Oversupplementation of Vitamin A and Osteoporotic Fractures in the Elderly: To Supplement or Not to Supplement With Vitamin A


  • John J. B. Anderson Ph.D.

    Corresponding author
    1. Department of Nutrition, Schools of Public Health and Medicine, University of North Carolina, Chapel Hill, North Carolina, USA
    • Department of Nutrition, University of North Carolina, Chapel Hill, NC 27599-7461, USA
    Search for more papers by this author
    • Dr. Anderson has financial interest in the form of stock ownership.


VITAMIN A, an essential micronutrient, is provided by animal foods, whereas β-carotene and other carotenoids, known as provitamin A molecules, are derived from plant sources. In addition, each of these dietary contributors to circulating metabolites and body pools of vitamin A may be obtained from nutrient supplements. Overconsumption of vitamin A by the elderly leading to hypervitaminosis has not been considered to have adverse skeletal effects until recently. However, excess β-carotene in the diet is not likely to result in toxicity. Because of a narrow window for a healthy vitamin A status between deficiency and toxicity, especially in the elderly, the chronic excessive ingestion of vitamin A and other retinoids now is a significant risk factor for reduced bone mass, osteoporosis, and subsequent fragility fractures (Fig. 1).

Figure FIG. 1.

The narrow window of healthy vitamin A intakes in the elderly. Too little vitamin A results in deficiency and too much results in toxicity. Emerging now is a serious concern about modest chronic increases in the consumption of vitamin A, primarily from supplements, that result in hypervitaminosis A with consequent adverse changes in bone tissue and even fractures.

Two recent reports of prospective investigations, one in the current issue of the Journal by Promislow and colleagues(1) and another published a few months ago in JAMA by Feskanich et al.,(2) raise the specter of excessive vitamin A consumption and increased osteoporotic fractures among elderly men and women from two different U.S. populations. Because these reports are population-based, the authors were not able to explore the mechanism(s) through which excessive vitamin A results in bone loss and, ultimately, fractures. Earlier cross-sectional epidemiological reports also implicated this linkage between vitamin A and bone.(3–5)

This editorial addresses three biological aspects of vitamin A that may help explain the observational findings of these two reports. Subsequently, policy implications for controlling the excessive use of vitamin A-rich supplements are raised.


Major food sources of vitamin A (retinol) include animal foods (liver, fishes and fish oils, dairy products, eggs, and margarines and butter) and fortified foods (cereals, juices, and others). U.S. intake data on vitamin A(6) suggest that ∼5-10% of the elderly are consuming this fat-soluble vitamin in excessive amounts relative to recommended amounts for adult men of 900 μg/day, and for adult women of 700 μg/day, slightly lower than the 1989 RDAs.(7) The tolerable upper limit of safety (UL) of vitamin A set by the Committee on Dietary Reference Intakes of the Institute of Medicine is 2800-3000 μg/day for adult women and men.(8) U.S. consumption trends have shown that vitamin A intake from fortified foods has increased since the 1980s(9) and that supplement use by adults has increased over the last 10 years to almost 50% of the population.(10) This relatively large change in the intake of vitamin A may help explain the high rates of osteoporotic hip fractures among the elderly, especially in Europe.(5)

An intake well below the recommended amount of vitamin A for an adult may result in deficiency symptoms. Despite the fortification of select foods with vitamin A and the ready availability of supplements containing the daily recommended amounts of vitamin A, a small percentage of the U.S. population is classified as deficient. Thus, at both extremes of intake, adverse health effects may occur. Although hypervitaminosis A or even toxicity has not been considered a problem in the United States as little as 10 years ago,(11,12) it has now been found to be a major concern because of excessive intakes of vitamin A, that is, the retinoids, rather than with the carotenoids.(13,14)

Despite the well-known difficulties of accurately assessing dietary intakes, especially retrospective exposures, the authors of these studies(1,2) present convincing relative intake data among the subgroups in their study populations. Therefore, the high consumers are truly classified as high consumers although some overestimation of intake may have occurred. Methodological considerations aside, 5-10% of elderly Americans may be consuming excessive amounts of vitamin A on a daily basis, although the intakes may not be exceeding the established UL, primarily because of supplement use.


Vitamin A in animal foods is absorbed typically by the small intestine in the retinol form as an ester, but small amounts of retinoic acid also may be absorbed from the gut lumen. Much of the β-carotenes, perhaps as high as 50%, are cleaved by an intestinal enzyme and the two retinol products absorbed. For other types of carotenoid molecules, it is less clear how many retinol molecules are generated by intestinal digestion. In general, almost 100% of the retinol molecules are absorbed by healthy individuals along with fat digestion products. Absorption at the intestinal lumen typically is achieved by entry of the retinyl esters into micelles and then at the basolateral membrane of absorbing epithelial cells after incorporation in chylomicrons. β-Carotenes also may be absorbed via micelles and chylomicrons. Vegans or strict vegetarians get all of their vitamin A from carotenoid-containing foods. Supplements of either retinol or carotenoids are digested and absorbed by the same mechanisms as the naturally occurring molecules.

Storage of vitamin A in the liver and fat tissue is extensive, and this capacity increases as body fat accumulation mounts with advancing age. In addition, circulating retinol has a longer half-life in the elderly than in young adults, which is related to the slower rate of chylomicron remnant turnover in the elderly.(15) The elevated postprandial serum remnant vitamin A concentration in the elderly compared with young adults signifies a greater time span for vitamin A to transfer to extrahepatic tissues, other circulating lipoproteins, or the liver, and an overall elevation in blood concentration of retinyl esters occurs. In addition, some of the excess vitamin A can no longer be bound by retinol-binding protein and remains “free” within the serum.(15) The increase in free serum vitamin A permits greater exposure of cells to this toxic molecule in elderly individuals, especially in those who use vitamin A supplements.(16) The longer exposure of cells to the free vitamin A makes the elderly more vulnerable for excessive vitamin A accumulation and chronic induction of bone loss, leading to osteoporosis. In general, long-term intakes of ∼1500 μg (4500 IU) per day or greater have resulted in fractures.(5)

The release of retinol from storage sites is not well understood, but the slow serum turnover of vitamin A in the elderly results in greater amounts of unbound vitamin A in serum and chronically increased exposures of target tissues, including bone cells. Retinol conversion to retinoic acid occurs in nearly all cells. This hormonal form of vitamin A is the culprit for the negative consequences at the cellular level when consumption of vitamin A exceeds the dietary reference intake (DRI), that is, hypervitaminosis.


Excessive intakes of vitamin A have long been known to have toxic effects on bone of animal models,(17) but recent rodent experiments have shown that even subclinical hypervitaminosis A can result in fracture risk.(18) However, in humans, only unusual clinical cases of bone loss resulting from overdosing, by oneself or iatrogenic, of vitamin A or other retinoids have been reported.(11,12) Therefore, the current bone studies(1,2) represent the first attempts to establish the vitamin A-fracture linkage by a prospective investigation. The mechanisms for bone loss and resultant fractures relate directly to actions of vitamin A on the DNA of bone cells, both osteoblasts and osteoclasts.

The adverse effects of vitamin A appear to result from the independent effects of retinoic acid on gene expression and negative interactions between retinoic acid and 1,25-dihydroxyvitamin D [1,25(OH)2D].(19,20) Excess binding of the activated retinoic acid receptors to their hormone response elements results in overproduction of cellular proteins in osteoblasts. These proteins, especially cytokines, are secreted by the osteoblasts and, as messengers, act on osteoclasts in bone and recruit their precursors residing in bone marrow. Both the number and the activity of osteoclasts are increased by the actions of cytokines. Continuous stimulation leads to an imbalance between bone resorption and formation and disturbs the microarchitectural details of cancellous bone tissue. If such high intakes of vitamin A continue, the progressive bone loss increases fracture risk and may result in fractures.

Low bone mineral content (BMC) and bone mineral density (BMD) are the first measurable effects of excessive vitamin A accumulation by elderly individuals. The careful observations of bone loss and fractures in the two reports(1,2) testify to the relatively rapid loss of the bone. The increase in hip fractures also suggests that significant loss of cortical bone tissue is occurring also in the affected individuals.


Vitamin A intakes accruing from the three sources of foods, fortified foods, and supplements can easily exceed the UL of safety. Excessive intakes usually are the result of supplement use by well-meaning individuals. On food and supplement labels, the per cent daily value (PDV) is based on the 1968 Recommended Dietary Allowances (RDA) not on the most recent recommendations.(8) (The Food and Drug Administration [FDA] has announced plans to update these values in the near future.) The problem for the consumer is the lack of information available on the UL for vitamin A. Hence, he/she may ingest vitamin A in much greater quantities than the DRIs,(8) but less than the UL. Vitamin A accumulates in storage sites of the body and over time, this “load” acts on bone and other tissues to exert its adverse cellular effects. After ≥5 years of vitamin A supplementation in amounts two- to fourfold greater than the RDA by the elderly, the negative effects on bone present as measurable endpoints, that is, low BMD and fractures.

The standard dose of vitamin A in a multi-mineral/vitamin supplement is 100% of the PDV or 5000 IU/day, whether provided entirely by retinol (retinyl palmitate) or by a combination of retinol and β-carotene. Although this amount is not so high by itself, it adds to the amounts from food sources of vitamin A, such as foods containing naturally occurring molecules and foods fortified with vitamin A. For example, a serving of milk (1 cup) contains 10% of the PDV or 50 IU because of fortification. The food sources containing vitamin A do not present a problem if fortification, such as for milk, remains fairly low, but supplements do present a problem because of the concentrated dosages. (Some ready-to-eat breakfast cereals, however, are fortified at 100% of the PDV.) The safety issue of vitamin A-supplement consumption by elderly individuals now needs to be scrutinized carefully by the FDA.

The current UL for the elderly or those beyond the age of 50 years may need to be lowered considerably below 2800-3000 μg/day(8) to prevent hypervitaminosis and improve bone turnover more in favor of formation. This reduction would make the window of optimal vitamin A consumption even narrower for the elderly. With this UL change, supplement use would need to be eliminated or greatly curtailed. In the future, vitamin A supplementation may require a prescription to assure safety and, hence, no longer be available over the counter.


Vitamin A may be consumed easily in excess if supplements are taken by the elderly. Based on the findings of adverse effects on the skeleton from hypervitaminosis in the prospective human studies(1,2) and in animal investigations,(17–19) the general rule for vitamin A consumption by elderly individuals should be foods solely and supplements of vitamin A only if a physician finds the patient to be truly deficient. Even the consumption of vitamin A-fortified foods is much safer than supplement use by the elderly because food portion size largely determines nutrient intakes; the elderly typically have modest to major reductions of total energy intake because of the consumption of less food from all food groups. However, the ingestion of supplements containing vitamin A is not governed by portion size and, therefore, the elderly can easily ingest excessive amounts. In recent years, most supplement manufacturers have removed iron from their formulations for the people who are over 50 years old; why not also remove vitamin A or its precursors from pill preparations for those over 50 years? Practically all elderly supplement users are modestly well-off or affluent, well educated, and health conscious. These individuals seem to be the same ones most at risk of bone loss and fractures from excessive vitamin A intakes that result in increased amounts of vitamin A circulating in the unbound state.