Vitamin D supplementation in older adults: Searching for specific guidelines in nursing homes

Abstract

Background

The prevalence of vitamin D insufficiency is very high in the nursing home (NH) population. Paradoxically, vitamin D insufficiency is rarely treated despite of strong clinical evidence and recommendations for supplementation. This review aims at reporting the current knowledge of vitamin D supplementation in NH and proposing recommendations adapted to the specificities of this institutional setting.

Design

Current literature on vitamin D supplementation for NH residents was narratively presented and discussed by the French Group of Geriatrics and Nutrition.

Result

Vitamin D supplementation is a safe and well-tolerated treatment. Most residents in NH have vitamin D insufficiency, and would benefit from vitamin D supplement. However, only few residents are actually treated. Current specific and personalized protocols for vitamin D supplementation may not be practical for use in NH settings (e.g., assessment of serum vitamin D concentrations before and after supplementation). Therefore, our group proposes a model of intervention based on the systematic supplementation of vitamin D (1,000 IU/day) since the patient's admission to the NH and throughout his/her stay without the need of a preliminary evaluation of the baseline levels. Calcium should be prescribed only in case of poor dietary calcium intake.

Conclusion

A population-based rather than individual-based approach may probably improve the management of vitamin D insufficiency in the older population living in NH, without increasing the risks of adverse health problems. The clinical relevance and cost effectiveness of this proposal should be assessed under NH real-world conditions to establish its feasibility.

Introduction

It has been estimated that 25% of the older person aged 85 and over live in nursing home (NH) in France (1). The mean age of NH residents is about 86 years. Most of them are dependent in basic activities of daily living, and present multiple comorbidities and polypharmacy. One out of five of NH residents dies during his/her first year of stay (2). Consistently, rates of hospitalization in NH residents are also very high (3). Although the functioning and characteristics of NHs in the whole health system vary across countries, similar profiles of residents (i.e., individuals characterized by an increased vulnerability of health status to stressors) are reported worldwide (1).

In NH, vitamin D supplementation is rarely considered as a health priority (2, 4, 5). Despite of the very high rates of falls and fractures in NH residents (2), osteoporosis is largely underdiagnosed and often remains untreated. Nevertheless, the risk of hip fractures has been estimated at 3.7 to 5.0 per 100 residents per year. This risk is about 2.5- to 10-fold greater than that measured in community-dwelling populations (6., 7., 8.). A fourth of the fractures over the age of 75 years occurs in NH (9). In the last two decades, numerous intervention studies and several meta-analyses (10., 11., 12., 13., 14., 15.) have repeatedly reported that vitamin D supplementation can reduce the risk of falls and fractures in older people including NH residents.

Vitamin D is a cheap treatment for which major clinically relevant benefits have been reported. Moreover, it is associated with low risk of adverse events (10, 16). There is no population with more convincing arguments to supplement with vitamin D than NH residents (17). For physicians working in NHs, the main challenge is to preserve functional ability and quality of life of residents who have multiple and complex comorbidities, polypharmacy, and reduced life expectancy. However, to reach these goals, clinical practices in NH currently rely more on empiricism (and sometimes even on fatalism) than on evidence-based medicine. Only 2% of the research published in older populations concerns NH residents (1). Therefore, our knowledge about the specific health care to provide for the growing population of this institutional setting is still largely lacking.

The aim of the present manuscript is to critically review articles that investigated the use of vitamin D supplementation in older persons, in particular those living in NH. Another purpose is to promote the use of vitamin D supplementation in NHs. We expect that the present review will increase physicians’ awareness about the importance of vitamin D supplementation in the care of older residents living in NHs. This work was conducted on the behalf of the French Group of Geriatrics and Nutrition composed by experts in the fields of nutrition, gerontology, and geriatrics.

Prevalence of vitamin D deficiency in NH

Although different cut-points have been provided, vitamin D levels below 50 nmol/L (or 20 ng/mL), 25 nmol/L (or 10 ng/mL), and 12.5 nmol/L (or 5 ng/mL) are generally considered to define vitamin D insufficiency, deficiency, and severe deficiency, respectively (18). Epidemiological studies have reported that many older people, especially those living in NH, present vitamin D deficiency (18., 19., 20., 21., 22., 23., 24., 25., 26., 27., 28., 29., 30., 31., 32., 33.). In fact, the vitamin D insufficiency has even reached the 100% prevalence in some studies in institutionalized older persons (34). Consistent data have been reported from different countries worldwide in NH populations (18., 19., 20., 21., 22., 23., 24., 25., 26., 27., 28., 29., 30., 31., 32., 33.) (See Table I for a summary of 25(OH)-vitamin D levels in nursing homes). The age-related reduction of cholecalciferol production in the skin, the limited exposure to direct sunlight, and the inadequate nutritional intake represents the main reasons for such phenomenon in NH (35).

The high prevalence of vitamin D insufficiency is widely reported and stable since decades in spite of education efforts and expert recommendations proposing the increase of vitamin D intake in NH (36). For example, mean 25(OH)-vitamin D levels of a sample of homebound subjects living in NH or in the community was 33.25 nmol/L (or 13.3 ng/mL) in the United States (37). Other studies reported that about 60% of NH residents presents a level of 25(OH)-vitamin D below 50 nmol/L (or 20 ng/mL) (38, 39). In a very recent study conducted in Austria, Piltz et al. showed that 92.8% of the 961 residents from 95 NH had vitamin D insufficiency (40). In the Netherlands, vitamin D level was found to be insufficient in 98% of the NH residents (27). Even in Australia or in the Southern Greece, countries with high sunlight exposure, similar prevalence of vitamin D insufficiency was reported among NH residents (26, 32, 41, 42). These findings suggest that a population-based vitamin D supplementation should be carefully considered in this population. Levels below 25 nmol/L (or 10 ng/mL) have been found in from 18% to 73% of nursing home residents with 10 out of 12 studies reporting a more than 30% of the population showing deficiency.

Background of vitamin D supplementation in the NH

The low prescription rate of vitamin D in NH residents is noteworthy. This unjustified undertreatment relies on wrong assumptions such as the lack of utility of a long-term treatment, the poor tolerance of patients to the supplementation, and polypharmacy issues due to the frailty or disability status (43). However, these assumptions are not supported by the available evidence (4, 43).

The prevalence of use of vitamin D supplements in NH varies among countries (and even among regions within the same country), but it constantly remains low. Vitamin D supplements were prescribed to 10.3% of NH residents in Australia (44), 6.2% in Austria (45), 7.0% in Spain (46), and 14.6% in France (47). In the United States, only one third of all residents of a New York university NH received vitamin D supplementation (48). In Arkansas, among 177 residents in a nursing facility, only 9% were prescribed vitamin D (49). Similarly, only 5% of 1,427 NH residents in Maryland were reported to receive vitamin D supplementation (50). In a large sample (n=186,221) of NH residents in Kansas, Maine, Missouri, Ohio, and South Dakota, Wright et al. reported that a combination of calcium and vitamin D supplements was prescribed to only 3.4% of residents (51). Moreover, when prescribed, the dosage of vitamin D supplement is often too low. In Finland, although 32.9% of the NH residents received vitamin D supplementation, only 3.6% received the recommended dose of 800 IU per day (52).

The prescription of vitamin D is strongly associated with the diagnosis of osteoporosis. In fact, the prevalence of supplementation increases when a diagnosis of osteoporosis is reported. As indicated by Colon-Emeric et al. in a study investigating 67 NHs, 69% of residents with osteoporosis or who recently experienced a fracture received vitamin D supplementation (53). Other studies also found higher use of vitamin D supplementation in persons with osteoporosis (48). In this context, it is important to recognize that the diagnosis of osteoporosis itself still remains largely under recognized in NH (4), potentially contributing to the limited use of vitamin D supplementation.

The optimal vitamin D requirement for elderly individuals remains unclear. Quantity, quality and procedures to deliver vitamin D in NH residents are not univocally established. Most guidelines and recommendations propose to continuously supplement all the institutionalized older persons, given the very high prevalence of vitamin D deficiency in the NH residents (Table 2). Moreover, higher doses of supplementation have been recommended in this population compared to what usually suggested in younger adults (4, 54., 55., 56.).

Table 2.

Current guidelines and recommendations for vitamin D supplementation, proposed by different groups and applicable in NH

Guidelines and recommendations	Years	Recommended Vitamin D supplement dosages
American Geriatrics Society (119)	2012	At least 1000 IU/day
National Academy of Medicine (France) (152)	2012	At least 1500 IU/day
American Endocrine Society (84)	2011	1500 to 2000 IU/day
Institute of Medicine (United State) (54)	2011	800 IU/day
Scientific Advisory Council of Osteoporosis in Canada (153)	2010	800 to 2000 IU/day
United State preventive Services Task Force (154)	2010	800 to 1000 IU/day
Consensus recommendations in Australia (116)	2010	1000 IU/day
International Osteoporosis Foundation (55)	2010	800 IU/day
Expert working group on Vitamin D (56)	2010	800 IU/day
Society for Sarcopenia, Cachexia, and Wasting Disease (155)	2010	Vitamin D should be supplemented in all persons with values less than 100 nmol/L (or 40 ng/mL)
Expert working group on Vitamin D (35)	2007	800 to 1000 IU/day of vitamin D3 or the bolus equivalent administered as 50,000 IU D2 doses every 2 months
Expert working group on Vitamin D (156)	2007	1000 IU/day
Japan Osteoporosis Society (157)	2006	200 a 400 IU/day
Nordic Nutrition Recommendation (158)	2006	800 IU/day
American Medical Director Association (159)	2005	800 to 1000 IU/day
The National Kidney Foundation’s Guidelines (US) (160)	2003	PTH level should be determined if the glomerular filtration rate is less than 60 ml/min. If PTH is elevated, the 25(OH)-vitamin D level should be determined. If that level is 75 nmol/L (or <30ng/mL), supplementation with vitamin D should be initiated to achieve a target level of 75nmol/L (or <30 ng/mL).
British Geriatrics Society Falls and Bone Health Special Interest Group (161)	2003	Use of vitamin D supplements in all nursing/residential home population.
FAO/WHO, Recommended Nutritient Intake (162)	2001	600 IU/day
Recommended Dietary Intake in France (163)	2001	600 IU/day
Referenzwerte in Germany, Switzerland, Austria (164)	2000	400 IU/day
Current Care guidelines in Finnish (165)	2000	700 to 800 IU/day
Dietary Reference Value, Recommended Nutritient Intake in United Kingdom (166)	1998	400 IU/day
Dietary Reference Intake and Adequate Intake in the United States of America and Canada (131)	1997	600 IU/day

Note: IU, International Unit; FAO, Food and Agriculture Organization; WHO, World Health Organization; PTH, parathormon.

Available recommendations are largely designed considering the prevention of bone resorption as the primary endpoint of interest and also as a co-adjuvant of all osteoporosis treatments. The level of parathyroid hormone increases when blood level of 25(OH)-vitamin D is lower than 50 nmol/L (or 20 ng/mL). However, vitamin D requirement may be best defined by a blood level of 25(OH)-vitamin D rather than an increased level of PTH. Interestingly, a meta-analysis of randomized clinical trials showed that a higher concentration of 25(OH)-vitamin D (i.e., above 72 nmol/L or 29 ng/mL) is required in order to prevent bone fractures (10). An optimal level of 75 to 100 nmol/L (or 30 to 40 ng/mL) of 25(OH)-vitamin D has been indicated by several groups of experts as the optimal range to prevent fracture events and maintain health benefits (57, 58). Studies in groups of young and old people generally show that blood concentrations of 75 to 100 nmol/L (or 30 to 40 ng/mL) of 25(OH)-vitamin D may be achieved with daily supplementation of 700 IU to 1000 IU of vitamin D (17, 59, 60), although some persons may require higher dosages (61). It has been calculated that a daily vitamin D3 supplementation with 2000 IU may bring the vast majority of elderly people to the desired level (61). This is a dosage that is well-below the threshold usually considered as toxic in older persons (i.e., 4,000IU/day) (62). However, further studies are needed to validate such data and confirm these dosages.

Most of the current guidelines generally advise that institutionalized older individuals should receive a supplementation of at least 1,000 IU of vitamin D per day to achieve the optimal vitamin D status (Table 2). At this dosage, the majority of residents, including those with the highest requirement of vitamin D supplementation, should meet their needs. Nevertheless, the patient’s baseline level of vitamin D, skin pigmentation, and amount of adipose tissue may influence the capacity to adequately responding to the treatment. Taking into account such variability of the population, the working group of the American Geriatric Society has proposed a daily dosage of 1,000 IU of vitamin D per day, plus individualized supplementation according to sun exposure, skin pigmentation, and obesity (Table 1).

Table 1.

Prevalence of 25(OH)-vitamin D Deficiency in Nursing Home Residents

Author	Country	Year	n	25(OH)-Vitamin D
Egsmose et al (135)	Denmark	1987	94	50% <12.5 nmol/L. Note: 400 IU VD* produced normo-calcemia but did not normalize PTH or alkaline phosphatase
Sem et al (136)	Norway	1987	35	83% of women and 77% of men <50 nmol/L. Note: Supplement users had values >20 ng/mL
Webb et al (137)	USA	1990	211	29% <25 nmol/L in non VD users and 3% <25 nmol/L in VD users Note: VD use (400IU) kept levels >37.5 nmol/L in 95%
Hemmelstein et al (138)	USA	1990	30	50% <50 nmol/L
McMarty et al (31)	USA	1992	57	mean VD level: 43.75 nmol/L
Löwik et al (139)	Holland	1992	51	73% <25 nmol/L
O’Dowd et al (33)†	USA	1993	109	VD intake 200-400 IU: 40% <37.5 nmol/L VD intake 400 IU†: None <37.5 nmol/L
Komar et al (140)	USA	1993	109	86% <50 nmol/L
Stein et al (141)	Australia	1996	99	52% <28 nmol/L
Plantalech et al (142)	Argentina	1997	77	40.5% <25 nmol/L
Kinyamu et al (143)	USA	1997	60	8% <12 nmol/L
Liu et al (24)	Canada	1997	155	9% in fall and 18% <25 nmol/L in spring 38% in fall and 60% <40 nmol/L in spring
Stein et al (42)	Australia	1999	83	mean VD level: 27 nmol/L. Note: Values lower than in those who fell
Krieg et al (144)	Switzerland	1999	103	mean VD level: 11.8 nmol/L. Note: 440 IU of VD3 for 2 years increased bone broadband ultrasound attenuations
Nashimoto et al (145)	Japan	2002	220	57.9% <30 nmol/L
Elliott et al (38)	USA	2003	49	60% <50 nmol/L
Sambrook et al (146)	Australia	2004	637	73.6% <39 nmol/L. Note: VD associated with falls
Flicker et al (41)**	Australia	2005	625	58% <40 nmol/L. Note: 1,000 IU daily or10,000 IU VD weekly reduced falls and fractures in those <25 nmol/L
Drinka et al (147)	USA	2007	302	32% <75 nmol/L
Hamid et al (35)	USA	2007	44	16% <50 nmol/L
Rinaldi et al (148)	Indonesia	2007	62	22.6% <50 nmol/L
Chel et al (27)	Holland	2008	338	98% <50 nmol/L
Perez-Llamas et al (46)	Spain	2008	86	58.2% <50 nmol/L and 32.6% <25 nmol/L
Papapetrou et al (26)	Greece	2008	58	mean VD level: 19 nmol/L
Braddy et al (149)	USA	2009	229	37% <50 nmol/L
Terabe et al (19)	Japan	2012	403	49.1% <40 nmol/L
Pilz et al (40)	Australia	2012	961	93% <50 nmol/L
Verhoeven et al (150)	Belgium	2012	589	75.6% <50 nmol/l
Kruavit et al (151)	Thailand	2012	93	38.7% <20 nmol/L

^*vitamin D; **Persons less than 25 nmol/L excluded; f Appears to be same population;Vitamin D levels below 50 nmol/L define vitamin D insufficiency, below 25 nmol/L define deficiency, and below 12.5 nmol/L define severe deficiency [18] (2.5 nmol/L = 1 ng/mL).

What can we expect from vitamin D supplementation in NH residents?

A recent study has suggested that the lowest rate of mortality is associated with 25(OH)-vitamin D of 75 to 87.5 nmol/L (or 30 to 35 ng/mL) (63). There is evidence that higher vitamin D levels decreased cardiovascular mortality (64, 65). Increased mortality has been found when 25(OH)-vitamin D levels are greater than 98 nmol/L (or 39.2 ng/mL) (66). Research evidence supports that vitamin D can reduce bone turnover, incidence of falls, risk of bone fractures, and improve balance and muscle strength (10, 67., 68., 69., 70.). In particular, a large number of randomized controlled trials (RCT) and several meta-analyses have reported that supplementation of vitamin D, when administered at sufficient dosages, reduces the risk of hip fracture by 13% to 26% in older NH residents (10, 11, 13, 14, 71., 72., 73.). For example, a recent meta-analysis pooling data from 31,022 individuals shows that vitamin D supplementation (at least 800 IU/day) has clinically significant effects on fracture reduction (30% less hip fractures) (12).

However, data from several study-level meta-analyses and one pooled individual participant-level (IPD) analysis are conflicting. While one trial-level meta-analysis of double-blind randomized trials (RCTs) suggested an 18% reduction of hip and 20% reduction of any non-vertebral fractures at a received dose of no less than 482 IU vitamin D per day (11), three study-level meta-analyses (15, 74, 75) and one pooled analysis of IPD (76) from open design and blinded trials, suggested that vitamin D may have a neutral effect on total fractures (74), or may reduce hip fractures by 7 to 16 % independent of its dose if combined with calcium supplementation (15, 74, 75). The discordant findings may in part be explained by different inclusion criteria of trials with respect to blinding and intake form (oral, injectable), or different accommodations for adherence.

Similarly, some well-designed RCTs individual trials included in these meta-analyses, did not report a reduction in the incidence of fractures among older people living in institutional care after supplementation with four-monthly 100,000IU of oral vitamin D2 (77) or equivalent daily dosages (78), while another large trial with 800 IU vitamin D3 per day (16) showed a significant 43 percent reduction in hip fracture risk among nursing home residents Moreover, another recent study showed that high dosages of vitamin D once a year (500,000 IU of cholecalciferol) ambulatory post-menopausal women was associated with a significant increase of risk for hip fracture (73). These discrepancies may, at least partly, explain why physicians are discouraged at adopting systematic supplementation of vitamin D in NH residents. However, the large range of dosages used in the current evidence, the supplementation of non-deficient individuals, the possible use of vitamin D2 supplements, and the variable rates of adherence/compliance to the treatment may explain the different available results (79). Studies that have used high dosages of vitamin D (700 to 800 IU per day) with or without calcium were more likely to be effective in preventing hip fracture (79). Trials with compliance rates of >80%, developed for people aged 70 years and older (especially in NH residents) were likely to report greater benefits from vitamin D supplementation (15).

It is possible that vitamin D supplementation seems particularly effective in NH residents because of their more severe deficiency, their lower dietary intake, and their better compliance to the intervention (80). In this context, it is noteworthy that while supplementation of vitamin D seems to improve both strength and physical performance in deficient older persons (81, 82), little or no effects are reported in individuals with normal concentrations of 25(OH)-vitamin D level (83), unless much higher 25(OH)-vitamin D levels are reached as tested in a small trial among early postmenopausal women (84).

Observational studies suggest that vitamin D may represent a key factor for promoting general health, even beyond its specific role played on bone and muscle tissues. In fact, vitamin D deficiency may play a pathophysiological role in cardiovascular (85, 86), cognitive (87), immunological (88), dental (89) and oncological (90) conditions, which are highly prevalent in NH residents (91). The general health benefits of vitamin D supplementation may also potentially include the 7% decreased risk of mortality reported in a meta-analysis of clinical trials (92), an association confirmed also in NH patients (40) and among senior hip fracture patients followed for 12 months after their fracture (93).

Vitamin D safety

The long-term safety and side effects associated with drug consumption are important sources of concern, especially in NH population. The fear of potential toxicity due to excessive vitamin D supplementation currently poses a barrier for the optimization of a standard nutritional policy (94). Vitamin D has potential health benefits, but it may also adverse effects if given in too high doses.

The large majority of available randomized controlled trials and reviews support that vitamin D supplementation is safe and well-tolerated. In a 2010 risk benefit assessment of vitamin D, a safe upper intake of up to 10,000 IU per day has been suggested based on the risk of hypercalcemia (56), which was the basis of the new safe upper intake recommendation of 4000 IU vitamin D per day among adults by the Institute of Medicine (safety factor of 2.5). Further, it has been reported that up to 4,800 UI/day of vitamin D (up to one year) result in no adverse events (62). Other authors have reported that administration of up to 11.000IU/day of vitamin D3 for 5 months did not result in serum calcium levels rising above the normal range (95). Moreover, it is commonly reported the absence of significant differences for the incident rates of adverse events between control and vitamin D supplementation groups (41, 78, 96). Rapid repletion of vitamin D levels should be avoided. Some authors did not reported an associated between a rapid repletion and an increased risk of side effects in a frail older persons (97). However, Sanders et al. showed the time dependent risk in the first 3 months of application of a 500,000 IU of vitamin D supplementation of frail elderly women at risk of fracture (73). These results have led the recent recommendations not to supplement with large bolus doses.

In practice, supplementation at high dosages, such as 2000 IU/day, is very unlikely to be associated with any toxicity (56, 98), also documented in detail in one 12-month double-blind RCT among acute hip fracture patients with a mean age of 84 years (56). Based on recent researches, the theoretical dose of 250 microg/d (or 10,000 IU/day) of vitamin D3 is proposed to be the highest intake a person may safely take (56, 99). This threshold is far above the 2000 IU/day usually proposed by international authorities and expert groups such as the Food and Nutrition Board (100).

Unfortunately, most of the available data rely on studies focused on healthy older adults. Further studies are needed to better define the safety threshold of vitamin D intake in order to maximize the positive effects of the intervention (101).

In studies using high dosages of vitamin D, the increased concentrations of serum 25(OH)-vitamin D did not follow a linear trend, but drew a curvilinear response suggestive of a rate-limiting mechanism (62). Moreover, very high dosages of vitamin D supplementation have been associated with increased risk of adverse events. In 2010, a large double-blind RCT by Sanders et al., including 2256 community-dwelling women age 70 years and older, tested the benefit of 500,000 IU vitamin D3 given orally once a year, on fall and fracture prevention (73). In those women, mean age 76, considered to be at risk of fracture, 500.000IU vitamin D once a year did not reduce but increase the risk of falls by 15% and the risk of fractures by 26% compared to placebo, with the greatest increase in falls occurring during the first 3 month after dosing. These findings are consistent with another trial that tested 300,000 IU vitamin D2 as an intra-muscular injection once a year (14).

Whether the large dose of vitamin D tested in the Sanders trial was too much of a good thing or not enough to provide a sufficient supply of vitamin D over 12 months is speculative (102). The temporal pattern of events may suggest that the high dose of vitamin D may have induced a “protective” reaction resulting in an acute decrease in 1,25-dihydroxyvitamin D (103). Alternatively, the undocumented potential effect of vitamin D on muscle strength (104) and overall health (i.e. less infections and less hospital admissions (105)) in the Sanders trial, may have improved mobility and decreased “down time”, ironically leading to an increased opportunity to fall and fracture.

As a result of the Sanders trial and given the half-life of vitamin D is 3 to 6 weeks, a daily, weekly, or monthly dosing interval may be most advantageous and safe (11, 27).

Calcium safety

Calcium is frequently associated with vitamin D in the treatment of osteoporosis, and may be responsible of adverse reactions. For example, the combination of calcium and vitamin D may alter the incidence of kidney stones. In the Women’s Health Initiative (WHI) RCT, vitamin D supplementation (at a dosage of 400 IU/day) significantly increased the risk of kidney stones (106) and cardiovascular events (107), especially among individuals who take calcium supplements although their intake from diet is sufficient (800 to 1000 mg/day).

Notably, in two meta-analyses of randomized trials of calcium supplements alone compared with placebo there was no significant effect on fracture risk (14, 39). In fact, in one meta-analysis of 4 double-blind RCTs calcium supplementation at a dose between 600 and 1200 mg per day increased the risk of hip fracture significantly by 64% compared with placebo (39).

In contrast to calcium supplements, dietary sources of calcium are not associated with an increased risk of cardiovascular disease and are recommended in combination with vitamin D supplementation. Further, larger cohort studies, mechanistic studies and small clinical trials support a cardiovascular-protective effect of vitamin D (64, 108).

Risk of active vitamin D metabolites

Active forms of vitamin D, i.e., 1,25(OH)-vitamin D (calcitriol or analogs), have been investigated in 2 meta-analyses with significant benefits on both falls (109) and fracture reduction (11). However, this benefit was not superior to vitamin D supplementation at a dose between 700 to 1000 IU per day.

Further, active vitamin D metabolites, especially in older individuals with renal failure may be associated with an increased risk of hypercalcaemia (110., 111., 112., 113.).

Should calcium be systematically associated with vitamin D

Most meta-analyses suggested that the dose of vitamin D is irrelevant when vitamin D is combined with calcium (15, 74., 75., 76.). In contrast, a most recent pooled meta-analysis of 8 double-blind RCTs that combined vitamin D with calcium supplements found that fracture reduction was present only at the highest actual intake level of vitamin D (792 to 2000 IU/day: median = 800 IU/day). Further, the authors’ findings suggest that at the highest actual intake level of vitamin D supplementation, a calcium dose <1000 mg/d may be preferred to a higher dose of calcium for fracture reduction. These findings are in line with epidemiologic studies suggesting that with sufficient vitamin D intake, a higher calcium intake may not be correlated with bone health (114, 115).

Including these most recent data and the potential risk of calcium supplements, calcium supplementation may be harmful in seniors who get enough calcium through their diet. Further, Calcium supplements have frequent side effects such as constipation, anorexia, and gastrointestinal irritation in older persons. This is also generally associated with a negative impact on long-term adherence to the treatment of osteoporosis and the global dietary intake. Calcium supplementation should be driven by daily dietary intakes from dairy products and patient’s specific tolerance. Calcium supplementation should be carefully considered, especially in residents with history of kidney stones and risk factors for cardiovascular events (116).

In other words, calcium requirements may be simply reached with diet in the NH and calcium supplementation should be prescribed only in case of specific malnutrition.

Testing vitamin D levels before and during supplementation

One approach to identify and manage vitamin D insufficiency is by determining its presence by measuring serum 25(OH)-vitamin D concentrations and then use the specific supplementation according to the patient’s needs. Specific formulas (for example, based on the Heaney’s regression analysis (95)) may be used to facilitate the reaching and maintenance of normal concentrations of 25(OH)-vitamin D (117, 118). This approach is rigorous and precise, but not easy to implement in NH because it implies the assessment of all the residents as well as a familiarity of the coordinating physician with the specific vitamin D protocols. Moreover, promotion of this clinical intervention is likely to be more successful if a unique, common, and standardized approach is always adopted for every NH resident. The rationale for the assessment of vitamin D status resides in the large variability of vitamin D. Such heterogeneity of results is largely explained by exogenous factors (such as sunlight exposure or previous supplementation), which are not particularly relevant in NH residents (who live in the same environment and have similar lifestyles). Therefore, improving the vitamin D status of NH residents with a common and standardized approach seems more suitable and feasible rather than looking for an resident-tailored and more expensive methodology (57).

In this context, it is also noteworthy that the assessment of vitamin D concentrations in NH residents has repeatedly been discouraged (117., 118., 119.). Souberbielle and colleagues proposed that dark-skinned or veiled individuals, older persons, and institutionalized individuals could be supplemented without the need of previous testing of vitamin D concentrations (117, 118). The assessment of vitamin D concentrations may rather be conducted after at least 3 months of daily supplementation to monitor levels (117, 118). Just recently, the American Geriatrics Society (AGS) also revised its recommendations for vitamin D supplementation (119), stating that routine laboratory testing for 25(OH)-vitamin D concentrations before starting supplementation is not necessary in NH populations. The AGS recommendations also indicate no need of monitoring 25(OH)-vitamin D for safety or efficacy reasons if supplementation is conducted at recommended dosages. Moreover, monitoring of serum concentrations of 25(OH)-vitamin D should be done at least after 4 months of vitamin D supplementation (119). Another justification to avoid measuring vitamin D is that substantial inter-assay differences in performance between most common commercially available assays have been reported (120).

Given the probable cost-effectiveness and safety of the intervention, a population- rather than an individual-based approach for vitamin D supplementation without need of preliminary monitoring of the serum concentrations seems acceptable for NH residents.

Proposals for NH care

On the basis of the present narrative review, our group proposes the following recommendations to encourage a wider use of vitamin D supplementation in NH residents. These recommendations represent an attempt to simplify the clinical translation of current evidence in the difficult (and largely unexplored) field of NH patients. Given the limitations of current evidence, alternative approaches to improve the vitamin D status of NH patients might be amenable as well if based on scientifically sound rationales and methodologies.

Proposal 1

Who should be treated in the NH?

All NH residents should be supplemented. For safety reasons, this recommendation may exclude residents at risk of hypercalcaemia (such as residents with sarcoidosis, or myeloma). Although current evidence does not fully support this statement, our group believes that bedridden residents, who are not exposed to falls, may also benefit from vitamin D supplementation as vitamin D has more benefits than just the prevention of non- vertebral fracture, and may support their rehabilitation and mobilization. Vitamin D supplementation may also be relevant for residents in palliative care or with short life expectancy (i.e., less than 6 months) especially if they are still mobile.

Proposal 2

When should residents be treated?

We suggest supplementing residents on admission to the NH. The supplementation should be conducted during their entire NH stay. Transition periods (e.g., the time immediately following NH admissions) have a particularly higher rate of negative health-related events (121). Prevalence of vitamin D supplementation may be proposed as a quality indicator of the preventive policy (122).

Proposal 3

Should baseline serum 25(OH)-vitamin D level be measured?

Our group recommends the implementation of vitamin D supplementation without previous assessment of vitamin D concentrations given the high prevalence of vitamin D deficiency in NH residents. Moreover, the measurement of 25(OH)-vitamin D concentrations is expensive compared to the cost of vitamin D supplementation (about two- to three-fold higher compared to the cost of an annual treatment of vitamin D in Europe).

Proposal 4

Which kind of vitamin D should be used?

As vitamin D requirements cannot be reached by diet (123), 25(OH)-vitamin D concentrations relies on specific supplements, especially in older persons. Both vitamin D2 (ergocalciferol) and vitamin D3 (cholecalciferol) present similar effects on serum concentrations of 25(OH)-vitamin D when daily administered (124). However, vitamin D2 has a shorter plasma half-life than vitamin D3 (125., 126., 127.). Thus, vitamin D3 may be more suitable than vitamin D2 for intermittent supplementation. For this reason, our group recommends the use of vitamin D3. It should be recognized that fortification of bread or orange juice can be equally effective in raising serum 25(OH)-vitamin D levels (128).

In healthy young adults, a one day sunbath can produce large amount of endogenous vitamin D (doses equivalent to an oral intake of up to 10000 to 14000 IU of vitamin D) (117). Sunlight exposure has beneficial effects on the NH residents’ overall health status, in addition to the specific vitamin D production (129, 130). Unfortunately, the organization of systematic sunbaths to NH residents may be particularly challenging especially during the cold seasons. Moreover, half body sun exposure once a week during 2 months was shown to be insufficient for reaching normal concentrations of vitamin D in NH residents (129). This result may probably be also due to the lower effect of sunlight on older persons’ skin. Finally, it cannot be ignored the risk of skin cancer due to sunlight exposure, which becomes higher in elders. Also, seniors tend to avoid the sun and produce 4 times less vitamin D in their skin undersunshine exposure if compared to young adults (84).

Proposal 5

What dose of vitamin D should be prescribed?

Evidence supports that daily oral supplementation with 1.000IU of vitamin D is associated with reduction of falls and fractures in NH residents (41). This dosage is far below the threshold indicated as associated with increased risk of adverse effects (10,000 IU/day) reported by The Food and Nutrition Board (131). Very high intermittent dosage (such as 500,000 IU per year or 200,000 IU every 4 months) should also be avoided. Therefore, our group proposes supplementing with 1,000 IU of vitamin D per day. This proposal is in accordance with the recent AGS recommendations which indicate such daily dosage as the minimum for older subjects (119). It has been reported that 1,000 IU/day is sufficient at maintaining the target concentration of 25(OH)-vitamin D of at least 50 nmol/L (or 20 ng/mL) and will shift at least 50% within the range of 70 and 100 nmol/L (or 28-40 ng/mL) (132). Specific recommendations for obese or dark-skin residents are to be established: these populations may require higher dosages.

At the individual level, it is recognized that a dose of 1,000 IU/day will take a very long time to correct serum levels in a very deficient resident and thus to prevent adverse events associated with vitamin D deficiency. However, at the NH level, our approach would benefit for all NH population without side-effect.

Proposal 6

What protocol of administration may be proposed?

A practical and easy-to-implement protocol is required in NH. The supplementation should be administered without individual assessment, or complex monitoring and adjustments.

Currently, different protocols can support the correction of vitamin D insufficiency and maintenance of normal values. However, these protocols are relevant for personalized care, but likely too complex and not feasible in the daily routine of institutional settings. For example, all of them rely on the preliminary assessment (and follow-up monitoring) of 25(OH)-vitamin D concentrations (56, 133). These protocols have existed for years and have not been implemented by NH staff (2). A controversy exists whether supplementation should be continuously (e.g., daily) or intermittently (e.g., weekly, monthly, quarterly) conducted to achieve the maximal benefits. In the absence of definitive data on this specific topic, the intermittent supplementation might be preferred. In fact, it has the dual advantage of improving compliance, and reducing both daily poly-pharmacy and the burden of the NH personnel.

The recommended target of 25-hydroxyvitamin D concentrations ranges between 75 and 100 nmol/L (or 30 to 40 ng/mL) (132). Our group proposes starting with an oral bolus of 100.000IU followed by 100,000 IU every 3 months.

Alternative protocols may include the monthly supplementation of 50,000 IU, or even other dosages (always maintaining the amount of supplemented vitamin D to an equivalent of 1,000 IU per day).

Proposal 7

Should calcium be added to the vitamin D supplementation?

Calcium should not be systematically prescribed. Current calcium DRI are 1,200 mg per day in older adults (131, 134). If calcium intake is lower than 1,200 mg per day, consequent modifications should be made to the diet. An adequate calcium intake should be indeed present in the diet of NH residents. Calcium supplementation should be prescribed to those residents that do not still reach the desirable intake despite of diet modifications, and rarely higher doses than 500 mg/day are needed. Given the adverse events associated with high dosages of calcium supplementation, caution is required before implementing an intervention specifically modifying calcium concentrations (134).

Proposal 8

How should vitamin D be monitored in NH population?

In a disabled population with multiple comorbidities, the exact correction of 25(OH)-vitamin D concentrations is not considered a priority. Our proposal is aimed at ensuring a sufficient vitamin D status to the vast majority of NH residents. Consistently with other panels of experts (116, 119), our group does not support follow-up monitoring of 25(OH)-vitamin D concentrations. The assessment of 25(OH)-vitamin D concentrations should only be performed when vitamin D intoxication is suspected.

Perspectives and future research

The present recommendations for the use of vitamin D supplementation in NH older residents are designed by the French Society of Geriatrics and Gerontology (SFGG). Moreover, they have been included in the SFGG tool designed for training healthcare professionals in NHs (i.e., MobiQual, Mobilization for Quality of care). The primary aim of MobiQual is to improve quality of life of older persons (including NH residents) by adhering to healthy diet recommendations. The MobiQual is supported by the Caisse Nationale de Solidarité pour l’Autonomie (i.e., the French national healthcare agency for older persons) and the French Minister of Health.

There is still a lack of knowledge on vitamin D supplementation, and further studies are needed in this area. Intense controversies currently exist on vitamin D supplementation, especially when the target population is composed by frail older persons (as NH residents). The ratio of risks/benefits of vitamin D supplementation needs to be carefully explored in specifically designed, long-term RCTs recruiting older persons (including those with increased vulnerability to stressors).

In conclusion, our population- rather than individual-directed approach represents an attempt to provide an immediate answer to the need of treating a common (as well as detrimental) clinical condition in an extremely understudied population. Of course, although based on a scientific rationale and on the available evidence, our recommendations should find confirmation in specifically designed trials, and be validated in complex “real world” of NH. The need of developing research with the specific aim of solving the never-ending “evidence-based medicine” issue existing in geriatric medicine is urgent.