.9±0.9 0.4±0.4 2.9±2.1 15.7±1.4 1.1±2.9 6.0±2.1 1.3±0.5 0.1±0.4 0.4±2.3 32.4±30.2

49 47 38 34 20 48 48 48 48 47

68.0±2.1 26.5±0.8 47.7±4.0 31.4±2.7 30.1±4.2 35.7±2.8 38.9±0.4 11.6±0.5 14.8±1.4 273.8±17.8

0.41±0.6 0.23±0.3 þ 3.0±2.0 13.5±1.4 3.8±2.9 þ 27.4±3.9 1.2±0.4 0.1±0.5 þ 13.0±1.6 þ 145.6±33.1

0.476 0.532 0.041 0.282 0.520 o0.001 0.764 0.111 o0.001 o0.001

Abbreviations: BMI, body mass index; HGS, hand grip strength; QUS, quantitative heel ultrasound; TUG, timed up and go. a Mean±s.e.m. characteristics at baseline. b Adjusted for baseline values (univariate general linear model). c Differences in the mean change between groups (mean absolute values, t-test).

Percent change in serum micronutrient

120% Placebo Multivitamin

100%

Table 4 Prevalence of serum/borderline deficiencies at baseline and at 6 months Placebo

80%

Multivitamin

60%

Baseline

6 months

Baseline

6 months

40%

n

%

n

%

n

%

n

%

25(OH)D p50 nmol l1 450 nmol l1

35 7

83 17

38 4

90 10*

37 11

77 23

11 36

23 77

Folate p7 nmol l1 47 nmol l1

4 28

13 87

9 33

21* 79

6 36

14 86

0 47

0 100

Vitamin B12 p200 pmol l1 4200 pmol l1

10 22

31 69

17 25

40* 60

13 30

30 70

3 44

6 94

20% 0%

¹

¹

²

²

-20% -40% 25(OH)D 1

Vitamin B12

Folate

Zinc

Albumin

2

Significantly lower vs multivitamin (P