Protein intake, bone mineral density and risk of hip fracture: a systematic literature review

Abstract

Background

The prevalence of osteoporosis is high and osteoporotic fractures affect a large proportion of the older population. A vital question is therefore how we can prevent bone loss and reduce the prevalence of osteoporosis. It has been proposed that protein intake could have an effect on bone mineral density and the risk of hip fracture.

Objectives

The main objective of this systematic review was to assess and give an overview of the existing literature on the associations between protein intake, bone mineral density and hip fractures.

Methods

A systematic search was conducted in PubMed up until 20th of February 2019, following predefined inclusion and exclusion criteria and relevant search terms.

Results

The initial search gave 1746 results, and twenty studies were included in the review, of which fifteen were observational, four were RCTs and one was a case-control study. The results indicated a potentially small beneficial effect of protein intake on bone mineral density, especially in the older populations, and a potentially protective effect of higher protein intake on hip fracture.

Conclusion

Protein intake may have a beneficial effect on bone loss, and a preventive effect on the risk hip fracture. Future research is needed to look further at the potentially beneficial effects of protein intake on bone health in a lifetime perspective

Introduction

The bone tissue is involved in several important functions in the body, such as the ability to move, support and protect muscle tissue and vital organs, calcium and phosphate metabolism and production of blood cells in the bone marrow (1). Bone health is therefore of major importance for overall health and physiological function.

Osteoporosis

An imbalance of bone resorption and bone formation could result in bone loss and development of osteoporosis (1). Osteoporosis can be defined as «a disease characterized by low bone mass and microarchitectural deterioration of bone tissue leading to enhanced bone fragility and a consequent increase in fracture risk» (2).It is the most prevalent bone disorder in the world, and it is estimated that over 200 million people worldwide are affected (3).

There are several risk factors associated with the risk of osteoporosis. Some of the non-modifiable factors are sex, ethnicity and age, while modifiable risk factors include low BMI, smoking, inactivity, diet and use of e.g. corticosteroids (4).

With aging several physiological changes occurs that could affect the regulation of the bone remodelling cycle, including decreased oestrogen levels, reduced ability to absorb calcium and changes in vitamin D metabolism due to reduced skin synthesis and kidney function (5). Bone mineral density (BMD) thus declines with age, and by age 70, bone mass has decreased by about 30 – 40 percent, leading to an increased risk of fractures (5). In addition, a low peak bone mass attained during growth, ie. the amount of bone gained by the time a stable skeletal state has been attained during young adulthood, is associated with increased risk of fractures later in life (6). Due to a lower peak bone mass during growth and the reduced levels of oestrogen with menopause associated with decrease of BMD, osteoporosis is far more common in women than in men (7).

The prevalence of osteoporosis-related fractures is high (7), especially in Caucasian populations like Norwegians (8). The most adverse osteoporotic fractures are hip fractures (9), that are associated with increased mortality and serious impaired physical functioning (4).

Associations between protein intake and bone health

Constituents in the diet, and especially calcium, vitamin D and protein, have been suggested to be associated with bone mineral density and risk of osteoporosis (10). The role of protein intake in bone health has been controversial, and both beneficial and detrimental effects of dietary protein have been proposed (10). Several mechanisms for both negative and positive associations between protein intake and bone health have been suggested, but there are still some uncertainties regarding the long-term effects of protein intake (11).

Based on the many possible connections between dietary protein intake, BMD and risk of fractures, it is highly relevant to investigate the current evidence for the potential effects of protein intake on BMD and risk of fractures. Therefore, the main objective of this systematic review was to give an overview and a critical evaluation of existing literature on the associations between amount and source of protein intake and BMD and risk of hip fractures in both genders of all age groups. Further, the objective was to investigate whether a higher protein intake could have a protective effect on bone loss, development of osteoporosis and accompanying fractures.

Methods

Literature search

The reporting of this systematic review adheres to the PRISMA guidelines (12). A systematic search and manual selection process was conducted in PubMed. The following search string was used for studies on protein intake and BMD: [((Protein intake OR Dietary protein OR Whey protein)) AND (Bone mineral density OR bone density OR bone mass density OR BMD)]. The search string used for the studies on protein intake and fracture was [((Protein intake OR Dietary protein OR Whey protein)) AND (hip fracture OR femoral neck fracture OR Trochanteric fracture)]. The search covered the period up until 20 February 2019.

Study eligibility criteria

The present review included relevant studies of both men and women without any recent diagnosed fracture or diseases that could have an impact upon BMD, like osteomalacia, Rickets, primary hyperparathyroidism, hypocalcaemia etc. There were no limitations of age in the included populations. Both observational studies and randomized controlled trials (RCTs) were included. We excluded studies only comparing protein intake from different sources and not the quantity of protein, studies that included weight loss interventions, and studies of pregnant or lactating women.

Results

Study selection process

The search strategy and study selection process are presented in the flow chart (Figure 1). The initial search in PubMed gave a total of 1746 results, of which 20 studies were included in the review. Thirteen of the studies (13 – 25) assessed the associations between protein intake and BMD, of which nine (13 – 15, 17 – 21, 25) were observational studies and four were RCTs (16, 22 – 24). The last seven were observational studies (26 – 32) investigating the effects of protein intake on hip fractures.

Figure 1. Search strategy for the selection of studies included in the literature review (12)

Study characteristics

Study population

Twelve of the studies (15 – 17, 19 – 21, 25 – 27, 29, 31, 32) included both men and women in the study population, seven (13, 14, 18, 22 – 24, 28) included only women, and one (30) included only men as a part of the study population.

The age ranges also varied in the study populations included. Thirteen of the studies (14 – 18, 21 – 23, 25, 28 – 32) inclued subjects aged 40 years and older. Four of the studies (19, 20, 26, 27) had a study population with a wide age range that included adults older than 25 years, and two of the studies (13, 24) were done in younger subjects with mean age 24 and 21 years old, respectively.

Study exposures

Five (14, 18, 19, 25, 27) of the studies only assessed the associations between total protein intake and BMD or hip fractures, one study (32) only assessed the effect of non-dairy animal protein, while ten (13, 15, 17, 20, 21, 26, 28 – 31) also looked at the effects of different sources of protein.

The quantification and categorization of levels of protein intake varied in the included observational studies. Eight studies (15, 18, 19, 21, 27, 29, 30, 32) quantified protein intake as percentage of energy intake (E %) and categorized into quartiles of protein intake, one study (31) categorized into quintiles, while one study (13) categorized protein intake as tertiles of intake. One study (20) categorized the levels of protein intake in terms of predefined values of E %, where below 15 E % was defined as low protein intake. In addition, two of the studies (17, 25) quantified protein intake as a continuous variable in terms of grams of protein per day, while another study (26) categorized protein intake into tertiles of grams of protein per day. The last study (28) quantified protein intake in terms of grams per MJ energy intake, categorized into quartiles of protein intake.

Study outcomes

Bone mineral density was measured with different methods and at different anatomic sites in the studies included. DXA-scan was used to measure BMD in twelve (13, 15 – 25) of the studies, one study (21) used Lunar dual photon absorptiometer (DP3) in addition to DXA, while the last one (15) used quantitative computed tomography.

Eleven of the studies included measurements of the BMD at the spine (13 – 21, 23, 24), six included total hip BMD (13, 14, 16, 17, 20, 22), four studies measured BMD of femoral neck (17 – 19, 21), three included BMD of trochanter (18, 19, 21), two included radius (18, 21), one included Ward’s area (21), one measured heel BMD (25), and four included measurement of total body BMD (13, 14, 17, 18).

The secondary outcome of interest, hip fractures, was reported in different ways across the five included studies. In two of the studies (26, 27), hip fractures were reported by interviews or phone calls, while four other studies (28, 30 – 32) assessed hip fracture by regularly follow-up and verification by physicians. The case-control study (29) used cases of hip fractures collected from 18 hospitals in Utah, and compared with randomly selected controls.

Protein intake and bone mineral density

Main results

Observational studies

Five (14, 15, 20, 21, 25) of the eight observational studies (13 – 15, 17 – 21) investigating the association between protein intake and BMD found positive associations between protein intake and BMD, while the last four (13, 17, 18, 26), showed no or negative associations (Table 1).

Table 1: Characteristics and results of the eight included observational studies investigating the longitudinal associations between protein intake and bone mineral density.

Author, year

Country

Follow‑up

Study design

Exposure

Outcome measure of interest

Study groups

Main results

Langsetmo, 2015

Canada

5 years

Cohort

Protein intake as percent of total energy intake, dairy protein, non-dairy animal protein, and plant protein assessed by FFQ

Total hip and lumbar spine BMD measured by DXA

6510 men and women aged 25 years and older:

Men 25 – 49 y:

No associations

Pre-menompausal women 25 – 49 y:

No associations

Men 50+ y:

No associations

Post- menopausal women 50+ y:

Spine BMD: plant protein = increased bone loss β-coefficient: – 0,003 (-0,005, 0,000)

Hip BMD: dairy protein = decreased bone loss β-coefficient: 0,002 (0,000, 0,003)

Hu, 2014

USA

3 years

Cohort

Total, animal and vegetable protein intake as percentage of energy intake assessed by FFQ

Lumbar spine trabecular volumetric bone mineral density (vBMD) measured by quantitative computed tomography

1658 men and women aged 45 – 84 years:

317 Non-Hispanic white women

Lumbar spine BMD: Higher Vegetable protein = greater lumbar spine vBMD

β-coefficient: 0,44 (0,07, 0,81)

Total and animal protein: no associations

340 Black, 230 Chinese, 430 Hispanic men and women

No associations

Sahni, 2014

USA

4 – 6 years

Cohort

Total protein intake as percent of energy intake assessed by FFQ

BMD of the hip (femoral neck and trochanter) and lumbar spine measured by DXA

1280 Men aged 26 – 86 years

Trochanter BMD: higher protein intake E % = increased bone loss.

β-coefficient: – 0,0498 (SE = 0,020)

1639 Women aged 26 – 86 years

No associations

Beasley, 2010

USA

2 – 3 years

Cohort

Amount and source (animal and vegetable) of dietary protein intake assessed by FFQ

BMD of the hip, spine and whole-body measured by DXA

560 women aged 14 – 40 years

Hip BMD: No associations

Spine BMD: No associations

Total body BMD: No associations

Absolute change in BMD for every percentage increase in E % from protein:

Hip: -0,0002 g/cm2

Total body: -0,0012 g/cm2

Spine: 0,0004 g/cm2

Rapuri, 2003

USA

3 years

Cohort

Protein intake as a percentage of energy assessed by 7-days food diaries. Quartiles of protein intake.

BMD of the hip, spine, radial midshaft, and total body measured by DXA

489 women aged 65 – 77 years

Spine BMD: No associations. Difference in BMD change in Q4 vs Q1 protein intake: 2,8 %

Total body BMD: No associations.

Difference in BMD change in Q4 vs Q1 protein intake: -1,1 % Midradius BMD: No associations.

Difference in BMD change in Q4 vs Q1 protein intake: 1,0 % Femoral neck BMD: No associations. Difference in BMD change in Q4 vs Q1 protein intake: 1,0 %

Trochanter BMD: No associations. Difference in BMD change in Q4 vs Q1 protein intake: 0,7 %

Total femur BMD: No associations. Difference in BMD change in Q4 vs Q1 protein intake: 2,0 %

Promislow, 2002

USA

4 years

Cohort

Total, animal and vegetable protein intake assessed by FFQ

BMD and bone loss at the hip, femoral neck, and lumbar spine measured by DXA

388 Men aged 55 – 92 years

Total hip BMD: No associations.

β-coefficient: -0,0003 (-0,0180, 0,0174)

Femoral neck BMD: No associations. β-coefficient: -0,0045 (-0,0202, 0,0112)

Total spine BMD: No associations. β-coefficient: -0,0095 (-0,0345, 0,0155)

Total body BMD: No associations. β-coefficient: -0,0078 (-0,0212, 0,0057)

572 Women aged 55 – 92 years

Total hip BMD: No associations. β-coefficient: 0,0094 (-0,0025, 0,0214)

Femoral neck BMD: No associations.

β-coefficient: 0,0063 (-0,0039, 0,0165)

Total spine BMD: No associations.

β-coefficient: 0,0084 (-0,0090, 0,0258)

Total body BMD: No associations.

β-coefficient: 0,0081 (-0,0017, 0,0179)

Beasley, 2014

USA

6 years

Cohort

Protein intake assessed by FFQ and calibrated by using biomarkers of energy and protein intakes.

BMD and bone loss at the hip, posterior-anterior spine, and total body measured by DXA

9062 women aged 50 – 79 year

Total body BMD: 20 % higher calibrated protein intake = decreased bone loss

BMD change per 20 % increase in E % from protein: 0,004 g/cm2 (0,001, 0,007)

Hip BMD: 20 % higher calibrated protein intake = decreased bone loss

BMD change per 20 % increase in E % from protein: 0,002 g/cm2 (0,001, 0,004)

Spine BMD: 20 % higher calibrated protein intake = no association

BMD change per 20 % increase in E % from protein: 0,003 g/cm2 (0,000, 0,008)

Hannan, 2000

USA

4 years

Cohort

Total and animal protein intake

BMD and bone loss at the spine, femur and radius

391 women and 224 men 68 – 91 years

Spine BMD: higher total and animal protein = decreased bone loss.

Percentage BMD change Q1- Q4:

Total protein: -2,61 Animal protein: -2,14

Femoral neck BMD: higher total and animal protein = decreased bone loss.

Percentage BMD change Q1- Q4:

Total protein: -2,29 Animal protein: -1,80

Trochanter BMD: higher total and animal protein = decreased bone loss.

Percentage BMD change Q1- Q4:

Total protein: – 1,35 Animal protein: -0,62

Wards area BMD: higher total and animal protein = decreased bone loss.

Percentage BMD change Q1- Q4:

Total protein: -2,66 Animal protein: -2,05

Radius BMD: No associations.

Percentage BMD change Q1- Q4:

Total protein: 0,10 Animal protein: -0,12

Steell, 2019

UK

Cross-sectional

Total protein intake, g/kg/day, 24-h recall questionnaire

Heel BMD

31149 women and 39066 men aged 40 – 69

Men: Total protein intake and heel BMD: positive association β-coefficient: 0,008 (0,000 – 0,015)

Women: Total protein intake and heel BMD: positive association β-coefficient: 0,010 (0,005 – 0,015)

RCTs

The first RCT included (16) found no significant differences in BMD between the high-protein supplement group and the placebo group, while the second RCT (22) showed a significant between-group difference between the protein group and the control group.

The last two RCTs included (23, 24) investigated the effect of milk basic protein (MBP) on BMD of the lumbar spine in menopausal (23) and young women (24). In both RCTs the results showed a significant positive effect of MBP on BMD of the lumbar spine (Table 2).

Table 2: Characteristics and results of the four included RCTs investigating the associations between protein intake and bone mineral density.

Author, year

Study groups

Country

Follow‑up

Study design

Intervention

Outcome measure of interest

Main results

Kerstetter, 2015

208 older women and men: Women 60 + years, Men 70 + years

USA

18 months

RCT

45 g whey protein or isocaloric maltodextrin supplement

BMD by dual energy X-ray absorptiometry

Lumbar spine BMD: no effect. Between-group difference ( %): 0,002 (-0,011, 0,014)

Total hip BMD: no effect. Between-group difference ( %): 0,001 (-0,007, 0,009)

Femoral neck BMD: no effect. Between-group difference ( %): 0,006 (-0,004, 0,016)

Zhu, 2011

219 healthy women aged 70 to 80 years.

Australia

2 years

RCT

A high-protein drink containing 30 g of whey protein (n ¼ 109) or a placebo drink identical in energy content

BMD at total hip measured by DXA and QCT

Total hip: no effect. Between group difference (mg/cm3): 0,19

Uenishi, 2007

35 healthy young women (mean age 21 years old)

Japan

6 months

RCT

Milk basic protein (40 mg/ day) or placebo

BMD of the lumbar spine measured by DXA

Lumbar spine: MBP-supplementation = increased BMD. Mean gain of BMD (control group – placebo group) = 1,44 %

Aoe, 2005

32 healthy menopausal women (mean age 50,5 years old)

Japan

6 months

RCT

Milk basic protein (40 mg/ day) or placebo

BMD of the lumbar spine measured by DXA

Lumbar spine: MBP-supplementation = increased BMD. Mean gain of BMD (control group – placebo group) = 1,87 %

Protein intake and hip fracture

Five (26 – 28, 30 – 32) of the seven studies (26 – 32) investigating the associations between protein intake and hip fractures were cohort studies, and the last one (29) was a case-control study.

Main results

Cohort studies

The six cohort studies gave diverse results regarding the effects of different protein sources, however, five of the six studies (26 – 28, 30, 31) showed protective effects of protein intake on hip fractures, while the last one showed an elevated risk of fracture in the group of women with high protein intake and low calcium intake (32) (Table 3).

Table 3: Characteristics and results from the five included studies investigating the associations between protein intake and hip fractures.

Author, year

Population

Country

Follow‑up

Study design

Exposure

Outcome measure of interest

Study groups

Main results

Sahni, 2010

1752 men and 1972 women aged 26 to 86 years

USA

12 years

Cohort

Total protein, animal protein and plant protein assessed by FFQ

Hip fractures reported by interview

Total calcium intake > 800 mg/day

Higher total protein: No associations HR = 0,54 (0,12, 1,30)

Higher animal protein = decreased risk of hip fracture

HR = 0,15 (0,02, 0,92)

Higher vegetable protein:

No associations

HR = 0,24 (0,06, 1,06)

Total calcium intake < 800 mg/day

Higher total protein: No associations. HR = 2,20 (0,88, 5,54)

Higher animal protein = increased risk of hip fracture.

HR = 2,84 (1,20, 6,74)

Higher vegetable protein:

No associations.

HR = 0,56 (0,19, 1,68)

Misra, 2011

576 women and 370 men aged 28 – 62 years

USA

11,6 years

Cohort

Total protein and protein as percent of energy intake assessed by FFQ. Quartiles of protein intake.

Hip fractures reported by interview or phone call

Men and women

Higher protein intake = decreased risk of hip fracture. Q4 vs Q1

HR = 0,63 (0,37, 1,09)

Munger, 1999

32050 women aged 55 – 69 years

USA

3 years

Cohort

Total protein, animal protein and vegetable protein intake assessed by FFQ. Quartiles of protein intake.

Hip fracture reported by follow-up questionnaires, and verification by physicians

Women

Higher total protein = decreased risk of hip fracture. RR Q4 vs Q1 = 0,44 (0,16, 1,22)

Higher animal protein = decreased risk of hip fracture.

RR Q4 vs Q1 = 0,31 (0,10, 0,93)

Higher vegetable protein = no associations.

RR Q4 vs Q1 = 1,92 (0,72, 5,11)

Langsetmo, 2017

5875 men aged 65 years and older

USA

15 years

Cohort

Protein intake as percent of total energy intake, dairy protein, non-dairy animal protein, and plant protein, assessed by FFQ. Quartiles of protein intake

Follow-up every 4th month and confirmation by physician review of medical records.

Men

Higher total protein intake = decreased risk of hip fracture.

HR = 0,92 (0,84, 1,00)

Higher dairy protein = decreased risk of hip fracture.

HR = 0,80 (0,65, 0,98)

Higher non-dairy animal protein = decreased risk of hip fracture.

HR = 0,84 (0,72, 0,97)

Plant protein = no associations.

HR = 0,99 (0,78, 1,24)

Wengreen, 2004

Men and women aged 50 – 89

1167 cases (831 women, 336 men)

1334 controls (885 women, 449 men)

USA

Case-control

Total protein as percent of energy intake, animal protein and vegetable protein intake assessed by FFQ

Cases of hip fracture collected from 18 Utah hospitals, and controls randomly selected from Utah Drivers License and Medicare databases

Men and women 50 – 69 years old

Higher total protein = decreased risk of hip fracture.

OR Q4 vs Q1 = 0,35 (0,21, 0,59)

Higher animal protein = decreased risk of hip fracture.

OR Q4 vs Q1 = 0,43 (0,22, 0,82)

Vegetable protein = decreased risk

of hip fracture.

OR Q4 vs Q1 = 0,52 (0,27, 0,997)

Men and women 70 – 89 years old

Total protein: no associations.

OR Q4 vs Q1 = 1,28 (0,97, 1,70)

Animal protein: no associations.

OR Q4 vs Q1 = 1,54 (0,93, 2,55)

Vegetable protein: no associations. OR Q4 vs Q1 = 0,79 (0,48, 1,30)

Meyer, 1997

20035 men and 19752 women aged 35 – 49

Norway

11,4 years

Cohort

Non-dairy animal protein assessed by FFQ. Quartiles of protein intake.

Hip fracture reported by follow-up questionnaires, and verification by review of medical records

Women

Quarter of calcium intake:

Q2-Q4

Q1(low)

Q1 (low)

Quarter of non-dairy animal protein intake:

Q4 (high): RR = 0,99 (0,63 – 1,56)

Q1-Q3: RR = 1,07 (0,69 – 1,65)

Q4 (high): RR = 1,96 (1,09 – 3,56)

Men

Quarter of calcium intake:

Q2-Q4

Q1 (low)

Q1 (low)

Quarter of non-dairy animal protein intake:

Q4 (high): RR = 1,27 (0,63 – 2,56)

Q1-Q3: RR = 1,04 (0,50 – 2,16)

Q4 (high): RR = 1,67 (0,64 – 4,39)

Fung, 2017

74443 Caucasian women and 35439 men aged 50 or older

USA

32 years women

26 years men

Cohort

Quintiles of total, animal, dairy, and plant protein assessed by FFQ

Hip fractures self-reported on biennial questionnaires

Women

Per 10 g increase in protein intake:

Total protein: RR = 0,97 (0,93 – 1,02)

Animal protein: RR = 0,97 (0,93 – 1,01)

Plant protein: RR = 0,87 (0,76 – 1,00)

Dairy protein: RR = 0,92 (0,86 – 0,99)

Men

Per 10 g increase in protein intake:

Total protein: RR = 0,92 (0,85 – 0,99)

Animal protein: RR = 0,91 (0,85 – 0,98)

Plant protein: RR = 0,89 (0,73 – 1,10)

Dairy protein: RR = 0,89 (0,79 – 1,00)

Case-control study

The case-control study (29) showed that the risk of hip fracture decreased with increasing protein intake as percentage of total energy intake for participants 50 – 69 years of age. The protective effect of protein was observed for both total, animal and vegetable protein intake.

Discussion

The nine observational studies (13 – 15, 17 – 21, 25) indicated no or potentially a small positive effect of protein intake on BMD. There seems to be a more positive effect in older populations compared to studies of young participants. Two of the RCTs (16, 22) showed no positive associations between protein intake and BMD, while the two others (23, 24) found a positive effect of milk basic protein on BMD of the lumbar spine.

Results from the seven observational studies (26 – 32) investigating the effects of protein intake on hip fracture, indicated that a higher total protein intake may have a protective effect on risk of hip fracture. Due to the uncertain effect of protein intake om BMD, it could indicate that the effect of protein intake on the risk of hip fracture could be affected by other factors than BMD, such as body weight, muscle strength, risk and consequence of falling.

Potential mechanisms

Several mechanisms have been proposed to contribute to the effects of protein intake on bone health. First, protein makes up about 50 % of the bone volume and one third of the bone mass (33). In addition, important effects of protein intake include maintaining bone structure and increasing insulin-like growth factor 1 (IGF-1) that promotes bone formation (34). Moreover, inadequate protein intake in the diet may lead to a negative nitrogen balance, which could negatively affect bone formation (35).

Protein intake also affects muscle mass and strength, which are important determinants of the risk and consequence of falling (36). A state of malnutrition and low bodyweight is also considered to increase the risk of hip fractures due to reduced movement ability, impaired reaction time and balance, and reduced muscle strength, which increase the risk of falling (37). A low bodyweight and reduced muscle mass also leads to a reduction in the layer of soft tissue protecting the skeletal, and thereby increases the risk of hip fractures when falling (37).

Historically, it was proposed that high protein intake could have negative effects by causing a systemic acid load, leading to increased bone resorption to buffer the acid load (20). In addition, it has been suggested that high protein intake could elevate urinary calcium excretion and produce a negative calcium balance, and thereby negatively affect BMD and bone health (38). However, recent findings do not support these negative effects of a high protein intake (30, 39, 40).

In addition to total amount of protein, it has been proposed that different protein sources could have different effects on bone health (41). One of the proposed mechanisms behind this hypothesis is that protein sources with different amino acid content could have different effects on BMD (34), and there has been found associations between intakes of specific amino acids, including alanine and glycine, and higher BMD (34). However, the included studies comparing different protein sources (13 – 24) do not support that consumption of either animal protein or vegetable protein is more advantageous for BMD than the other. These results are in line with earlier meta-analyses (41, 42) that showed no significant difference between the effect of animal protein and vegetable protein on BMD. Similar results are present for the risk of hip fracture (26 – 30), and are also supported by previous meta-analyses (39, 43).

Comparison with other reviews

There has been performed a number of systematic reviews that addresses both risk of hip fracture and BMD (39, 40, 44). Addressing BMD and protein intake, a recent meta-analysis by Shams-White et al. (40) concluded that a higher protein intake might cause less BMD loss of the lumbar spine compared to a lower protein intake in elderly, while Darling et al. (39) found a small positive association between protein intake and BMD. The included study populations and the exposures used in the meta-analysis by Darling et al. (39) had similar characteristics as in the present literature review, while in Shams-White et al. (40) they included several RCTs involving weight-loss programs and studies including subjects with a recent fracture. Although the study populations affectedthe conclusions drawn in these reviews, they presented similar findings regarding the effect of protein intake on BMD as in the present.

In systematic reviews on the risk of hip fractures, a meta-analysis by Wu et al. (43) found that a higher protein intake had a statistically significant protective effect on hip fractures compared to a lower protein intake, but there was no difference between the effect of animal or plant protein intake (43). The conclusion supports the results from the present review.

In contrast, meta-analyses by Darling et al. (39) and Shams-White (40) found no associations between dietary protein and risk of hip fracture. The included studies differed in the duration, age range and size of the study populations, and analyses and this might explain the different conclusion compared to the present review.

Strengths and limitations

A strength of this study is that the review is based on longitudinal data, which gives the opportunity to investigate the long-term effects of protein intake on the outcomes of interest.

In addition, this literature review has a great variation in study populations, with a wide range of age, gender and ethnicities, which makes the results applicable and thereby more generalizable to a larger population. This is different from other reviews that include only older populations with age above 65 years.

However, this variation in study population does not take into account the possible different effects of protein intake across genders and key stages in a lifetime that affects bone development and loss, like adolescence, menopause and old age, which may be considered a limitation.

A second limitation is that the exposures, amounts and sources of protein vary between the different studies. In addition, the quantification and categorization of protein intake varied. This may influence the observed effects and makes it more difficult to compare the results across studies.

Another possible limitation is that a large proportion of the studies included are observational studies. Observational studies can be affected by confounding, but at the same time these studies are useful to investigate long term effects of an exposure, in this case the effect of protein intake. RCTs included also have their limitations, by only studying a limited time period compared to a lifelong exposure.

Conclusion

In conclusion, no harmful effect of a high protein intake and potentially a small beneficial effect of protein intake on BMD was observed, especially in the older populations. The results indicated a protective effect of a higher total protein intake on the risk of hip fracture. Fractures are the most important outcome from a clinical point of view, and studies with longer duration and large populations are necessary to assess the potential long-term effects of protein intake.

Conflict of interest: The authors declare no conflict of interest.

References

  1. Florencio-Silva R, Sasso GRdS, Sasso-Cerri E, Simões MJ, Cerri PS. Biology of Bone Tissue: Structure, Function, and Factors That Influence Bone Cells. BioMed Res Int. 2015;2015:421746.

  2. Bonjour JP, Ammann P, Rizzoli R. Importance of Preclinical Studies in the Development of Drugs for Treatment of Osteoporosis: A Review Related to the 1998 WHO Guidelines. Osteoporos Int. 1999;9(5):379 – 93.

  3. Reginster J-Y, Burlet N. Osteoporosis: A still increasing prevalence. Bone. 2006;38(2, Supplement 1):4 – 9.

  4. LeBlanc KE, Muncie HL, Jr., LeBlanc LL. Hip fracture: diagnosis, treatment, and secondary prevention. Am Fam Physician. 2014;89(12):945 – 51.

  5. Gallagher JC, Tella SH. Prevention and treatment of postmenopausal osteoporosis. J Steroid Biochem Mol Biol. 2014;142:155 – 70.

  6. Weaver CM, Gordon CM, Janz KF et al. The National Osteoporosis Foundation’s position statement on peak bone mass development and lifestyle factors: a systematic review and implementation recommendations. Osteopor Int. 2016;27(4):1281 – 386.

  7. Reid DM. Introduction. Handbook of Osteoporosis. Tarporley: Springer Healthcare Ltd.; 2011. p. 1 – 11.

  8. Cauley JA. Defining ethnic and racial differences in osteoporosis and fragility fractures. Clin Orthop Relat Rese. 2011;469(7):1891 – 9.

  9. Kanis JA, McCloskey EV, Johansson H et al. European guidance for the diagnosis and management of osteoporosis in postmenopausal women. Osteopor Int. 2013;24(1):23 – 57.

  10. Shah M. Role of Nutrition in Musculoskeletal Health. In: Sinaki M, Pfeifer M, editors. Non-Pharmacological Management of Osteoporosis: Exercise, Nutrition, Fall and Fracture Prevention. Cham: Springer International Publishing; 2017. p. 53 – 7.

  11. Thorpe MP, Evans EM. Dietary protein and bone health: harmonizing conflicting theories. Nutr Rev. 2011;69(4):215 – 30.

  12. Moher D, Liberati A, Tetzlaff J, Altman DG et al. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLOS Med. 2009;6(7):e1000097.

  13. Beasley JM, Ichikawa LE, Ange BA et al. Is protein intake associated with bone mineral density in young women? Am J Clin Nutr. 2010;91(5):1311 – 6.

  14. Beasley JM, LaCroix AZ, Larson JC et al. Biomarker-calibrated protein intake and bone health in the Women’s Health Initiative clinical trials and observational study. Am J Clin Nutr. 2014;99(4):934 – 40.

  15. Hu T, Rianon NJ, Nettleton JA et al. Protein intake and lumbar bone density: the Multi-Ethnic Study of Atherosclerosis (MESA). Br J Nutr. 2014;112(8):1384 – 92.

  16. Kerstetter JE, Bihuniak JD, Brindisi J et al. The Effect of a Whey Protein Supplement on Bone Mass in Older Caucasian Adults. J Clin Endocrinol Metab. 2015;100(6):2214 – 22.

  17. Promislow JH, Goodman-Gruen D, Slymen DJ, Barrett-Connor E. Protein consumption and bone mineral density in the elderly : the Rancho Bernardo Study. Am J Epidemiol. 2002;155(7):636 – 44.

  18. Rapuri PB, Gallagher JC, Haynatzka V. Protein intake: effects on bone mineral density and the rate of bone loss in elderly women. Am J Clin Nutr. 2003;77(6):1517 – 25.

  19. Sahni S, Broe KE, Tucker KL et al. Association of total protein intake with bone mineral density and bone loss in men and women from the Framingham Offspring Study. Public health nutr. 2014;17(11):2570 – 6.

  20. Langsetmo L, Barr SI, Berger C et al. Associations of Protein Intake and Protein Source with Bone Mineral Density and Fracture Risk: A Population-Based Cohort Study. J Nutr Health Aging. 2015;19(8):861 – 8.

  21. Hannan MT, Tucker KL, Dawson?Hughes B et al.. Effect of Dietary Protein on Bone Loss in Elderly Men and Women: The Framingham Osteoporosis Study. J Bone Mineral Res. 2000;15(12):2504 – 12.

  22. Zhu K, Meng X, Kerr DA, et al. The effects of a two?year randomized, controlled trial of whey protein supplementation on bone structure, IGF?1, and urinary calcium excretion in older postmenopausal women. J Bone Min Res. 2011;26(9):2298 – 306.

  23. Aoe S, Koyama T, Toba Y, Itabashi A, Takada Y. A controlled trial of the effect of milk basic protein (MBP) supplementation on bone metabolism in healthy menopausal women. Osteoporos Int. 2005;16(12):2123 – 8.

  24. Uenishi K, Ishida H, Toba Y, Aoe S, Itabashi A, Takada Y. Milk basic protein increases bone mineral density and improves bone metabolism in healthy young women. Osteopor Int. 2007;18(3):385 – 90.

  25. Steell L, Sillars A, Welsh P et al. Associations of dietary protein intake with bone mineral density: An observational study in 70,215 UK Biobank participants. Bone. 2019;120:38 – 43.

  26. Sahni S, Cupples LA, McLean RR et al. Protective effect of high protein and calcium intake on the risk of hip fracture in the Framingham offspring cohort. J Bone Mineral Res. 2010;25(12):2770 – 6.

  27. Misra D, Berry SD, Broe KE et al. Does dietary protein reduce hip fracture risk in elders? The Framingham Osteoporosis Study. Osteoporos int. 2011;22(1):345 – 9.

  28. Munger RG, Cerhan JR, Chiu BC. Prospective study of dietary protein intake and risk of hip fracture in postmenopausal women. Am J Clin Nutr. 1999;69(1):147 – 52.

  29. Wengreen HJ, Munger RG, West NA, Cutler DR, Corcoran CD, Zhang J, et al. Dietary protein intake and risk of osteoporotic hip fracture in elderly residents of Utah. J Bone Mineral Res. 2004;19(4):537 – 45.

  30. Langsetmo L, Shikany JM, Cawthon PM et al. The Association Between Protein Intake by Source and Osteoporotic Fracture in Older Men: A Prospective Cohort Study. J Bone Mineral Res. 2017;32(3):592 – 600.

  31. Fung TT, Meyer HE, Willett WC, Feskanich D. Protein intake and risk of hip fractures in postmenopausal women and men age 50 and older. Osteoporos int. 2017;28(4):1401 – 11.

  32. Meyer HE, Pedersen JI, LØken EB, Tverdal A. Dietary Factors and the Incidence of Hip Fracture in Middle-aged Norwegians: A Prospective Study. American Journal of Epidemiology. 1997;145(2):117 – 23.

  33. Heaney RP. Effects of protein on the calcium economy. International Congress Series. 2007;1297:191 – 7.

  34. Amy J, Alexander M, Tim S, Aedín C. Amino Acid Intakes Are Associated With Bone Mineral Density and Prevalence of Low Bone Mass in Women: Evidence From Discordant Monozygotic Twins. J Bone Mineral Res. 2016;31(2):326 – 35.

  35. Koutsofta I, Mamais I, Chrysostomou S. The effect of protein diets in postmenopausal women with osteoporosis: Systematic review of randomized controlled trials. J Women Aging. 2018:1 – 23.

  36. Bonjour J-P, Chevalley T, Amman P, Rizzoli R. Protein Intake and Bone Health. In: Holick MF, Nieves JW, editors. Nutrition and Bone Health. New York, NY: Springer New York; 2015. p. 301 – 17.

  37. Bonjour JP, Schurch MA, Rizzoli R. Nutritional aspects of hip fractures. Bone. 1996;18(3, Supplement 1):S139-S44.

  38. Abelow BJ, Holford TR, Insogna KL. Cross-cultural association between dietary animal protein and hip fracture: A hypothesis. Calcified Tissue Int. 1992;50(1):14 – 8.

  39. Darling AL, Millward DJ, Torgerson DJ, Hewitt CE, Lanham-New SA. Dietary protein and bone health: a systematic review and meta-analysis. Am J Clin Nutr. 2009;90(6):1674 – 92.

  40. Shams-White MM, Chung M, Du M et al. Dietary protein and bone health: a systematic review and meta-analysis from the National Osteoporosis Foundation. Am J Clin Nutr. 2017;105(6):1528 – 43.

  41. Shams-White MM, Chung M, Fu Z et al. Animal versus plant protein and adult bone health: A systematic review and meta-analysis from the National Osteoporosis Foundation. PloS One. 2018;13(2):e0192459.

  42. Shams-White M, Sackey J, Fu Z et al. Protein Intake and Bone Mineral Density- a Systematic Review and Meta-Analysis of Randomized Controlled Trials. FASEB J. 2016;30(1_supplement):678.6-.6.

  43. Wu A-M, Sun X-L, Lv Q-B et al. The Relationship between Dietary Protein Consumption and Risk of Fracture: a subgroup and dose-response meta-analysis of prospective cohort studies. Scientific Reports. 2015;5:9151.

  44. Wallace TC, Frankenfeld CL. Dietary Protein Intake above the Current RDA and Bone Health: A Systematic Review and Meta-Analysis. J Am Coll Nutr. 2017;36(6):481 – 96.