SEARCH

SEARCH BY CITATION

Keywords:

  • PARITY;
  • BREASTFEEDING;
  • HIP FRACTURE;
  • POSTMENOPAUSAL WOMEN;
  • BONE MINERAL DENSITY

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Despite reported bone loss during pregnancy and lactation, no study has shown deleterious long-term effects of parity or breastfeeding. Studies have shown higher bone mineral density and reduced risk for fracture in parous than in nulliparous women or no effect of parity and breastfeeding, so long-term effects are uncertain. We studied the effect of parity and breastfeeding on risk for hip, wrist and non-vertebral fragility fractures (hip, wrist, or proximal humerus) in 4681 postmenopausal women aged 50 to 94 years in the Tromsø Study from 1994–95 to 2010, using Cox's proportional hazard models. During 51 906 person-years, and a median of 14.5 years follow-up, 442, 621, and 1105 of 4681 women suffered incident hip, wrist, and fragility fractures, and the fracture rates were 7.8, 11.4, and 21.3 per 1000 person-years, respectively. The risk for hip, wrist, and fragility fracture did not differ between parous (n = 4230, 90.4%) and nulliparous women (n = 451, 9.6%). Compared with women who did not breast-feed after birth (n = 184, 4.9%), those who breastfed (n = 3564, 95.1%) had 50% lower risk for hip fracture (HR 0.50; 95% CI 0.32 to 0.78), and 27% lower risk for fragility fracture (HR 0.73; 95% CI 0.54 to 0.99), but similar risk for wrist fracture, after adjustment for age, BMI, height, physical activity, smoking, a history of diabetes, previous fracture of hip or wrist, use of hormone replacement therapy, and length of education. Each 10 months longer total duration of breastfeeding reduced the age-adjusted risk for hip fracture by 12% (HR 0.88; 95% CI 0.78 to 0.99, p for trend = 0.03) before, and marginally after, adjustment for BMI and other covariates (HR 0.91; 95% CI 0.80 to 1.04). In conclusion, this data indicates that pregnancy and breastfeeding has no long-term deleterious effect on bone fragility and fractures, and that breastfeeding may contribute to a reduced risk for hip fracture after menopause. © 2011 American Society for Bone and Mineral Research


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Pregnancy and lactation have been thought of as possible risk factors for bone fragility for many reasons. First, women sustain more fragility fractures than men, and bone loss starts in young adulthood.1–5 Second, mineralization of the fetal and neonatal skeleton may call on maternal skeletal calcium stores.6–8 Third, the physiological hypoestrogenemia and hyperprolactinemia during lactation may result in deleterious effects on the skeleton.6, 7 Pregnancy and lactation could thus result in reduced peak bone mass, bone loss, or both.6, 9, 10

During pregnancy, slight bone loss of 1% to 3% is reported in humans,11–14 whereas bone loss of 5% to 10% is reported during lactation.6, 7, 12–17 Whether this bone loss is completely restored or has any long-term effect on the skeleton is not clear, because there are few clinical studies with sufficient follow-up after weaning.13–16 In epidemiological studies, nulliparity is associated with higher risk for fracture and lower bone mineral density (BMD).18–22 No deleterious long-term effect on the skeleton is reported by higher parity. However, there are few studies of the effect of breastfeeding on risk for fracture in a prospective setting, and results from mostly small case-control studies are not clear.20, 23–25 In addition, few population-based studies have sufficient power to detect an independent effect of parity and breastfeeding, if there truly is one.

Therefore, it is not clear whether pregnancy and lactation means higher risk of fragility fracture later in life or protection against fractures. We tested the hypothesis that higher parity and longer duration of breastfeeding is associated with lower risk for hip, wrist, or fragility fracture (hip, wrist, or humerus), in a prospective study of postmenopausal women in Tromsø, Norway.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

The Tromsø Study is a single-center, population-based health study in Northern Norway, which conducted six surveys in 1974, 1979 to 1980, 1986 to 1987, 1994 to 1995, 2001, and 2007 to 2008, respectively.4, 26–28 At the fourth survey in 1994–95, all 37,558 eligible inhabitants in Tromsø, Norway, older than 24 years were invited, and 27,158 subjects (72%) participated (Fig. 1). From the main survey, all men aged 55 to 74 years and all women aged 50 to 74 years and 5% to 10% random samples of the other age groups of both sexes were invited to an extended examination. At the second part of this survey 7,948 (4558 women and 5795 men) of the 10,213 invited subjects (response rate 78%) had distal forearm bone mineral density (BMD) measured.26 All of those who were still living in Tromsø were invited to a distal forearm and hip BMD in 2001, and 5771 (57%) of the original cohort attended. All participants gave informed written consent. The study was approved by the Regional Committee of Research Ethics and the Norwegian Data Inspectorate and was conducted in accordance with the World Medical Association Declaration of Helsinki.

thumbnail image

Figure 1. The Tromsø Study 3 to 6; with all the 27 158 subjects in Tromsø 4 included in the registry of non-vertebral fractures from 1994–95 to 2010, the women ≥ 50 years who had data on parity and were postmenopausal at Tromsø 4 baseline in 1994 to 95, who we included in this study.

Download figure to PowerPoint

Of all 5433 women aged 50 to 94 years at baseline at the fourth survey in 1994–95, 4633 had self-reported information on parity available, and 800 women were missing data; 75, 284, and 53 women had data available from the third, fifth, or sixth survey, respectively. After excluding 388 women with missing data on parity at all four surveys, 5045 women were left. We further excluded premenopausal women who reported a menstrual period within the last year (n = 363) and one woman with a pathological fracture, so 4681 postmenopausal women were included in the main analyses of the effect of parity on the risk for fracture. Among those 4681, 3748 reported their duration of breastfeeding. Because exclusion of 452 women using hormone replacement therapy (HRT) did not change the results, we included them. Among those 4681 women, distal forearm BMD was measured in a subgroup of 3556 women in 1994–95 and the total hip BMD was measured in a subgroup of 2358 women in 2001.

At baseline, two self-administered questionnaires were filled in. If the women had given birth, the number of children they had given birth to, and the months of breastfeeding of each child were reported. They reported whether they were currently smoking (yes/no), teetotaler (yes/no), used HRT (yes/no), had diabetes (yes/no), previous fracture of hip or wrist (yes/no), length of education of more than 7 to 10 years (yes/no), and hours of moderate and hard physical activity during leisure time. A physical activity score was made by adding the hours/week of moderate and hard physical activity, giving the hours with hard activity double weight: score = moderate + 2hard. Definition of premenopausal status was based on the self-reported date for their last menstrual period. No participant reported use of bisphosphonates at baseline.

All non-vertebral fractures were registered from the X-ray archives of the University Hospital of North Norway in Tromsø between their participation in 1994–95 survey and 1 January 2010. All fractures are registered here, because this is the only X-ray service in the city or within 250 km. The only exception would be fractures occurring while traveling with no control X-ray after returning home. The validation of the fracture registration was previously reported.29 Follow-up time was assigned from baseline to the first fracture, to death, to when the participant moved, or to the end of follow-up. In this study we included the first fragility fractures, which we defined as a fracture at the hip, wrist, or proximal humerus. Analyses of fractures at the hip included the first hip fracture that happened during follow-up, with the same procedure for wrist fractures.

Height and weight were measured in light clothing without shoes, and body mass index (BMI) was calculated as weight divided by the square of height (kg/m2). BMD was measured at the non-dominant distal forearm, with Single X-ray Absorptiometric (SXA)-devices (DTX-100; Osteometer MediTech, Inc., Hawthorne, California. USA), and at the total hip by Dual-energy X-ray Absorptiometry (DXA, GE Lunar Prodigy; Lunar Corporation, Madison, WI, USA). The coefficients of variation (CV) were 0.8% and 1.2% for the distal forearm and total hip, respectively; details of the measurement methods, the strict quality control procedures for densitometry, and long-term performance of the densitometers were previously published.26, 27, 30

The SAS Software package, v9.1 (SAS Institute Inc., Cary, NC, USA) was used for statistical analyses. The number of children and total period of breastfeeding showed a skewed distribution. The women were therefore divided into five groups according to parity (nulliparous, 1, 2, 3, 4 or more children) to test the effects per child. The parous women were further divided into four groups according to the total duration of breastfeeding. Of these four groups, one group consisted of women who did not breast-feed after birth, whereas the breastfeeding women were divided into pre-specified tertiles. These four groups had breastfed for 0, 1 to 9, 10 to 19, 20 or more months, respectively. We used analysis of variance (ANOVA) and Scheffé post hoc tests to compare the groups and to adjust for multiple comparisons of means, and linear regression analyses for p for trend. Cox's proportional hazards model were used to determine whether parity and total duration of breastfeeding predicted fractures. We used the same models to determine the effect of parous versus nulliparous, breastfeeding versus not breastfeeding, and the effect of duration of breastfeeding per child on the risk for fracture. The proportionality assumptions of the models were verified. We controlled for age, BMI, height, current smoking, alcohol use, HRT use, physical activity, a history of diabetes, previous hip or wrist fracture, and length of education, known to be associated with fracture risk.31, 32 We could not control the risk for hip fracture for total hip BMD because of too few fractures in the subgroup with hip BMD measurements. When we controlled the risk for wrist fracture for distal forearm BMD, the results did not change. We reran all the analyses after exclusion of women with high-energy trauma involved because of, for example, traffic accidents (n = 72).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Participants mean age at baseline was 63.9 years (range 50 to 94). The parous women (90.4%), had given birth to a median of three children (range 1 to 12), and the breastfeeding women (95.1%) had breastfed for a median of 13 months (range 1 to 213). Higher-parity women had higher weight and BMI, and a larger proportion of them were smokers, teetotalers, had diabetes, and fewer years of education (all p for trend < 0.05, Table 1). Women with longer duration of breastfeeding were older, had higher weight and BMI, and had less HRT use and fewer years of education (all p trend < 0.01, Table 1).

Table 1. Age-adjusted Means, 95% CI for the Means, or Percentages for Baseline Characteristics by Parity and Duration of Breastfeeding (Months)
Parity (n = 4681)0 (n = 451)1 (n = 480)2 (n = 1195)3 (n = 1237)≥ 4 (n = 1318)p Trend
Age (years)65.5 (64.6–66.3)64.5 (63.7–65.3)63.0 (62.5–63.5)62.6 (62.1–63.1)65.3 (64.8–65.8)0.80
Height (cm)160.4 (159.8–161.0)161.1 (160.5–161.6)161.1 (160.8–161.4)161.4 (161.0–161.7)160.1 (159.8–160.5)0.09
Weight (kg)66.3 (65.2–67.5)67.8 (66.7–68.9)67.1 (66.4–67.8)67.7 (67.1–68.4)69.2 (68.5–69.8)<0.001
Body mass index (kg/m2)25.7 (25.3–26.2)26.1 (25.7–26.5)25.9 (25.6–26.1)26.0 (25.8–26.3)27.0 (26.7–27.2)<0.001
Physical activity score2.6 (2.4–2.8)2.6 (2.4–2.8)2.7 (2.6–2.9)2.7 (2.6–2.8)2.6 (2.5–2.7)0.55
Smoking, %22.731.427.228.431.80.04
Teetotaler, %31.830.524.727.839.1<0.001
History of diabetes, %2.93.03.53.44.90.02
Previous fracture,* %15.117.312.913.615.40.96
Using HRT, %8.98.812.010.87.10.10
Education > 7–10 yrs, %40.433.241.037.125.9<0.001
Feeder, % 87.794.397.196.5<0.001
Breastfeeding (months) 5.6 (4.4–6.9)10.9 (10.2–16.6)16.4 (15.7–17.1)27.0 (26.3–27.7)<0.001
Breastfeeding duration (n = 3748)0 (n = 184)1–9 (n = 1282)10–19 (n = 1111)≥ 20 (n = 1171)p Trend
  • Physical activity score, hours of moderate activity + 2 × hours of hard activity; HRT, hormone replacement therapy.

  • *

    Fracture of hip or wrist.

Age (years) 62.9 (61.6–64.2)61.4 (60.9–61.9)63.0 (62.5–63.5)65.9 (65.4–66.4)<0.001
Height (cm) 160.4 (159.6–161.3)161.2 (160.9–161.5)161.4 (161.0–161.7)161.0 (160.7–161.3)0.94
Weight (kg) 68.4 (66.7–70.2)67.4 (66.7–68.1)67.5 (66.8–68.2)69.5 (68.8–70.2)<0.001
Body mass index (kg/m2) 26.6 (25.9–27.2)25.9 (25.7–26.2)25.9 (25.7–26.2)26.8 (26.6–27.1)<0.001
Physical activity score 2.6 (2.3–2.9)2.8 (2.6–2.9)2.7 (2.6–2.8)2.8 (2.7–2.9)0.51
Smoking, % 27.230.031.126.90.26
Teetotaler, % 29.025.825.736.4<0.001
History of diabetes, % 2.93.63.04.60.16
Previous fracture,* % 12.013.513.616.00.28
Using HRT, % 13.014.29.67.20.005
Education > 7–10 yrs, % 36.038.538.932.00.005
Breastfeeding (months)  5.9 (5.5–6.4)14.0 (13.5–14.5)32.0 (31.5–32.5)<0.001

BMD at the distal forearm and total hip did not differ significantly between groups when nulliparous women were compared with groups of parity, or when women who did not breast-feed after birth were compared with groups of breastfeeding women (Table 2). BMD at the distal forearm and total hip showed no significant trend by parity or by duration of breastfeeding.

Table 2. Bone Mineral Density (BMD) at the Distal Forearm and Total Hip by Parity and by Total Duration of Breastfeeding (Months)
ParityDistal Forearm BMD (mg/cm2)Total Hip BMD (mg/cm2)
Mean (95% CI)aMean (95% CI)bMean (95% CI)cMean (95% CI)aMean (95% CI)bMean (95% CI)c
n = 3556n = 3549n = 3500n = 2358n = 2354n = 2320
0398 (391–404)399 (393–405)399 (393–406)888 (870–906)895 (879–910)893 (878–909)
1398 (392–404)399 (393–405)399 (393–405)877 (860–895)877 (862–892)880 (865–896)
2399 (395–402)399 (396–403)399 (395–403)885 (875–895)889 (880–898)888 (879–897)
3396 (392–399)396 (393–400)396 (393–400)878 (868–888)882 (873–891)879 (871–888)
≥ 4402 (398–406)400 (396–403)401 (397–405)893 (883–903)883 (874–892)888 (879–897)
p for trend0.290.980.810.410.370.65
Breastfeeding durationn = 2933n = 2928n = 2887n = 1965n = 1961n = 1931
  • a

    Adjusted for age.

  • b

    Adjusted for age and body mass index.

  • c

    Adjusted for age, body mass index, height, smoking, alcohol, physical activity, length of education, history of diabetes, and use of hormone replacement therapy in analyses of variance, using Scheffé post hoc tests to adjust for multiple comparisons of means as a weighted test to determine which pair of means differed significantly from each other.

  • *

    p < 0.05 for differences in BMD between women who had breastfed ≥ 20 months and those who had breastfed 10 to 19 months after birth.

0394 (385–403)394 (385–403)396 (387–405)862 (836–888)865 (842–889)872 (849–895)
1–9402 (398–405)402 (399–406)402 (399–406)887 (877–897)887 (879–896)886 (878–895)
10–19396 (392–400)397 (393–400)398 (394–401)884 (874–894)888 (878–897)890 (881–899)
≥ 20405 (401–409)*403 (400–407)404 (400–407)894 (883–905)889 (879–899)888 (878–898)
p for trend0.130.410.580.100.290.48

During 51 906 person-years and a median of 14.5 years follow-up, 442, 621, and 1105 of 4681 women suffered incidents of hip, wrist, and fragility fractures (hip, wrist, or proximal humerus), and the fracture rates were 7.8, 11.4, and 21.3 per 1000 person-years (Table 3). There were no differences in risk for hip, wrist, and fragility fracture between parous and nulliparous women (n = 451, 9.6%, Table 3). Parity was not associated with risk for fracture at any site, after adjustment for age, BMI, and other covariates. Additional adjustment for breastfeeding made no change in the results.

Table 3. Risk for Hip, Wrist, or Fragility Fracture for Parous Versus Nulliparous Women, and by Parity
Parous vs. nulliparousnFractureRateHR (95% CI)aHR (95% CI)bHR (95% CI)c
n = 4681n = 4663n = 3844
  Hip 0.95 (0.71–1.28)1.01 (0.75–1.36)1.04 (0.75–1.46)
  Wrist 1.03 (0.79–1.36)1.07 (0.81–1.40)1.20 (0.88–1.63)
  Fragility 1.08 (0.88–1.32)1.11 (0.91–1.36)1.20 (0.96–1.51)
Parity Hip    
0451509.7111
1480519.01.07 (0.72–1.58)1.10 (0.74–1.63)0.96 (0.61–1.50)
211951188.01.08 (0.77–1.50)1.09 (0.78–1.52)1.17 (0.81–1.70)
31237966.20.88 (0.62–1.24)0.91 (0.65–1.29)1.00 (0.68–1.47)
≥ 413181278.00.88 (0.63–1.22)0.98 (0.71–1.37)1.00 (0.69–1.46)
All/p trend46814427.80.140.480.89
  Wrist    
04515711.4111
14806812.61.13 (0.79–1.60)1.14 (0.80–1.62)1.28 (0.87–1.90)
2119515310.90.99 (0.73–1.34)0.99 (0.73–1.35)1.16 (0.83–1.63)
3123717511.91.10 (0.81–1.48)1.12 (0.83–1.51)1.28 (0.91–1.79)
≥ 4131816811.10.98 (0.73–1.33)1.06 (0.78–1.43)1.11 (0.78–1.57)
All/p trend468162111.40.780.720.80
  Fragilityd    
  • a

    Hazard ratio (HR) adjusted for age.

  • b

    HR adjusted for age and body mass index (BMI).

  • c

    HR adjusted for age, height, BMI, smoking, alcohol use, physical activity, history of diabetes and previous wrist or hip fracture, use of hormone replacement therapy, and length of education.

  • d

    Fragility fractures; hip, wrist, or proximal humerus fracture.

045110421.5111
148011421.91.08 (0.83–1.40)1.09 (0.84–1.42)1.12 (0.83–1.51)
2119528121.01.10 (0.88–1.38)1.10 (0.88–1.38)1.21 (0.94–1.56)
3123729521.01.13 (0.91–1.42)1.16 (0.92–1.45)1.28 (0.99–1.64)
≥ 4131831121.61.02 (0.82–1.28)1.10 (0.88–1.37)1.15 (0.89–1.48)
All/p trend4681110521.30.970.450.28

Compared with women who did not breast-feed after birth (n = 184, 4.9%), women who breastfed their offspring had a 50% lower risk for hip fracture (HR 0.50; 95% CI 0.32 to 0.78), and a 27% lower risk for fragility fracture (HR 0.73; 95% CI 0.54 to 0.99), after multivariable adjustment (Table 4). Each 10 months longer duration of breastfeeding reduced the age-adjusted risk for hip fracture by 12% (HR 0.88; 95% CI 0.78 to 0.99, p for trend = 0.03). This trend remained significant after adjustment for parity (p < 0.05). However, the trend was no longer statistically significant after adjustment for BMI and other covariates (HR 0.91; 95% CI 0.80 to 1.04, Table 4). Longer duration of breastfeeding per child reduced the risk for hip fracture similarly to longer total duration of breastfeeding. Breastfeeding was not associated with risk for wrist fracture. Exclusion of high-energy fracture did not change the results substantially.

Table 4. Risk for Hip, Wrist, and Any Non-vertebral Fragility Fracture for Breastfeeding Versus Non-breastfeeding Women, and Risk for Fracture by Total Duration of Breastfeeding (Months)
Breastfeeding versus non-breastfeeding womennFractureRateHR (95% CI)aHR (95% CI)bHR (95% CI)c
n = 3748n = 3734n = 3216
  Hip, 0.62 (0.40–0.96)0.63 (0.41–0.97)0.50 (0.32–0.78)
  Wrist 1.02 (0.69–1.53)1.01 (0.68–1.52)1.06 (0.68–1.66)
  Fragilityd 0.80 (0.60–1.06)0.80 (0.60–1.05)0.73 (0.54–0.99)
Breastfeeding duration Hip    
0184229.7111
1–91282996.10.69 (0.43–1.09)0.67 (0.42–1.06)0.51 (0.31–0.83)
10–191111936.70.62 (0.39–0.99)0.61 (0.38–0.97)0.49 (0.30–0.80)
≥ 2011711218.70.57 (0.36–0.90)0.61 (0.39–0.97)0.50 (0.31–0.81)
All/p trend37483357.30.030.130.15
  Wrist    
01842511.3111
1–9128217411.41.04 (0.68–1.58)1.00 (0.66–1.53)1.08 (0.68–1.73)
10–19111116012.11.08 (0.71–1.64)1.05 (0.69–1.60)1.05 (0.66–1.69)
≥ 20117115011.30.96 (0.63–1.47)0.99 (0.65–1.51)1.04 (0.65–1.66)
All/p trend374850911.60.620.950.83
  Fragilityc    
  • Statistical significant models are bold.

  • a

    Hazard ratio (HR) adjusted for age.

  • b

    HR adjusted for age and body mass index (BMI).

  • c

    HR adjusted for age, height, BMI, smoking, alcohol use, physical activity, history of diabetes and previous wrist or hip fracture, use of hormone replacement therapy, and length of education.

  • d

    Fragility fractures; hip, wrist or proximal humerus.

01845225.3111
1–9128227518.70.80 (0.59–1.07)0.78 (0.58–1.05)0.71 (0.52–0.98)
10–19111126320.70.82 (0.61–1.10)0.80 (0.59–1.07)0.72 (0.53–0.99)
≥ 20117129423.40.79 (0.58–1.06)0.81 (0.60–1.09)0.76 (0.56–1.05)
All/p trend374888421.10.380.760.75

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

We report that breastfeeding reduced the risk for hip fracture, but was not associated with BMD at the distal forearm and total hip. By parity, there were no differences in distal forearm or hip BMD or risk for hip, wrist, or fragility fractures in this study.

To the best of our knowledge, this is the first report of breastfeeding associated with lower risk for hip fracture in a large prospective population-based study. The modest protective effect of breastfeeding is in agreement with the Melton et al. report of breastfeeding for more than eight months associated with greater BMD at the femur shaft and lumbar spine with no effect of parity,33 and the Cumming et al. report of lower risk for hip fracture in women who breastfed their offspring than women who did not.23

Fox et al. reported parity associated with higher radius BMD; with each additional birth, the radius BMD was 1.4% higher in 2330 women ≥ 65 years of age, whereas there were no differences in BMD between women who were breastfeeding and those who were not.34 Kritz-Silverstein et al. reported hip and wrist BMD associated with parity and lactation before, not after, adjustment for age and BMI,35 whereas Murphy et al. reported higher hip and spine BMD associated with higher parity before and after adjustment for age and BMI, but no association between BMD and breastfeeding.36 In brief, these cross-sectional studies showed either no relationship between BMD and breastfeeding,34, 36 or the relationship disappeared after controlling for BMI,33, 35 but only 33% to 55% of the parous women breastfed their child.33, 34, 36 In our study, 95% of parous women breastfed their child, and total breastfeeding for 20 months or more was not associated with any significant difference in BMD at the distal forearm and total hip compared with those who did not breast-feed. Even in adolescents there appear to be no long-term adverse effect of parity or lactation on bone mass; as Chantry et al. reported in a prospective study, women aged 20 to 25 years who breastfed during adolescence had higher hip BMD than those who did not breast-feed, and their BMD levels were equivalent to nulliparous women.37

Most,24, 25 but not all,20 of the studies of the relationship of breastfeeding and fracture risk are limited to small case-control studies without any significant relationship. However, even a large case-control study of 1328 incident cases of hip fractures and 3312 controls, showed no relationship of breastfeeding with fracture risk.20 In contrast, we report a protective threshold effect of breastfeeding on hip fracture risk, which persisted after adjustment for BMI and other covariates. Our findings of lower risk for hip fracture in breastfeeding women than in women who did not breast-feed, is similar to the report from a small case-control study by Cumming et al. suggesting a lower risk for hip fracture by longer duration of breastfeeding per child after multivariable adjustment.23 Because our data showed no effect of breastfeeding on risk for wrist fractures, and the group of humerus fractures was too small for a site-specific risk estimate, the reduced risk for fragility fracture of 27% is accounted for by the 50% reduction in hip fractures.

In the Study of Osteoporotic Fracture, during a 10-year follow-up of 9704 women aged ≥ 65 years, nulliparous women had 44% increased risk of hip fracture independent of weight, height, and BMD, and in parous women, each additional birth reduced hip fracture risk by 9%.18 Similarly, the Dubbo Osteoporosis Study of 1091 women over 60 years of age reported higher risk of all atraumatic fractures in nulliparous women independent of BMD.19 In a population-based study by Michaelsson et al. of 1328 cases with hip fractures and 3321 controls, the age-adjusted risk for hip fracture was reduced by 10% per child, similar to the 7% reduction observed by Paganini-Hill et al. during a 21-year follow-up of 8877 women with a mean age of 73 years.20, 21 However, BMI explained about half of the risk reduction.

In our study, including 10% nulliparous women, we could not confirm any protective effect of parity on hip fractures risk as reported in large prospective studies with 19% to 26% nulliparous women.18, 21 Some nulliparous women could be subfertile with reduced estrogen levels during their menstrual cycle that may contribute to greater risk for fracture.18 In addition, the effect of parity in those studies was not controlled for breastfeeding, which could be one reason why parous women are protected from fragility fracture. Parity, breastfeeding, or both may be causally related to increased weight; if this is true, this may protect against fracture at least in part because of the tendency for women to become heavier with increasing parity, breastfeeding, or both. Nevertheless, Paganini-Hill et al. reported parity to be protective even after controlling for BMI, suggesting that increased BMI may not be the only way pregnancy protects against fracture.21 The most likely mechanism by which childbearing protects is through increases in bone formation rates leading to increases in bone mass. However, other mechanisms, may lead to a favorable bone structure and strength with or without increased bone mass. Specker et al. suggested that greater femoral neck size, without higher BMD may explain the reduced hip fracture risk in women with higher parity.38 They reported no protective effect of breastfeeding, but that study may be limited by a small sample size.

The study has some limitation because most parous women breastfed after birth; the group of parous women who did not breast-feed was small. In addition, the effect of parity and breastfeeding on fracture risk independent of BMD could not be assessed because BMD only was measured in subgroups of participants with low numbers of fractures.

In summary, this study adds to the growing evidence that pregnancy and breastfeeding has no long-term deleterious effect on bone fragility and fractures, and breastfeeding probably confers some additional advantage, and may provide modest protection against hip fracture.

Disclosures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

All the authors state that they have no conflicts of interest.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

We thank Dr GKR Berntsen and Nina Emaus for collecting and quality assessment of the BMD data and the Tromsø Study group headed by Professor Inger Njølstad for providing the data. The study was funded, in part, by the Research Council of Norway (ÅB).

Authors' roles: Study design and conduct: ÅB and LAA. Data collection: ÅB, LAA, LJ, JS, and RMJ. Data analysis: ÅB and LAA. Data interpretation: ÅB, LAA, LJ, JS, and RMJ. Drafting manuscript: ÅB. Revising manuscript: ÅB, LAA, LJ, JS, and RMJ. Approving final version of manuscript: ÅB, LAA, LJ, JS, and RMJ. ÅB takes responsibility for the integrity of the data analysis.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References