Dietary Protein and Bone Health

INTRODUCTION

UNDERNUTRITION, PARTICULARLY PROTEIN malnutrition, is frequent in the elderly with osteoporotic hip fracture. On the other hand, high protein intake has been claimed to be harmful for bone health. This editorial addresses the issue of dietary protein in the pathogenesis and/or management of osteoporosis.

DIETARY PROTEIN AND BONE METABOLISM

High protein intake has been claimed to be a risk factor for osteoporosis. The proposed underlying mechanism implies that increasing protein intake increases acid production and renal acid excretion, because of protons released during the oxidation of sulfur-containing amino acids such as methionine, cysteine, and cystine. Because urinary calcium excretion directly varies with acid excretion, urinary calcium is positively correlated with protein intake.⁽¹⁾ These associations would thus suggest that high protein intake will in turn induce a negative calcium balance and consequently would favor bone loss.⁽²⁾ Furthermore, nutrition-generated acid load would lead to an increased bone dissolution in healthy individuals, analogous to the classical physico-chemical in vitro observation indicating that lowering pH favors the dissolution of calcium phosphate crystals, including those of hydroxyapatite. However, further studies indicate that a reduction in dietary protein may lead to a decline in calcium absorption and to secondary hyperparathyroidism.^{(3, 4)} A low (0.7 g/kg body weight), but not a high (2.1 g/kg), protein intake was associated with an increase in biochemical markers of bone turnover compared with a diet containing 1.0 g/kg of protein.⁽⁵⁾ High meat diets (1.6 g/kg body weight of protein) compared with 0.9 g/kg for 8 weeks did not affect calcium retention or indices of bone metabolism.⁽⁶⁾ The question as to whether the source of proteins, animal versus vegetal, would differently affect calcium metabolism has been the object of more emotional belief than serious scientific demonstration.⁽⁷⁾ This belief lies on the hypothesis that animal proteins would generate more sulfuric acid from sulfur-containing amino acids than a vegetarian diet. That animal protein in contrast to vegetal protein would be consistently detrimental for bone health is not supported by chemical and experimental evidence. Indeed, a vegetarian diet, with protein derived from grains and legumes, would deliver as many millimoles of sulfur per gram proteins as would a purely meat-based diet.^{(7, 8)} On the other hand, it is true that meats contain substances other than sulfur acid-producing substances. However, the net release of proton buffers from bone mineral does not seem to significantly contribute to blood acid-base equilibrium, even in response to a conspicuous proton load, unless renal function is severely impaired.⁽⁹⁾ High protein intakes are not associated with significant changes in blood pH.⁽¹⁰⁾ However, in favor of this endogenous acid production in bone metabolism, it seems that neutralization of this endogenous acid production with potassium bicarbonate is associated with positive calcium balance.⁽¹¹⁾ In a cross-sectional survey, BMD was higher in subjects with diets rich in fruits and vegetables, presumably rich in alkali.^{(8, 12)} This issue is further complicated by the fact that the vegetable intake-induced decrease in bone resorption⁽¹³⁾ has been shown to be independent of acid-base changes⁽¹⁴⁾ and that potassium, but not sodium bicarbonate (i.e., the same anion) or citrate, administration reduces urinary calcium excretion.^{(1, 15)} Finally, if high protein intake was really harmful for the skeleton, some consistent inverse relationship should be detected between BMD and dietary protein.

DIETARY PROTEIN AND BONE MASS

In fact, nearly all the studies have found a positive rather than a negative association between bone mineral mass and dietary protein intake^(16-31): the higher the protein intake, the lower the bone loss. Unadjusted BMD was greater in the group with the higher protein intake in data collected in the Study of Osteoporotic Fracture.⁽²⁸⁾ In a longitudinal follow-up in the frame of the Framingham study, the rate of bone mineral loss was inversely correlated to dietary protein intake.⁽³²⁾ Recently, spontaneous higher protein intake was associated with an increase in femoral neck BMD in a group of postmenopausal subjects receiving calcium supplements.⁽³³⁾ These findings are in agreement with an observation done in middle-aged women, in whom protein intake and forearm bone mineral mass changes over a 4-year follow-up were positively correlated.⁽³⁴⁾ On the other hand, several studies have not found any relationship, neither positive nor negative, between bone mineral mass and dietary protein intake in cross-sectional^{(12, 35-38)} or longitudinal designs.^{(27, 39, 40)} In contrast, very few surveys associated high protein intake with a lower bone mass. Indeed, in a cross-sectional study conducted in young college women, a protein intake close to 2 g/kg body weight was associated with reduced BMD only at one of two forearm sites measured.⁽⁴¹⁾ A negative association was found between BMD and dietary protein of vegetable origin, whereas it was positive for animal protein.⁽²⁶⁾ Only with various adjustments did BMD changes become significantly negative in the group with the higher animal-to-vegetal protein ratio.⁽²⁸⁾ An animal-to-vegetal protein ratio remains mechanistically difficult to interpret, because the same ratio can be achieved with a large variety of protein contents and thus of sulfur-containing amino acids. More importantly, the statistically negative relationship between the animal-to-vegetal protein ratio and bone loss was obtained only after multiple adjustments, making it difficult to generalize these findings in terms of nutritional recommendations for bone health and osteoporosis prevention.⁽⁷⁾ However, two studies that specifically addressed the issue of animal versus vegetal protein came to the conclusion that beneficial effects of protein on bone health were more related to the former than the latter component.^{(32, 33)}

Randomized controlled intervention trials provides a higher level of evidence than observational studies. Hence, in a double-blind, placebo-controlled study, the correction of a low protein intake with 20 g protein supplement daily for 6 months compared with an isocaloric placebo attenuated proximal femur BMD decrease by 50%.⁽⁴²⁾ In this trial, all 82 patients (80.7 ± 1.2 years of age) were given 200,000 IU vitamin D once at baseline and 550 mg/day of calcium, starting within 1 week after hip fracture. Thus, in undernourished elderly with a recent hip fracture, an increase in the protein intake, from low to normal, was beneficial for bone. In another randomized controlled trial conducted in elderly with a body mass index (BMI) below 21 kg/m², all receiving calcium and vitamin D, a 12 g protein supplement significantly decreased a serum biochemical marker of bone resorption, increased osteoprotegerin levels, and was associated with a trend toward increased BMD.⁽⁴³⁾

DIETARY PROTEIN AND FRACTURE RISK

It has been reported that hip fracture incidence was greater in countries with high protein intake of animal origin.^{(44, 45)} However, as expected, the countries with the highest incidence of hip fracture are those with the longest life expectancy and thus the greatest lifetime risk of fracture. On the other hand, in a similar cross-cultural type of survey, a higher incidence of hip fracture was positively correlated with the number of books in the public library in the different countries assessed, casting some doubt about any causal relationship in such cross-cultural studies (S. Maggi, personal communication, 2002). In another survey, fracture risk was increased when a high protein diet was accompanied by a low calcium intake, whereas no association between hip fracture and nondairy animal protein intake could be overall detected.⁽⁴⁶⁾ In the large Nurses Health Study, a trend for hip fracture incidence inversely related to protein intake was reported.⁽⁴⁷⁾ In this study, forearm fracture risk was increased with dietary protein intakes higher than 95 g/day. In the NHANES I study, hip fractures were associated with low energy intake, low serum albumin levels, and low muscle strength, all reflecting low protein intakes.⁽⁴⁸⁾ In a prospective longitudinal cohort study carried out on more than 40,000 women in Iowa, followed up over a 3-year period, higher protein intake was associated with a significantly reduced risk of hip fracture.⁽⁴⁹⁾ The reduction in hip fracture risk reached 56-67% in the highest protein intake quartile. Similarly, a reduced relative risk of hip fracture was found with a higher intake of milk as a source of both calcium and protein.^{(50, 51)} A longitudinal study has shown that protein of vegetable origin was protective.⁽²⁸⁾ However, in this study, hip fracture incidence was positively related to the animal-to-vegetable ratio of protein intake. As pointed out above, the significance of a ratio, irrespective of the absolute amount of protein ingested, is difficult to interpret.

Thus, considering the large number of studies having addressed this issue, there is a near consensus that increasing dietary protein in the range of usual intakes is not harmful to bone health, but rather, beneficial.^{(52, 53)} This contention is strongly supported by the results of a large case-control study reported in this issue of the Journal.⁽⁵⁴⁾ In this study performed in 1167 cases and 1334 controls of both genders, increasing protein intake was associated with a decreased risk of hip fracture of 65% in the highest quartile in the 50- to 69-year-old age class. There was no interaction with calcium or potassium intakes. However, no similar association could be observed among the participants of the 70- to 89-year-old age class. An age-specific association between protein and bone health is unlikely. Indeed, intervention studies in which protein supplements were shown to positively influence bone mass or remodeling were conducted in patients older than 70.^{(42, 43)} As stated by Wengreen et al.,⁽⁵⁴⁾ the inability to detect protective effects of dietary protein in the older group might be related to some selection bias in favor of the healthiest hip fracture cases because of the higher mortality rate in the more frail patients and the greater difficulty to complete the interview and to report accurate dietary intakes in this class of age.

DIETARY PROTEIN AND FRACTURE OUTCOME

Undernutrition is often observed in the elderly, and it seems to be more severe in patients with hip fracture than in the general aging population.^(55-57) A state of undernutrition at admission is consistently documented in elderly patients with hip fracture,^{(55, 56, 58)} which cannot only have contributed to the risk of fracture, but may also adversely influence the clinical outcome.^{(55, 57, 59)} Intervention studies using a simple oral dietary preparation that normalizes protein intake can improve the clinical outcome after hip fracture.⁽⁵⁷⁾ Follow-up showed a significant difference in the clinical course in the rehabilitation hospitals, with the supplemented patients doing better. The significantly lower rate of complication (bedsore, severe anemia, intercurrent lung or renal infections) and death was still observed at 6 months.⁽⁵⁷⁾ In this study, the total length of stay in the orthopedic ward and convalescent hospital was significantly shorter in supplemented patients than in controls. It was shown that normalization of protein intake, independently of that of energy, calcium, and vitamin D, was responsible for this more favorable outcome.^{(42, 60)} The normalization of protein intake was found to increase insulin-like growth factor (IGF)-I.⁽⁴²⁾ In a multiple regression analysis, baseline serum IGF-I levels, biceps muscle strength, and protein supplementation accounted for more than 30% of the variance of length of stay in rehabilitation hospitals, which was shortened by 25%.

DIETARY PROTEIN AND IGF-I

Experimental and clinical studies indicate that dietary proteins influence both the production and action of IGF-I and thereby bone anabolism.^{(61, 62)} Protein restriction reduces IGF-I plasma levels by inducing a resistance to the action of GH at the hepatic level^{(61, 63)} and by an increase of IGF-I metabolic clearance rate.⁽⁶⁴⁾ Protein restriction seems to render target organs less sensitive to IGF-I. Indeed, the administration of pharmacologic doses of IGF-I, producing a 5-fold increase in IGF-I circulating levels in adult rats, in an attempt to correct the negative influence of protein deficiency, is without effects on bone if the protein intake is insufficient.⁽⁶⁵⁾ Decreasing protein intake from 0.9 to 0.4 g/kg body weight (this protein intake is quite common in elderly) decreases muscle fiber cross-sectional area.⁽⁶⁶⁾ A low serum IGF-I level is a risk factor for osteoporotic fracture.^{(67, 68)} On the other hand, increasing protein intake prevents the decrease in IGF-I usually observed in hypocaloric states,^{(58, 69)} and refeeding of undernourished patients increases IGF-I.^{(42, 70, 71)}

In an experimental model using adult female rats, a selective protein deprivation with isocaloric low protein diets containing identical amounts of minerals decreased BMD at the level of skeletal sites formed by trabecular or cortical bone.⁽⁷²⁾ In this model, the decrease in bone mass and bone strength is related to an early decrease in plasma IGF-I (40%) and bone formation and a later acceleration of bone resorption. Histomorphometric analysis and biochemical markers of bone remodeling results indicate that the low protein intake-induced decrease in bone mineral mass and bone strength is related to an uncoupling between bone formation and resorption.⁽⁷²⁾ Protein replenishment by administering essential amino acid supplements in the same relative proportion as in casein increases IGF-I and biochemical bone formation and decreases markers of bone resorption. Under these conditions, bone strength increases more than bone mineral mass, probably in relation with an increase in cortical thickness, as demonstrated by μCT.⁽⁷³⁾

CONCLUSION

There is a large body of evidence linking nutritional intakes, particularly protein undernutrition, to bone homeostasis and osteoporotic fractures. Whereas a gradual decline in caloric intake with age can be considered as an adequate adjustment to the progressive reduction in energy expenditure, a commensurate reduction in protein intake may be detrimental for maintaining the integrity and function of bone.⁽⁷⁴⁾ Thus, increasing dietary protein to the normal intake is beneficial to bone health.