Skip to main content
PMC home page
Scientific Reports logoLink to Scientific Reports
. 2015 Mar 16;5:9151. doi: 10.1038/srep09151

The Relationship between Dietary Protein Consumption and Risk of Fracture: a subgroup and dose-response meta-analysis of prospective cohort studies

Ai-Min Wu 1,*, Xiao-Lei Sun 2,*, Qing-Bo Lv 1, Yong Zhou 1, Dong-Dong Xia 3, Hua-Zi Xu 1, Qi-Shan Huang 1,a, Yong-Long Chi 1
PMCID: PMC5376209  PMID: 25779888

Abstract

It is still debate of the relationship between the dietary protein consumption and risk of fracture. We searched Medline and Embase to assess the effects of dietary protein consumption on risk of fracture. Twelve prospective cohort studies with 407,104 participants were included, higher total protein consumption may be decrease 11% risk of hip fractures, with adj. RR of 0.89 (0.82, 0.97), no significant difference was found for total protein and risk of all fractures and limb fracture; for animal protein consumption and risk of all fractures and hip fracture, with adj.RR of 0.79 (032, 1.96) and 1.04 (0.70, 1.54); for vegetable protein consumption and risk of all fractures, hip fracture and limb fractures with adj.RR of 0.77 (0.52, 1.12), 1.00 (0.53, 1.91), and 0.94 (0.40, 2.22), the subgroup of vegetable protein consumption and risk of all fractures of postmenopausal women with adj.RR of 0.78(0.52,1.16). Dose-response meta-analysis the relationship of total/animal/vegetable protein and hip fracture was consistent to the results of forest plot, the line of total protein and hip fracture was below the Y = 1.0 line. This meta-analysis showed that total dietary protein consumption may be decrease the risk of hip fracture, but not for animal or vegetable protein.

Fracture is a significant cause of morbidity and mortality, especial for aged patients, and it is a challenging global burden1,2,3. One study predicted that may be the number of hip fractures will rise to about 6.26 million world widely at 20504. How to prevent to fracture is a big issue among current researchers and doctors.

Protein is one of important factors that involved in bone metabolism. Beasley reported5 that the higher protein consumption could decrease the risk of hip and forearm fracture, and some other studies6,7 reported higher protein consumption was not associated with a decrease of fracture. Feskanich et al8 reported the high protein consumption may be increase the forearm fracture; overall, the reports were inconsistent.

Another problem is that may be different source of protein may be effect the risk of fracture. It was reported that animal protein might have a greater negative effect on bone health than vegetable protein9, because animal protein increase the urinary calcium excretion. However, the results were inconsistent to others studies10,11. Therefore, the relationship between protein consumption and risk of fracture was still debate.

The aim of this review is to evaluate the evidence from prospective studies on the relation between protein consumption and the risk of fracture, and to subgroup evaluate animal protein and vegetable protein consumption and the risk of fracture. To clear the risk of different site fracture, we evaluate the fracture by all fractures, hip fracture, vertebral fracture and limb fracture.

Methods

The present study was accorded to the preferred reporting items for systematic review and meta-analyses (PRISMA) guidelines (Checklist S1)12.

Search strategy

We searched the database of Medline and Embase on July 20, 2014, using the Key words of “dietary protein,” or “dietary animal protein,” or “dietary vegetable protein,” or “dietary plant protein” and “fracture,” or “hip fracture,” or “vertebral fracture,” or “limb fracture”. The function of “related article” was also used for search. The references of retrieved articles were manually searched to avoid initial miss.

Selection criteria

Studies were included in this meta-analysis according to the following criteria: 1) designed as a prospective cohort study; 2) the exposure of interest in protein consumption or animal protein consumption or vegetable/plant protein consumption; 3) the primary outcome of interest in all fractures of the whole body or hip fracture or vertebral fracture or limb fracture; 4) the relative risk (RR) estimates with 95% confidence intervals (CI) were reported or could be calculated by data reported. If the data were duplicated and reported in more than one study, only the study of the largest number of cases was included. All potential studies were reviewed independently for eligibility by two authors (AMW and ZY), and any disagreement was discussed and resolved with the third independent author (XLS). If the data of dose, case of fracture and person-years could be extracted, it will be included into dose-response meta-analysis.

Data extraction

Two reviewers (QSH and ZYH) independently extracted data for analysis, and the third reviewer checked the consistency between them. A standard data extracted form was used, including the first author's last name, publication year, sample size, country where the study was performed, the gender and age of participants, measure or exposure (Total protein or animal protein or vegetable protein), variables adjusted for analysis, and RR estimates with corresponding 95% CIs for each category of protein. If there were two or more RRs of different potential confounders, we extracted the RRs that reflected the greatest degree of control for potential confounders. If necessary, the primary authors were contacted to retrieve additional information. The study quality was assessed by using the Nine-Star Newcastle-Ottawa Scale13.

Statistical analysis

The methods of statistical analysis of this study are refer to previous similar studies14,15. In STATA software, on the use of fixed effects model and random effects model for homogeneity data have the same results, therefore, we combine Study-specific RR using a random effects model, which considers both between study variation and within study16.

The protein was divided into three types: total protein, animal protein and vegetable protein; and the fractures were divided into four types: all fractures of whole body, hip fracture, vertebral fracture and limb fracture. If there were more than two studies report the same type protein and the risk of the same type fracture, the data will be pooled into meta-analysis, and the highest protein consumption category (Quartile or Tertile) vs. the lowest category were pooled for synthesis. If data of dose, case of fracture and person-years could be extracted from more than 2 studies about the relationship between one type protein consumption and risk of one type fracture, dose-response meta-analysis will be performed to analyze the relation of them. The method of dose-response meta-analysis was according to Orsini and colleagues, whereas the methods of random-effects meta regression models were according to Greenland and colleagues17,18.

In subgroup of vegetable protein consumption for the risk of all fractures of whole body, two studies19,20 reported the menopause status of women; therefore, we pooled them in another subgroup meta-analysis for postmenopausal women only.

Q and I2 statistics were used to evaluate the statistical heterogeneity21. Sensitivity analysis involved removing one study and evaluating whether the rest results would be markedly affected. Potential publication bias was evaluated by the method of Egger's regression asymmetry test22. All statistical tests were performed with the STATA software (version 12.0; StataCorp, College Station, TX, USA).

Results

Literature search

The selection of literature for included studies is shown in Figure 1, total of 1071 potential records were identified from the databases, and 162 duplicated articles were excluded first, then the 836 articles was excluded by abstract screen, 73 full articles were retrieved, at last, 12 prospective cohort studies included for synthesis and meta analysis5,6,8,11,19,20,23,24,25,26,27,28 were included for meta analysis.

Figure 1. The selection of literatures for included studies.

Figure 1

Open in a new tab

Study characteristics

The characteristics of the included dietary protein consumption studies are showed in Table 1, and the dose of protein consumption of included studies were showed in Table 2. The total number of participants is 407,104 with 41,659 all fractures of the whole body, 4,000 hip fractures, and 9,599 limb fractures in the included 12 studies. Nine studies were conducted in the United States and Canada, two in Europe, and one in Asia. The Nine-Star Newcastle-Ottawa Scale results of included studies were showed in Table 3.

Table 1. Characteristics of Prospective Studies on Protein intake and Risk of Fracture.

Author Year No. of participants Location/Period Gender Age (years) No. of casesa Measure/Exposure Study Qualityb Adjustment for Covariatesc
Mussolino et al. 1998 2,879 United States M 45–74 HF: 79 Total protein 7 Smoking status, alcohol consumption, physical activity, chronic disease, calcium intake, calories, weight loss, bronchitis, thyroid disease, diabetes, kidney disease, coronary heart disease and stroke.
Beasley et al. 2014 144,580 United States F 50–79 AF:36,166 HF:3,286 VF:4,836 LF:7,800 Total protein 7 Age, BMI, race-ethnicity, calibrated energy intake, general health, physical activity, history of fracture, history of parental fracture, smoking, corticosteroid, glucocorticoid use, treated diabetes, rheumatoid arthritis, and hormone use.
Meyer et al. 1997 39,787 Norway F:19,752 M:20,035 35–49 HF:213 Animal Protein 7 Age, body height, BMI, serf-reported physical activity at work and during leisure time, diabetes mellitus, disability pension, marital status, and smoking.
Misra et al. 2011 946 United States F:576 M:370 28–62 HF:100 Total Protein 7 Age, sex, weight, height and total energy intake.
Thorpe et al. 2007 1,865 United States and Canada F Post-menopausal or >45 Y LF:171 Vegetable Protein 6 Education, BMI, practitioner-diagnosed medical conditions, coronary heart disease, stroke, high blood pressure, diabetes, diverticulitis, cancer, rheumatoid arthritis, other arthritis, alcohol use, smoking, nulliparity, menopausal status, age at menopause, hormone use and physical activity.
Nieves et al. 2010 125 United States F 18–26 AF: 17 Animal protein Vegetable Protein 6 Baseline nutrient intake, beverage consumption, dietary patterns, treatment group assignment, Menstrual status, spine bone density, age, and fracture history.
Zhang et al. 2005 24,403 China F 40–70 AF:1,170 Vegetable Protein 7 Age, BMI, hours of exercise, cigarette smoking, alcohol consumption, diabetes mellitus, education, family income, season of recruitment, calories intakes, calcium, fruits, and vegetables.
Munger et al. 1999 32,050 United States F 55–69 HF: 44 Total Protein Animal Protein Vegetable Protein 7 Age, BMI, number of pregnancies, smoking, alcohol use, estrogen use, and physical activity.
Sahni et al. 2010 3,656 United States F:1931 M:1725 26–86 HF: 44 Total Protein Animal Protein Vegetable Protein 7 Sex, menopause status, age, weight and height at baseline, physical activity index, intake of energy, vitamin D, smoking status, energy intake, dietary calcium intake.
Feskanich et al. 1996 85,900 United States F 35–59 HF:234 LF:1,628 Total Protein Animal Protein Vegetable Protein 6 Questionnaire time period; Age; BMI, hours of activity; menopause status and HT use; cigarette smoking; use of thyroid hormone medication and thiazlde diuretics; alcohol and caffeine Intakes.
Dargent-Molina et al. 2008 36,217 France F 40–65 AF: 2,408 Total Protein Animal Protein Vegetable Protein 7 BMI, physical activity, parity, maternal history of hip fracture, HT use, smoking status, and alcohol intake
Key et al. 2007 346,96 United Kingdom F:26,749 M:7,947 20–89 AF:1,898 Total Protein 6 Age, smoking, intakes of energy and each other nutrient, alcohol consumption, BMI, walking, cycling, vigorous exercise, other exercise, physical activity at work, marital status and, for women, parity and HT use.
Open in a new tab

Note: a: AF: All fracture; HF: Hip fracture; VF: Vertebral fracture; LF: Limb fracture.

b: Study quality was judged based on the Newcastle-Ottawa Scale (range, 1–9 stars).

c: BMI: Body mass index; HT use: Hormone replacement therapy.

Table 2. The dose of different protein consumption of included studies.

Study   Total protein Animal protein Vegetable Protein Quartile
Mussolino et al. 1998 (g/day) highest dose >98 - - Q4
  lowest dose <56 - - Q1
Beasley et al. 2014 (g/day) highest dose 20% increased - -  
  lowest dose - - -  
Meyer et al. 1997 (g/day) highest dose - >20.6(Woman) >21.6 (Man) - Q4
  lowest dose - <13.6(Woman) <14.2(Man) - Q1
Misra et al. 2011 (g/day) highest dose 82.74 ± 10.27 - - Q4
  lowest dose 46.45 ± 7.29 - - Q1
Thorpe et al. 2007 (Eight level food frequency) highest dose >1/day - - Q3
  lowest dose <3/week - - Q1
Nieves et al. 2010 (g/day/kg) highest dose - 1 g/day/kg increased 1 g/day/kg increased -
  lowest dose - - - -
Zhang et al. 2005 (g/day) highest dose - - >13.27 Q5
  lowest dose - - <4.98 Q1
Munger et al. 1999 (g/day) highest dose >95.5 >75.14 >26.2 Q4
  lowest dose <67.38 <43.74 <17.5 Q1
Sahni et al. 2010 (g/day) highest dose - 68 29 Q3
  lowest dose - 38 18 Q1
Feskanich et al. 1996 (g/day) highest dose >95 >80 >19 Q5
  lowest dose <68 <51 <12 Q1
Dargent-Molina et al. 2008 (g/1000 kcal) highest dose >50.11 >33.52 >14.12 Q4
  lowest dose <40.75 <22.42 <10.07 Q1
Key et al. 2007 (g/day) highest dose >90 - - Q5
  lowest dose <55 - - Q1
Open in a new tab

Table 3. Assessment of quality of included studies on the use of Nine-Star Newcastle-Ottawa Scale.

     Selection      Outcome assessment
Study (authors, year) Representativeness of the exposed cohort Selection of the nonexposed cohort Ascertainment of exposure Incident disease Comparability Assessment of outcome Length of follow up Adequacy of follow up Score
Mussolino et al. 1998 * * - - ** * * * *******
Beasley et al. 2014 * * - * ** - * * *******
Meyer et al. 1997 * * - * * * * * *******
Misra et al. 2011 * * - - ** * * * *******
Thorpe et al. 2007 * * - - ** - * * ******
Nieves et al. 2010 - * - * ** * - * ******
Zhang et al. 2005 * * * - ** - * * *******
Munger et al. 1999 - * - * ** * * * *******
Sahni et al. 2010 * * - - ** * * * *******
Feskanich et al. 1996 - * - - ** * * * ******
Dargent-Molina et al. 2008 * * - - ** * * * *******
Key et al. 2007 - * - - ** * * * ******
Open in a new tab

Note: One asterisk means one score, studies with more scores on behalf of higher quality.

Dietary total protein consumption and risk of fracture

Only one study5 concerns the relation between dietary total protein consumption and risk of vertebral fracture, therefore, cannot reach a meta-analysis. The total protein consumption and risk of all fractures of the whole body, hip fractures and limb fracture are shown in Figure 2. Our present meta-analysis of highest vs. lowest category shows that the adjusted relative risk (adj.RR) of total protein consumption for hip fractures is 0.89 (0.82, 0.97), has a statistic significantly difference, and decrease 11% risk of hip fractures. Others, for all fractures is 0.99 (0.97, 1.02), limb fractures is 1.05 (0.81, 1.37), no significant difference is found. Heterogeneity is observed at subgroup study of total protein consumption and risk of limb fractures (I2 = 90.0%, P = 0.002).

Figure 2. Adjusted Relative Risk of fracture (all or hip or limb fracture) for the highest vs. the lowest category of total dietary protein consumption.

Figure 2

Open in a new tab

Dietary animal protein consumption and risk of fracture

No study reported the dietary animal protein consumption and risk of vertebral fracture, only one study reported the dietary animal protein consumption and risk of limb fracture8, therefore, cannot reach meta-analysis of above two indications. The highest vs. lowest category shows that adj.RR of animal protein consumption for risk of all fractures of whole body is 0.79 (0.32, 1.96), and for risk of hip fractures is 1.04 (0.70, 1.54). Heterogeneity is observed at studies of risk of all fractures (I2 = 69.8%, P = 0.069), of hip fracture (I2 = 51.6%, P = 0.083). (Figure 3)

Figure 3. Adjusted Relative Risk of fracture (all or hip or limb fracture) for the highest vs. the lowest category of dietary animal protein consumption.

Figure 3

Open in a new tab

Dietary vegetable protein consumption and risk of fracture

No study reported the dietary vegetable protein consumption and risk of vertebral fracture. The adj.RR of highest vs. lowest category of all fractures is 0.77 (0.52, 1.12), of hip fractures is 1.00 (0.53, 1.91), of limb fractures is 0.94 (0.40, 2.22). Heterogeneity is observed at studies of risk of all fractures (I2 = 86.4%, P = 0.001), of hip fracture (I2 = 56.9%, P = 0.098), of limb fracture (86.1%, P = 0.001). Two studies reported the vegetable protein consumption of postmenopausal women and risk of all fractures, the adj.RR of subgroup meta-analysis of postmenopausal women only is 0.78 (0.52, 1.16, I2 = 93.1%, P = 0.000). (Figure 4)

Figure 4. Adjusted Relative Risk of fracture (all or hip or limb fracture) for the highest vs. the lowest category of dietary vegetable protein consumption.

Figure 4

Open in a new tab

Dose-response meta-analysis

Only the data of three sub-studies (total protein intake and risk of hip fracture, animal protein intake and risk of hip fracture, vegetable protein intake and risk of hip fracture) meet dose-response meta-analysis. The adj.RR of total protein intake and risk of hip fracture is below the line of RR = 1 (Figure 5A), others two adj.RRs of animal protein intake and risk of hip fracture, vegetable protein intake and risk of hip fracture is spanning the line of RR = 1 (Figure 5B/C). The result is consistent to the forest plot of Figure 2, 3 and 4.

Figure 5. Dose-response relationship between total protein (A) or animal protein (B) or vegetable protein (C) and relative risk of hip fracture.

Figure 5

Open in a new tab

Solid line represents adjusted relative risk and dotted lines represent the 95% confidence intervals for the fitted trend. The adj.RR of total protein intake and risk of hip fracture is below the line of RR = 1 (A), others two adj.RRs of animal protein intake and risk of hip fracture, vegetable protein intake and risk of hip fracture is spanning the line of RR = 1 (B and C). The result is consistent to the forest plot of Figure 2, 3 and 4.

Sensitivity analysis and publication bias

The results of sensitivity analysis suggest that either the study of Dargent-Molina et al.20 or Zhang et al.19 omitted could decrease the heterogeneity of subgroup meta analysis of vegetable protein consumption and risk of all fractures, however, the studies of Dargent-Molina et al.20 or Zhang et al.19 both reported the risk of postmenopausal women, however, the study of Nieves et al.27 is not about postmenopausal women, therefore, combine study of Nieves et al.27 to either Dargent-Molina et al.20 or Zhang et al.19 is unreasonable. Therefore, we only did subgroup meta-analysis of postmenopausal women. The influence of each individual data set to the pooled RRs is not significant for all of other subgroup meta analysis (Supplemental Figures File 1). The Egger's test shows no evidence of publication bias of the total protein for all fractures or hip fracture (P = 0.286; P = 0.054), animal protein for hip fractures (P = 0.855), vegetable protein for all fractures or hip fractures or limb fractures (P = 0.701; P = 0.905; P = 0.949). Only two studies included in the subgroup meta-analysis of total protein for limb fractures and animal protein for all fractures, therefore, the Egger's test is error for them and the P value of Begg's test is P = 1.000 for both of them.

Discussion

Fracture is a major global health problem. Many factors were supposed to decrease or increase the risk of fracture, such as age, BMD, physical activity, smoke, calcium, Vitamin D, Vitamin A and Vitamin K29,30,31,32,33. Protein is an important source of amino acid which to maintain bone structure, or stimulate some growth factors such as insulin-like growth factor I (IGF-I), then to increase the activity of osteoblast and the mineralization of bone matrix34,35, and the inadequate dietary protein may influence the bone strength and increase risk of fracture36,37. Some other concerns the relationship of high protein consumption and bone health are: 1) may be the protein will increase urinary calcium; 2) the protein could increase intestinal calcium absorption; 3) high dietary protein consumption may be act indirectly through preservation of muscle, and decrease falls and fractures38,39,40. However, the associate between the dietary protein consumption and risk of fracture is still dispute.

In 2009, Darling et al.7 meta analyzed the associate between the dietary protein and risk of hip fractures, three studies of total protein, three animal protein and two vegetable protein prospective reports were included by their study, and no associate between of dietary protein and risk of hip fracture was found at that time. In our present meta-analysis, added the recent publications, six studies of total protein, four of animal protein and three of vegetable protein prospective studies are pooled for analysis. We find that adjusted relative risk (adj.RR) of total protein consumption for hip fractures is 0.89 (0.82, 0.97), has a statistic significantly decrease 11% risk of hip fracture.

However, no benefit is found at meta-analysis of total protein for all fractures of the whole body and limb fracture. May be the included studies of them are still too small, only two or three different reports. Moreover, the hip fracture is more fragility than other sites, especially for aged participates; therefore, it may be prior fracture than other sites.

Some reports suggested that the different source of protein from animal or vegetable will be effect the risk of fracture varies. Sellmeyer et al.9 reported that more vegetable protein intake and less animal protein intake may decrease bone loss and the risk of hip fracture, however, Hannan et al.41 reported that higher animal protein consumption was not associated with a decrease in bone mineral density. In study of Munger et al.10, the higher animal protein intake had a lower risk of fracture than the lower animal protein intake category. In this meta-analysis, no difference is found by subgroup meta-analysis of animal protein and vegetable protein for all fractures, hip fracture and limb fracture. Because the higher total protein consumption is benefit for risk of hip fracture, may be this benefit is doesn't matter what the protein source from animal or vegetable.

The strength of our present meta-analysis study is that our quantitative assessment is based on prospective studies, compared to retrospective and case-control studies, these prospective studies minimizes the possibility of the recall and selection bias. In 2009, Darling et al.7 reported a meta-analysis of the associate between the dietary protein and risk of hip fracture; only four prospective studies were included for fracture risk meta-analysis at that time, without dose-response analysis. To the best of our knowledge, this is the first meta-analysis of the relationship between total/animal/vegetable protein and risk of all fractures, hip fracture and limb fracture based on prospective cohort studies, and a quantitative dose-response assessment of the relationship between protein consumption and risk of both hip fractures. Moreover, our study including the large number of participants, long duration of follow-up, and most individual studies are well powered.

There are also many limitations of our present study. Only one study report the risk of vertebral fracture5, therefore, cannot be meta-analysis. For some subgroup meta-analysis, such as total protein and limb fractures, animal protein and all fractures, the included studies are only two, and more prospective studies needed to be taken in future. Only the data of total/animal/vegetable protein and risk of hip fracture is sufficient for dose-response meta-analysis, others don't have enough data, and can't reach a dose-response meta-analysis.

Another limitation of this meta analysis is that: although the significant result data of total protein consumption for hip fracture without heterogeneity (I2 = 0.0%, P = 0.439), many other subgroup meta analysis have significantly heterogeneity, if these data have a significant result, which is suspected, the fortunate is that all of these heterogeneity data do not show any significant results.

Conclusion

Total dietary protein consumption may be decrease the risk of hip fracture, but not for all fractures and limb fracture. No current evidence shows the animal or vegetable protein could decrease or increase the risk of fracture.

Supplementary Material

Supplementary Information

Supplementary file

srep09151-s1.pdf (57.5KB, pdf)
Supplementary Information

Checklist S1

srep09151-s2.doc (68.5KB, doc)
Supplementary Information

Dataset 1

srep09151-s3.xlsx (9.1KB, xlsx)
Supplementary Information

Dataset 2

srep09151-s4.xlsx (9.3KB, xlsx)
Supplementary Information

Dataset 3

srep09151-s5.xlsx (9.5KB, xlsx)
Supplementary Information

Dataset 4

srep09151-s6.xlsx (10.7KB, xlsx)
Supplementary Information

Dataset 5

srep09151-s7.xlsx (10.3KB, xlsx)
Supplementary Information

Dataset 6

srep09151-s8.xlsx (10.3KB, xlsx)

Acknowledgments

This work is funded by National Natural Science Foundation of China (81372014, 81371988); and Xinmiao talent plan of Zhejiang Province (2014R413053). The funders had no role in the design, execution, and writing of the study.

Footnotes

The authors declare no competing financial interests.

Author Contributions A.M.W., X.L.S., Q.S.H. and Y.L.C. contributed to the conception and design of the study. A.M.W., X.L.S., D.D.X., Q.B.L., H.Z.X. and Y.Z. contributed to the analysis and interpretation of data. A.M.W., X.L.S. and Q.S.H. contributed to the drafting of the article. All authors revised the manuscript critically for important intellectual content and gave final approval of the version to be published. A.M.W. and X.L.S. are co-first authors on this study.

References

  1. Johnell O. & Kanis J. A. An estimate of the worldwide prevalence and disability associated with osteoporotic fractures. Osteoporos Int 17, 1726–1733, 10.1007/s00198-006-0172-4 (2006). [DOI] [PubMed] [Google Scholar]
  2. Cummings S. R. & Melton L. J. Epidemiology and outcomes of osteoporotic fractures. Lancet 359, 1761–1767, S0140-6736(02)08657-9 (2002). [DOI] [PubMed] [Google Scholar]
  3. Burge R. et al. Incidence and economic burden of osteoporosis-related fractures in the United States, 2005–2025. J Bone Miner Res 22, 465–475, 10.1359/jbmr.061113 (2007). [DOI] [PubMed] [Google Scholar]
  4. Cooper C., Campion G. & Melton L. J. 3rd Hip fractures in the elderly: a world-wide projection. Osteoporos Int 2, 285–289 (1992). [DOI] [PubMed] [Google Scholar]
  5. Beasley J. M. et al. Biomarker-calibrated protein intake and bone health in the Women's Health Initiative clinical trials and observational study. Am J Clin Nutr 99, 934–940, 10.3945/ajcn.113.076786 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Mussolino M. E., Looker A. C., Madans J. H., Langlois J. A. & Orwoll E. S. Risk factors for hip fracture in white men: The NHANES I Epidemiologic Follow-up Study. Journal of Bone and Mineral Research 13, 918–924, DOI 10.1359/jbmr.1998.13.6.918 (1998). [DOI] [PubMed] [Google Scholar]
  7. Darling A. L., Millward D. J., Torgerson D. J., Hewitt C. E. & Lanham-New S. A. Dietary protein and bone health: a systematic review and meta-analysis. Am J Clin Nutr 90, 1674–1692, 10.3945/ajcn.2009.27799 (2009). [DOI] [PubMed] [Google Scholar]
  8. Feskanich D., Willett W. C., Stampfer M. J. & Colditz G. A. Protein consumption and bone fractures in women. Am J Epidemiol 143, 472–479 (1996). [DOI] [PubMed] [Google Scholar]
  9. Sellmeyer D. E., Stone K. L., Sebastian A. & Cummings S. R. A high ratio of dietary animal to vegetable protein increases the rate of bone loss and the risk of fracture in postmenopausal women. Am J Clin Nutr 73, 118–122 (2001). [DOI] [PubMed] [Google Scholar]
  10. Munger R. G., Cerhan J. R. & Chiu B. C. H. Prospective study of dietary protein intake and risk of hip fracture in postmenopausal women. Am J Clin Nutr 69, 147–152 (1999). [DOI] [PubMed] [Google Scholar]
  11. Sahni S. et al. Protective effect of high protein and calcium intake on the risk of hip fracture in the Framingham offspring cohort. J Bone Miner Res 25, 2770–2776, 10.1002/jbmr.194 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Liberati A. et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol 62, e1–34, 10.1016/j.jclinepi.2009.06.006S0895-4356(09)00180-2 (2009). [DOI] [PubMed] [Google Scholar]
  13. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 25, 603–605, 10.1007/s10654-010-9491-z (2010). [DOI] [PubMed] [Google Scholar]
  14. Larsson S. C., Orsini N. & Wolk A. Vitamin B6 and risk of colorectal cancer: a meta-analysis of prospective studies. JAMA 303, 1077–1083, 10.1001/jama.2010.263303/11/1077 (2010). [DOI] [PubMed] [Google Scholar]
  15. Luo J., Shen L. & Zheng D. Association between vitamin C intake and lung cancer: a dose-response meta-analysis. Sci Rep 4, 6161, 10.1038/srep06161 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. DerSimonian R. & Laird N. Meta-analysis in clinical trials. Control Clin Trials 7, 177–188, 0197-2456(86)90046-2 (1986). [DOI] [PubMed] [Google Scholar]
  17. Orsini N., Li R., Wolk A., Khudyakov P. & Spiegelman D. Meta-analysis for linear and nonlinear dose-response relations: examples, an evaluation of approximations, and software. Am J Epidemiol 175, 66–73, 10.1093/aje/kwr265kwr265 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Greenland S. & Longnecker M. P. Methods for trend estimation from summarized dose-response data, with applications to meta-analysis. Am J Epidemiol 135, 1301–1309 (1992). [DOI] [PubMed] [Google Scholar]
  19. Zhang X. et al. Prospective cohort study of soy food consumption and risk of bone fracture among postmenopausal women. Archives of Internal Medicine 165, 1890–1895 (2005). [DOI] [PubMed] [Google Scholar]
  20. Dargent-Molina P. et al. Proteins, Dietary Acid Load, and Calcium and Risk of Postmenopausal Fractures in the E3N French Women Prospective Study. Journal of Bone and Mineral Research 23, 1915–1922, Doi 10.1359/Jbmr.080712 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Higgins J. & Thompson S. G. Quantifying heterogeneity in a meta-analysis. Statistics in medicine 21, 1539–1558 (2002). [DOI] [PubMed] [Google Scholar]
  22. Egger M., Davey Smith G., Schneider M. & Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 315, 629–634 (1997). [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Key T. J. et al. Calcium, diet and fracture risk: A prospective study of 1898 incident fractures among 34696 British women and men. Public Health Nutrition 10, 1314–1320 (2007). [DOI] [PubMed] [Google Scholar]
  24. Meyer H. E., Pedersen J. I., Loken E. B. & Tverdal A. Dietary factors and the incidence of hip fracture in middle-aged Norwegians: A prospective study. American Journal of Epidemiology 145, 117–123 (1997). [DOI] [PubMed] [Google Scholar]
  25. Misra D. et al. Does dietary protein reduce hip fracture risk in elders? The Framingham osteoporosis study. Osteoporosis Int 22, 345–349, DOI 10.1007/s00198-010-1179-4 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Munger R. G., Cerhan J. R. & Chiu B. C. Prospective study of dietary protein intake and risk of hip fracture in postmenopausal women. Am J Clin Nutr 69, 147–152 (1999). [DOI] [PubMed] [Google Scholar]
  27. Nieves J. W. et al. Nutritional factors that influence change in bone density and stress fracture risk among young female cross-country runners. PM R 2, 740–750; quiz 794, 10.1016/j.pmrj.2010.04.020S1934-1482(10)00340-0 (2010). [DOI] [PubMed] [Google Scholar]
  28. Thorpe D. L., Knutsen S. F., Beeson W. L., Rajaram S. & Fraser G. E. Effects of meat consumption and vegetarian diet on risk of wrist fracture over 25 years in a cohort of peri- and postmenopausal women. Public Health Nutr 11, 564–572, 10.1017/s1368980007000808 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kanis J. A. et al. Smoking and fracture risk: a meta-analysis. Osteoporos Int 16, 155–162, 10.1007/s00198-004-1640-3 (2005). [DOI] [PubMed] [Google Scholar]
  30. De Laet C. et al. Body mass index as a predictor of fracture risk: a meta-analysis. Osteoporos Int 16, 1330–1338, 10.1007/s00198-005-1863-y (2005). [DOI] [PubMed] [Google Scholar]
  31. Wu A. M. et al. The relationship between vitamin a and risk of fracture: meta-analysis of prospective studies. J Bone Miner Res 29, 2032–2039, 10.1002/jbmr.2237 (2014). [DOI] [PubMed] [Google Scholar]
  32. Bischoff-Ferrari H. A. et al. Fracture prevention with vitamin D supplementation: a meta-analysis of randomized controlled trials. JAMA 293, 2257–2264, 293/18/2257/jama.293.18.2257 (2005). [DOI] [PubMed] [Google Scholar]
  33. Qu X. et al. Association between physical activity and risk of fracture. J Bone Miner Res 29, 202–211, 10.1002/jbmr.2019 (2014). [DOI] [PubMed] [Google Scholar]
  34. Millward D. J. & Rivers J. P. The need for indispensable amino acids: the concept of the anabolic drive. Diabetes Metab Rev 5, 191–211 (1989). [DOI] [PubMed] [Google Scholar]
  35. Thorpe M. P. & Evans E. M. Dietary protein and bone health: harmonizing conflicting theories. Nutr Rev 69, 215–230, 10.1111/j.1753-4887.2011.00379.x (2011). [DOI] [PubMed] [Google Scholar]
  36. Bonjour J. P. Dietary protein: An essential nutrient for bone health. Journal of the American College of Nutrition 24, 526S–536S (2005). [DOI] [PubMed] [Google Scholar]
  37. Mangano K. M., Sahni S. & Kerstetter J. E. Dietary protein is beneficial to bone health under conditions of adequate calcium intake: An update on clinical research. Current Opinion in Clinical Nutrition and Metabolic Care 17, 69–74 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Geinoz G. et al. Relationship between bone mineral density and dietary intakes in the elderly. Osteoporosis International 3, 242–248 (1993). [DOI] [PubMed] [Google Scholar]
  39. Sattin R. W. et al. The incidence of fall injury events among the elderly in a defined population. Am J Epidemiol 131, 1028–1037 (1990). [DOI] [PubMed] [Google Scholar]
  40. Tinetti M. E., Speechley M. & Ginter S. F. Risk factors for falls among elderly persons living in the community. N Engl J Med 319, 1701–1707, 10.1056/NEJM198812293192604 (1988). [DOI] [PubMed] [Google Scholar]
  41. Hannan M. T. et al. Effect of dietary protein on bone loss in elderly men and women: The Framingham Osteoporosis Study. Journal of Bone and Mineral Research 15, 2504–2512 (2000). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information

Supplementary file

srep09151-s1.pdf (57.5KB, pdf)
Supplementary Information

Checklist S1

srep09151-s2.doc (68.5KB, doc)
Supplementary Information

Dataset 1

srep09151-s3.xlsx (9.1KB, xlsx)
Supplementary Information

Dataset 2

srep09151-s4.xlsx (9.3KB, xlsx)
Supplementary Information

Dataset 3

srep09151-s5.xlsx (9.5KB, xlsx)
Supplementary Information

Dataset 4

srep09151-s6.xlsx (10.7KB, xlsx)
Supplementary Information

Dataset 5

srep09151-s7.xlsx (10.3KB, xlsx)
Supplementary Information

Dataset 6

srep09151-s8.xlsx (10.3KB, xlsx)

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES