Changes in cortical volumetric bone mineral density and thickness, and trabecular thickness in lactating women postpartum.

ORIGINAL

ARTICLE

Changes in Cortical Volumetric Bone Mineral Density and Thickness, and Trabecular Thickness in Lactating Women Postpartum P. Brembeck, M. Lorentzon, C. Ohlsson, A. Winkvist, and H. Augustin Department of Internal Medicine and Clinical Nutrition, Sahlgrenska Academy, University of Gothenburg, 405 30 Gothenburg, Sweden

Context: Lactation is associated with decreased areal bone mineral density (aBMD). Replenishment occurs especially after ceased lactation. Changes in volumetric bone mineral density (vBMD), microstructure, and dimensional parameters are unknown and may clarify the role of lactation for skeletal health. Objective and Main Outcomes: The objective of the study was to test the hypothesis that lactation is associated with changes in aBMD, vBMD, microstructure, and dimensional parameters. Design: At baseline (0.5 mo after delivery) and 4, 12, and 18 months thereafter, bone was assessed using dual-energy x-ray absorptiometry and high-resolution peripheral quantitative computed tomography. Participants and Setting: Eighty-one fair-skinned postpartum women and 21 controls aged 25– 40 years were recruited. The completion ratio was 73%. Postpartum women were categorized depending on duration of lactation: 0 –3.9, 4 – 8.9, and 9 months or longer. Results: During the first 4 months, aBMD decreased at several sites (geometric mean ⫾ SE; ⫺0.73% ⫾ 0.21% to ⫺3.98% ⫾ 0.76%) in women lactating at least 4 months. During the same time, cortical vBMD at the ultradistal tibia decreased in women lactating 4 – 8.9 months (⫺0.26% ⫾ 0.08%) and 9 months or longer (⫺0.49% ⫾ 0.10%). At 12 months postpartum, cortical thickness (ⱖ9 mo, ⫺2.48% ⫾ 0.41%) and trabecular thickness (4 – 8.9 mo, ⫺2.14% ⫾ 0.92%; ⱖ 9 mo, ⫺2.56% ⫾ 1.21%) also were lower than baseline. No decreases were found in women lactating less than 4 months or in controls in these parameters. At 18 months postpartum, both cortical vBMD (ⱖ9 mo, ⫺0.77% ⫾ 0.17%) and trabecular thickness (4 – 8.9 mo, ⫺2.25% ⫾ 1.25%; ⱖ 9 mo, ⫺3.21% ⫾ 1.41%) were lower in women with long lactation. Conclusions: Decreases in cortical vBMD, thickness, and trabecular thickness at the ultradistal tibia were found in women lactating 4 months or longer. Longer follow-up is needed to confirm whether women with extended lactation recover fully or whether the changes could potentially lead to an increased risk of fracture in later life. (J Clin Endocrinol Metab 100: 535–543, 2015)

L

actation requires substantial amounts of calcium every day. It is estimated that mothers who are fully lactating secrete 200 –300 mg/d calcium into their breast milk and some even more (1–3). Several longitudinal studies have shown a decrease in maternal femoral neck and lum-

bar spine areal bone mineral density (aBMD) of 2%– 6% during the first months of lactation (1, 3–7), as assessed by dual-energy x-ray absorptiometry (DXA). Furthermore, decreases in aBMD postpartum have been shown to be greater in women with longer duration of lactation, com-

ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2015 by the Endocrine Society Received July 3, 2014. Accepted November 7, 2014. First Published Online November 11, 2014

Abbreviations: aBMD, areal bone mineral density; DXA, dual-energy x-ray absorptiometry; HR-pQCT, high-resolution peripheral quantitative computed tomography; PAL, physical activity level; vBMD,volumetric bone mineral density.

doi: 10.1210/jc.2014-2825

J Clin Endocrinol Metab, February 2015, 100(2):535–543

jcem.endojournals.org

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 February 2015. at 05:05 For personal use only. No other uses without permission. . All rights reserved.

535

536

Brembeck et al

Lactation and Volumetric Bone Mineral Density

pared with women with a shorter duration of lactation or formula-feeding mothers (1, 3–5, 8 –10). Several studies indicate that the influence of lactation varies between different skeletal sites. During the first months of lactation, aBMD is suggested to initially decrease at sites consisting mainly of trabecular bone (3–5, 11). During extended lactation, a few studies indicate that bone loss is mostly located to the cortical-rich bones (4, 5, 11). It is speculated that a recovery of aBMD at the end of extended lactation or after weaning, therefore, is detected earlier in the trabecular-rich lumbar spine than in the cortical-rich femoral neck (4, 5, 11). Whether aBMD fully recovers after weaning is still not clear. Some studies observed a recovery to baseline (after delivery) 6 months after weaning (5, 12–14) and also in women with extended lactation (5, 13, 14). These studies show a decrease in lumbar spine aBMD during the first 9 (5) to 12 (13, 14) months postpartum in women with the longest duration of lactation, which 18 months postpartum was no longer significantly different compared with baseline (5, 13, 14). Furthermore, in Gambian women, periods of long lactation are not associated with consistent decreases in bone minerals (15). Another study, however, does not show a complete recovery at all skeletal sites at 6 months after weaning in women with long duration of lactation (16). Hypothetically, repeated pregnancies and/or long lactations could lead to an increased fracture risk later in life (1, 17). However, several studies show no or even an inverse relation between parity, lactation, and fractures (8, 18 –20). In Sweden, a long duration of lactation is common, and at 6 months postpartum, 63% of all infants are breast-fed (21). Therefore, studies of the skeletal recovery after weaning in women with extended lactation require a long follow-up period. Only a few earlier studies have investigated changes in aBMD postpartum in women with an extended lactation and a follow-up period longer than 12 months (5, 13, 14). Of these studies, only Moller et al (5) also included a control group. Inclusion of a comparable group of women who are neither pregnant nor lactating is necessary to handle the variations in increasing age, changes in weight, and possible technical limitations (22). Most previous studies have used DXA to report changes in aBMD postpartum (1, 3–5, 14, 18 –20, 23). Only a two-dimensional areal measurement of the bone is reported by DXA. In addition, DXA does not separate between trabecular and cortical bone. Thus, currently it is not possible to draw conclusions regarding the relation between duration of lactation and changes in trabecular and/or cortical bone. However, the new technique highresolution peripheral quantitative computed tomography (HR-pQCT) measures volumetric bone mineral density


(vBMD) and separates cortical and trabecular bone. Furthermore, HR-pQCT provides information about microstructural changes, such as trabecular thickness and number and trabecular bone volume fraction, and dimensional changes, such as cortical thickness and area. No such data from the postpartum period has previously been published, and thus, changes in vBMD, microstructural and dimensional parameters postpartum, and its clinical implications are unknown. Such information would increase the understanding of skeletal changes during and after lactation. So far, a previous study among postmenopausal women has shown that cortical and trabecular vBMD and cortical thickness are significant determinants of fracture risk (24). The same study showed that HR-pQCT may detect small changes in bone mineral density that are not detectable by DXA (24). This is another advantage of HR-pQCT. Thus, our aim is to test the hypothesis that lactation is associated with changes not only in aBMD, but also in vBMD, microstructural and dimensional parameters from 2 weeks to 18 months postpartum in fair-skinned women living in Sweden. A control group consisting of nonpregnant and nonlactating fair-skinned women was included in this study.

Subjects and Methods Subjects Pregnant and nonpregnant and nonlactating women (controls) were recruited from July 2008 to July 2011 using posters in maternity health care clinics and in public places near Gothenburg, Sweden, and through advertisements on a Swedish web page addressing pregnant women in western Sweden. In total, 81 pregnant women and 21 nonpregnant and nonlactating controls agreed to participate. Inclusion criteria were age 25– 40 years and declaring oneself as healthy. Exclusion criteria were intake of prescribed medicine known to effect calcium and bone metabolism, recent bone fractures, pregnancy during the last 1.5 years before current pregnancy/before entering the study as control, miscarriage after week 12 of pregnancy during the last 1.5 years, breast-feeding during the last year before current pregnancy/before entering the study as control, current twin pregnancy, and current development of gestational diabetes or preeclampsia. This study was conducted according to the guidelines laid down in the Declaration of Helsinki (25), and all procedures involving the subjects were approved by the Regional Ethics Committee in Gothenburg and the Swedish Radiation Safety Authority. Written informed consent was obtained from all women.

Study design All women visited the Department of Internal Medicine and Clinical Nutrition, University of Gothenburg, Sweden, at baseline (2 wk after delivery for postpartum women) and 4, 12, and 18 months thereafter. Body weight in underwear was measured


doi: 10.1210/jc.2014-2825

(Tanita; BWB-800MA, Rex Frederiksbergs Vaegtfabrik), and height was assessed using a standardized wall stadiometer. Selfreported prepregnancy body weight was recorded. At all visits, women were asked about their lactation status, physical activity level (PAL), and use of hormonal contraceptives. If lactating, detailed information about lactation habits was collected, including the number of lactation sessions per day, number and amount of formula feedings per day, date of introduction of solid foods, and daily amount of solid foods given. Women were asked to record the date of their last lactation. Women were classified as postpartum women or controls. Postpartum women were categorized prior to statistical analyses according to their duration of total lactation; 0 –3.9 months, 4 – 8.9 months, and 9 months or longer. Total lactation refers to any extent of lactation. Full lactation was defined as when 90% or more of the infants’ daily energy intake came from breast milk.

Measurements At all occasions, bone status was measured at the Osteoporotic Laboratory, Sahlgrenska University Hospital, Gothenburg, Sweden. The DXA (Lunar Prodigy, GE Lunar Corp; software version 11.400.004) was used to measure aBMD at the ultradistal radius, lumbar spine (L1-L4), femoral neck, femoral shaft, femoral trochanter, total femur, and whole-body. The radius was measured on the nondominant upper limb. The coefficients of variation for the DXA measurements were 0.5%–3%, depending on the measuring site. HR-pQCT (XtremeCT, software version 5.3; Scanco Medical AG) was used to measure cortical vBMD, cortical thickness, cortical area, trabecular vBMD, trabecular thickness, trabecular bone volume fraction, and trabecular number at the ultradistal tibia on the lower limb at the same side as the nondominant radius. Briefly, a reference line was manually placed at the center of the scan of the end plate of the distal tibia. The first computed tomography slice started 22.5 mm proximal to the reference line for the tibia, as previously described (26). Standard evaluations were used. The coefficients of variation were determined from three repeated measurements according to a standardized protocol on two subjects, as previously reported (27). The coefficient of variation ranged from 0.1% to 1.6%: cortical vBMD (0.1%), cortical thickness (0.3%), cortical area (0.4%), trabecular vBMD (0.2%), trabecular thickness (0.7%), trabecular number (1.6%), and trabecular bone volume fraction (0.3%). For long-term precision, we refer to the study by Burghardt et al (28).

Statistical analyses Comparisons between postpartum women and controls at baseline were made using a Student’s t test. Data are presented as mean ⫾ SE, if not otherwise described. Because the data from the DXA and HR-pQCT were not normally distributed, they were log transformed when analyzed. The geometric mean (quartile 1-quartile 3) was used when presenting the means for postpartum women and controls at each time point. Repeated measures were analyzed by linear longitudinal models to study both between- and within-group differences over time. We assumed fixed effects, ie, that the model holds true across the sample, and with the same intercept and slope, we could estimate the outcome. Unstructured covariance structure of our longitudinal data was used. The longitudinal analysis yielded estimates of least square means with SE and P values. The repeated-measures ANOVA from mixed procedure yielded the differences of these


537

least square means and whether these differences were statistically significant. Longitudinal analyses were adjusted for body weight, PAL, age, and use of hormonal contraceptives when significant in the models. Percentage changes in aBMD, vBMD, and microstructural and dimensional parameters were calculated by the difference in log-transformed data between two time points and multiplied by 100. This approximates the percentage change as shown by Cole (29). The significance level was set to P ⬍ .05 (two sided). The mixed-procedure, repeated-measures ANOVAs were conducted using SAS for Windows (version 9.2; SAS Institute Inc). All other analyses were conducted using SPSS Statistics Software (version 21.0; IBM).

Results Descriptive characteristics of postpartum women and controls at baseline are shown in Table 1. Mean self-reported prepregnancy body weight of postpartum women was 64 ⫾ 1 kg. Eighty percent of the postpartum women and 71% of the controls had studied for 3 years or more at the university level. The only descriptive characteristic that differed significantly was body weight at baseline, which was higher for postpartum women compared with controls. The median duration of total lactation was 8.1 months (q1– q3 6.8 –10.4 mo, range 0 –20.2 mo). The median duration of total lactation within each lactation group was 2.7 months (q1– q3 1.0 –3.7 mo) for women lactating 0 –3.9 months, 7.6 months (q1– q3 6.8 – 8.2 mo) for women lactating 4 – 8.9 months, and 11.8 months (q1– q3 10.1–12.6 mo) for women lactating 9 months or longer. The percentage of postpartum women using hormonal contraceptives 0.5, 4, 12, and 18 months postpartum were 0%, 27%, 30%, and 31%, respectively. Corresponding percentage of controls were 52%, 48%, 58%, and 41%, respectively. Numbers of postpartum women and controls with a full set of data were, respectively, at 4 months postpartum 81 and 21 women, at 12 months postpartum 79 and 19 women, and at 18 months postpartum 58 and 18 women (see Table 2). Reasons for leaving the study were a new pregnancy (n ⫽ 17), time constraint (n ⫽ 8), and the development of disease (n ⫽ 3). Two of these three women were excluded from the analyses from the time point when they started their medication. The third women was excluded due to the potential development of osteoporosis. Because of a high dropout rate between 12 and 18 months postpartum in the group lactating 0 –3.9 months, the measurement at 18 months postpartum was excluded from the analyses for this lactation group. Mixed-procedure, repeated-measures ANOVAs showed significant differences in change over time in aBMD between postpartum women and controls at the lumbar spine, femoral neck, shaft, and trochanter and total femur


538

Brembeck et al

Table 1.



Descriptive Characteristics of the Participating Women at Baselinea

Age, y Height, cm Weight, kg Parity aBMD, g/m2 Radius ultradistal Lumbar spine Femoral neck Femoral shaft Femoral trochanter Total femur Whole body Cortical vBMD, mg/cm3 Cortical thickness, mm Cortical area, mm2 Trabecular vBMD, mg/cm3 Trabecular thickness, mm Trabecular number, mm⫺1 Trabecular bone volume fraction

Postpartum Women (n ⴝ 81) Mean (SE)

Controls (n ⴝ 21) Mean (SE)

P Value

32.9 (0.4) 168.4 (0.7) 70.2 (1.0) 0.6 (0.1)

32.3 (1.0) 168.9 (1.0) 64.2 (1.8) 0.7 (0.4)

.504 .754 .009b .867

0.334 (0.306 – 0.372) 1.167 (1.080 –1.252) 0.993 (0.920 –1.072) 1.204 (1.091–1.309) 0.783 (0.710 – 0.866) 1.012 (0.913–1.090) 1.180 (1.131–1.232) 892.5 (867.8 –918.7) 1.134 (0.977–1.300) 117.6 (106.6 –132.0) 156.9 (133.0 –177.5) 0.070 (0.060 – 0.079) 1.860 (1.696 –2.016) 0.131 (0.111– 0.148)

0.348 (0.317– 0.368) 1.209 (1.105–1.300) 1.009 (0.928 –1.101) 1.202 (1.149 –1.260) 0.775 (0.696 – 0.863) 1.011 (0.935–1.039) 1.195 (1.151–1.246) 906.3 (893.4 –942.0) 1.158 (1.000 –1.489) 118.2 (106.1–133.4) 164.4 (151.3–172.3) 0.074 (0.066 – 0.084) 1.849 (1.680 –2.020) 0.137 (0.126 – 0.144)

.230 .210 .647 .828 .736 .884 .472 .135 .751 .993 .445 .249 .896 .450

Statistical analyses for comparisons between postpartum women and controls were performed with a Student’s t test. For all bone parameters, geometric means (quartiles 1–3) are shown. a

Two weeks after delivery.

b

Significant difference between postpartum women and controls.

(P ⬍ .001 for all sites) (Supplemental Table 1). Body weight, age, and the use of hormonal contraceptives remained as significant confounders in the longitudinal analyses of changes in aBMD at some sites (Supplemental Tables 1 and 2), whereas PAL did not. When grouping the women according to the duration of total lactation (0 –3.9 mo, 4 – 8.9 mo, ⱖ 9 mo, and controls), repeated-measures ANOVAs showed that decreases in aBMD were significant only in the groups lactating 4 – 8.9 months and 9 months or longer. During the first 4 months postpartum, these two lactation groups had significant decreases in aBMD in the range ⫺0.73% ⫾ 0.21% to ⫺3.98% ⫾ 0.76% at the lumbar spine, femoral neck, shaft and trochanter, total femur, and whole body as compared with baseline (Table 2). At 12 months postpartum, aBMD at femoral neck, shaft, and total femur were still significantly lower in women lactating 4 – 8.9 months and 9 months or longer compared with baseline, as well as femoral trochanter and whole body for women lactating 9 months or longer. For women lactating 4 – 8.9 months, aBMD at the lumbar spine was instead significantly higher than baseline at 12 months postpartum. For women with a short duration of lactation (0 –3.9 mo), aBMD at the lumbar spine and radius ultradistal was also significantly higher compared with baseline at 12 months postpartum (Table 2). At 18 months postpartum, aBMD at the lumbar spine was significantly higher than baseline for women lactating both 4 – 8.9 months and 9 months or longer, as well as the

femoral trochanter and radius ultradistal for women lactating 4 – 8.9 months (Table 2). No significant decreases compared with baseline were found at 18 months postpartum in any lactation group. No significant changes were found in controls at any time point. Mixed-procedure, repeated-measures ANOVAs revealed significant differences in change over time at the ultradistal tibia bone parameters between postpartum women and controls in cortical thickness (P ⫽ .011), cortical area (P ⫽ .005), trabecular bone volume fraction (P ⫽ .044), and trabecular vBMD (P ⫽ .049) (Supplemental Table 2). No confounders were found. Repeated-measures ANOVAs showed significant decreases only in women with a long duration of lactation (4 – 8.9 and ⱖ 9 mo) for cortical vBMD (⫺0.26% ⫾ 0.08% and ⫺0.49% ⫾ 0.10%) and cortical area (⫺0.71% ⫾ 0.24% and ⫺1.54% ⫾ 0.33%) during the first 4 months postpartum at the tibia (Table 3). In women lactating 4 – 8.9 months, significant decreases compared with baseline were also evident for cortical thickness (⫺2.46% ⫾ 1.63%) and trabecular bone volume fraction (⫺0.66% ⫾ 0.38%). No changes were found in women with a short duration of lactation (0 –3.9 mo). In controls, significant changes were found only in the trabecular vBMD (⫺1.03% ⫾ 0.35%) and trabecular bone volume fraction (⫺1.12% ⫾ 0.37%) during the first 4 months, as compared with baseline. At 12 months postpartum, significant decreases compared with baseline were still evident for women lactating


doi: 10.1210/jc.2014-2825

Table 2.


539

Mean Percentage Change ⫾ SE in aBMD Compared With Baseline,a as Assessed With DXA Duration of Lactation

Percentage Change in aBMD (g/cm2) Compared With Baseline 4 mo postpartum, n 12 mo postpartum, n 18 mo postpartum, n Radius ultradistalb 4 mo postpartum 12 mo postpartum 18 mo postpartumd Lumbar spineb 4 mo postpartum 12 mo postpartum 18 mo postpartum Femoral neckb,f 4 mo postpartum 12 mo postpartum 18 mo postpartum Femoral shaftf,g 4 mo postpartum 12 mo postpartum 18 mo postpartum Femoral trochanter 4 mo postpartum 12 mo postpartum 18 mo postpartum Total femurf 4 mo postpartum 12 mo postpartum 18 mo postpartum Whole body 4 mo postpartum 12 mo postpartum 18 mo postpartum

0 –3.9 Months

4 – 8.9 Months

> 9 Months

Controls

10 10 2

42 41 34

29 28 22

21 19 18

0.70 ⫾ 1.48 2.10 ⫾ 1.37c

⫺0.35 ⫾ 0.62 ⫺1.82 ⫾ 0.73 0.91 ⫾ 0.73c

⫺1.54 ⫾ 0.73 ⫺2.65 ⫾ 0.88 ⫺1.82 ⫾ 1.36

⫺1.09 ⫾ 0.81 ⫺0.04 ⫾ 1.19 ⫺0.19 ⫾ 1.01

⫺0.16 ⫾ 1.41 3.88 ⫾ 0.94e

⫺3.71 ⫾ 0.53e 1.91 ⫾ 0.61e 4.27 ⫾ 0.65e

⫺3.98 ⫾ 0.76e ⫺0.76 ⫾ 0.71 2.30 ⫾ 0.69c

⫺0.23 ⫾ 0.36 0.31 ⫾ 0.38 0.33 ⫾ 0.48

⫺1.45 ⫾ 0.98 ⫺1.02 ⫾ 1.25

⫺3.80 ⫾ 0.40e ⫺2.92 ⫾ 0.60c ⫺1.60 ⫾ 0.70

⫺3.98 ⫾ 0.66e ⫺4.00 ⫾ 0.69e ⫺2.28 ⫾ 0.89

⫺0.27 ⫾ 0.47 ⫺0.21 ⫾ 0.56 ⫺0.48 ⫾ 0.56

0.05 ⫾ 0.54 1.47 ⫾ 0.81h

⫺3.12 ⫾ 0.34e ⫺1.23 ⫾ 0.40h 0.59 ⫾ 0.51

⫺3.01 ⫾ 0.44e ⫺3.15 ⫾ 0.59e ⫺0.53 ⫾ 0.68

0.48 ⫾ 0.49 0.66 ⫾ 0.63 0.50 ⫾ 0.66

1.65 ⫾ 0.56 2.01 ⫾ 1.04

⫺2.57 ⫾ 0.39e ⫺0.09 ⫾ 0.58 1.92 ⫾ 0.72c

⫺2.85 ⫾ 0.65e ⫺2.19 ⫾ 0.78c 1.35 ⫾ 0.89

⫺0.73 ⫾ 0.57 ⫺0.38 ⫾ 0.75 ⫺0.52 ⫾ 0.61

0.31 ⫾ 0.55 1.42 ⫾ 0.77

⫺2.98 ⫾ 0.31e ⫺1.05 ⫾ 0.44h 0.77 ⫾ 0.52

⫺2.93 ⫾ 0.47e ⫺2.89 ⫾ 0.56e ⫺0.13 ⫾ 0.64

0.09 ⫾ 0.29 0.26 ⫾ 0.37 0.15 ⫾ 0.39

0.63 ⫾ 0.46 0.19 ⫾ 0.41

⫺0.73 ⫾ 0.21c ⫺0.06 ⫾ 0.33 0.54 ⫾ 0.37

⫺0.84 ⫾ 0.30c ⫺1.58 ⫾ 0.38e ⫺0.14 ⫾ 0.40

⫺0.23 ⫾ 0.38 0.36 ⫾ 0.41 0.41 ⫾ 0.55

Statistical analyses were performed with mixed-procedure, repeated-measure ANOVAs with least square means. The confounders body weight, physical activity level, age, and use of hormonal contraceptives were adjusted for if significant in the analyses. a


b

Adjusted for body weight.

c

P ⬍ .01, compared with baseline.

d

Because of the small number of women at 18 months postpartum in the group lactating 0 –3.9 months, this measurement was excluded from all analyses. e


f

Adjusted for age.

g

Adjusted for use of hormonal contraceptives.

h


4 – 8.9 and 9 months or longer in cortical vBMD (⫺0.38% ⫾ 0.15% and ⫺1.24% ⫾ 0.16%) (Figure 1) and also in trabecular thickness (⫺2.14% ⫾ 0.92% and ⫺2.56% ⫾ 1.21%). In women lactating 9 months or longer, significant decreases at 12 months postpartum compared with baseline were further evident for cortical thickness (⫺2.48% ⫾ 0.41%) (Figure 2), cortical area (⫺2.49% ⫾ 0.41%), trabecular vBMD (⫺2.57% ⫾ 0.72%), and trabecular bone volume fraction (⫺2.60% ⫾ 0.70%). No significant changes were found in women with a short duration of lactation (0 –3.9 mo) or in controls at 12 months postpartum, as compared with baseline.

Still at 18 months postpartum, cortical vBMD was significantly lower compared with baseline (⫺0.77% ⫾ 0.17%) for women with the longest duration of lactation (ⱖ9 mo), and trabecular thickness was significantly lower compared with baseline for women lactating both 4 – 8.9 and 9 months or longer (⫺2.25% ⫾ 1.25% and ⫺3.21% ⫾ 1.41%) (Figure 3). In women lactating 4 – 8.9 months, cortical thickness (⫹1.90% ⫾ 0.57%) and area (⫹1.99% ⫾ 0.57%) were instead significantly higher compared with baseline as well as trabecular number (⫹2.31% ⫾ 1.06%). In controls, significant changes were found only in trabecular vBMD (⫺2.38% ⫾ 0.81%) and trabecular


540

Brembeck et al



Table 3. Mean Percentage Change ⫾ SE in Tibia Bone Parameters Compared With Baseline,a as Assessed With HR-pQCT Duration of Lactation Percentage Change Compared With Baseline 4 mo postpartum, n 12 mo postpartum, n 18 mo postpartum, n Cortical vBMD, mg/cm3 4 mo postpartum 12 mo postpartum 18 mo postpartumd Cortical thickness, mm 4 mo postpartum 12 mo postpartum 18 mo postpartum Cortical area, mm2 4 mo postpartum 12 mo postpartum 18 mo postpartum Trabecular vBMD, mg/cm3 4 mo postpartum 12 mo postpartum 18 mo postpartum Trabecular thickness, mm 4 mo postpartum 12 mo postpartum 18 mo postpartum Trabecular number, mm⫺1 4 mo postpartum 12 mo postpartum 18 mo postpartum Trabecular bone volume fraction 4 mo postpartum 12 mo postpartum 18 mo postpartum

0 –3.9 Months

4 – 8.9 Months

> 9 Months

Controls

10 10 2

42 41 34

29 28 22

21 19 18

0.29 ⫾ 0.18 0.24 ⫾ 0.20

⫺0.26 ⫾ 0.08b ⫺0.38 ⫾ 0.15b ⫺0.02 ⫾ 0.18

⫺0.49 ⫾ 0.10c ⫺1.24 ⫾ 0.16c ⫺0.77 ⫾ 0.17c

⫺0.13 ⫾ 0.09 ⫺0.23 ⫾ 0.12 ⫺0.27 ⫾ 0.11

0.71 ⫾ 0.57 1.04 ⫾ 0.74

⫺2.46 ⫾ 1.63e ⫺0.15 ⫾ 0.46 1.90 ⫾ 0.57c

⫺1.52 ⫾ 0.33 ⫺2.48 ⫾ 0.41c ⫺0.70 ⫾ 0.46

⫺0.17 ⫾ 0.18 0.22 ⫾ 0.40 0.38 ⫾ 0.37

0.44 ⫾ 0.48 0.79 ⫾ 0.66

⫺0.71 ⫾ 0.24b ⫺0.04 ⫾ 0.45 1.99 ⫾ 0.57c

⫺1.54 ⫾ 0.33c ⫺2.49 ⫾ 0.41c ⫺0.59 ⫾ 0.46

⫺0.01 ⫾ 0.16 0.27 ⫾ 0.36 0.40 ⫾ 0.32

0.12 ⫾ 0.43 ⫺0.55 ⫾ 0.83

⫺0.55 ⫾ 0.39 ⫺1.14 ⫾ 0.76 ⫺0.03 ⫾ 0.94

⫺0.49 ⫾ 0.32 ⫺2.57 ⫾ 0.72b ⫺1.00 ⫾ 0.88

⫺1.03 ⫾ 0.35d ⫺1.81 ⫾ 0.76 ⫺2.38 ⫾ 0.81e

1.88 ⫾ 2.52 ⫺0.08 ⫾ 1.65

⫺1.35 ⫾ 0.82 ⫺2.14 ⫾ 0.92e ⫺2.25 ⫾ 1.25e

⫺1.89 ⫾ 0.98 ⫺2.56 ⫾ 1.21e ⫺3.21 ⫾ 1.41e

0.09 ⫾ 1.79 ⫺0.36 ⫾ 1.64 ⫺1.60 ⫾ 2.03

⫺1.53 ⫾ 2.57 0.17 ⫾ 1.59

0.92 ⫾ 0.78 0.96 ⫾ 0.70 2.31 ⫾ 1.06e

1.42 ⫾ 1.04 ⫺0.17 ⫾ 1.18 2.21 ⫾ 1.09

⫺1.28 ⫾ 1.77 ⫺1.53 ⫾ 1.56 ⫺0.84 ⫾ 2.02

0.20 ⫾ 0.37 ⫺0.43 ⫾ 0.84

⫺0.66 ⫾ 0.38e ⫺1.15 ⫾ 0.77 0.00 ⫾ 0.93

⫺0.52 ⫾ 0.32 ⫺2.60 ⫾ 0.70b ⫺0.92 ⫾ 0.90

⫺1.12 ⫾ 0.37e ⫺1.85 ⫾ 0.77 ⫺2.45 ⫾ 0.83e

Statistical analyses were performed with mixed-procedure, repeated-measure ANOVAs with least square means. Body weight, physical activity level, age, and use of hormonal contraceptives were not significant as confounders in the analyses. a


b


c


d Because of the small number of women at 18 months postpartum in the group lactating 0 –3.9 months, this measurement was excluded from all analyses. e


bone volume fraction (⫺2.45% ⫾ 0.83%) at 18 months postpartum, as compared with baseline.

Discussion This study describes, using HR-pQCT measurements for the first time, compartmental changes in bone microstructural and dimensional parameters and vBMD during differing durations of lactation. The novelty of this study is that among women lactating 4 months or longer, we show that cortical vBMD, cortical thickness, and trabecular thickness at the ultradistal tibia decreased significantly during the first 12 months postpartum as compared with

baseline. At 18 months postpartum, cortical vBMD as well as trabecular thickness was still significantly lower than at baseline in women lactating 9 months or longer. To our knowledge, no data on changes in vBMD or microstructural and dimensional parameters postpartum have previously been published. The strengths of our study are the inclusion of a control group and longitudinal measurements throughout and after lactation with an 18-month follow-up. In accordance with previous studies (1, 3–5, 8, 9), we found significant bone decreases postpartum, especially in women with extended lactation. We also confirm previous results by showing that the decreases in aBMD are at first


doi: 10.1210/jc.2014-2825


Cortical vBMD

Trabecular thickness 4,0

1,0 Duration of lactation

0,0

0-3.9 mo (n=10, 10, 10)a

-0,5

4-8.9 mo (n=42, 42, 41, 34)

-1,0

* 0.5 mo

4 mo

§ 12 mo

‡

Controls (n=21, 21, 19, 18)

1,0

0-3.9 mo (n=10, 10, 10)a

0,0 -1,0

4-8.9 mo (n=42, 42, 41, 34)

the most pronounced in the lumbar spine, whereas the decrease is more longlasting in the femoral neck (3–5, 11). At 18 months postpartum, significant decreases in aBMD compared with baseline were no longer found at any skeletal site for any lactation group in this study. Instead, the lumbar spine aBMD was significantly higher than baseline in all lactation groups at 18 months postpartum. This increase corresponds well with the decreases in aBMD at the lumbar spine during pregnancy previously reported by others (4 – 6, 10, 30). Hence, the higher values at 18 months postpartum compared with baseline probably reflects a recovery of the bone lost during pregnancy. The HR-pQCT data showed that cortical vBMD, thickness, and area at the ultradistal tibia was markedly reCortical thickness 3,0 2,0 1,0

Controls (n=21, 21, 19, 18)

0,0

0-3.9 mo (n=10, 10, 10)a

-1,0 -2,0

4-8.9 mo (n=42, 42, 41, 34)

-3,0

9 mo (n=29, 29, 23, 22)

-4,0 0.5 mo

*

§

‡

4 mo

12 mo

18 mo

Figure 2. Mean percentage change from baseline ⫾ SE in cortical thickness at the ultradistal tibia postpartum, as assessed with HR-pQCT in relation to the duration of total lactation. *, Significant change compared with baseline for lactation group 4 – 8.9 months; §, significant change compared with baseline for lactation group 9 months or longer; ‡, significant change compared with baseline for lactation group 4 – 8.9 months. a, Because of the small number of women at 18 months postpartum in the group lactating 0 –3.9 months, this measurement was excluded from all analyses.

9 mo (n=29, 29, 23, 22)

-3,0 -4,0

18 mo

Figure 1. Mean percentage change from baseline ⫾ SE in cortical vBMD at the ultradistal tibia postpartum, as assessed with HR-pQCT in relation to duration of total lactation. *, Significant change compared with baseline for lactation groups 4 – 8.9 months and 9 months or longer; §, significant change compared with baseline for lactation groups 4 – 8.9 months and 9 months or longer; ‡, significant change compared with baseline for lactation group 9 months or longer. a, Because of the small number of women at 18 months postpartum in the group lactating 0 –3.9 months, this measurement was excluded from all analyses.

% change

2,0

-2,0 9 mo (n=29, 29, 23, 22)

-1,5

-5,0

3,0

Controls (n=21, 21, 19, 18)

% change

% change

0,5

-2,0

541

0.5 mo

4 mo

12 mo

§

18 mo

‡

Figure 3. Mean percentage change from baseline ⫾ SE in trabecular thickness at the ultradistal tibia postpartum, as assessed with HR-pQCT in relation to the duration of total lactation. §, Significant change compared with baseline for lactation groups 4 – 8.9 and 9 months or longer; ‡, significant change compared with baseline for lactation groups 4 – 8.9 and 9 months or longer. a, Because of the small number of women at 18 months postpartum in the group lactating 0 –3.9 months, this measurement was excluded from all analyses.

duced during the first 12 months postpartum in women lactating 4 months or longer. Potentially the decreases in cortical thickness and area may be due to the trabecularization of the cortical bone or a result of increased porosity because of a high bone turnover state (31). Thus, it seems that lactation reduces both cortical bone quality and dimensional parameters, at least temporally. At 18 months postpartum, only cortical vBMD was significantly lower than baseline and only in women lactating 9 months or longer. Thus, the decrease in the cortical thickness and area is transient and is not observed approximately 6 months after weaning, whereas the effect of lactation is more persistent in the cortical vBMD. No decreases in cortical bone were found in women who lactated for less than 4 months or in controls. Trabecular bone volume fraction and trabecular thickness decreased in women lactating 4 months or longer, with the nadir at 12 and 18 months, respectively. A decrease in trabecular bone volume fraction was also observed in controls. This makes the interpretation of these results difficult. No changes in trabecular bone parameters were found in women lactating for less than 4 months. Hence, we found an effect of lactation mainly on cortical vBMD, thickness, and area, but also on trabecular thickness in women lactating 4 months or longer, during the first 18 months postpartum. The finding that cortical vBMD still at 18 months postpartum was lower than baseline in women with longest lactation also supports our findings from the DXA measurements, ie, that extended lactation has a long-lasting effect on the cortical-rich bone, such as the femoral neck. It further supports previous suggestions that an extended lactation causes the bone loss of the cortical-rich skeletal sites, as measured by DXA (4, 5, 11). However, the finding that trabecular thickness still was lower than baseline at


542

Brembeck et al


18 months postpartum in women lactating 4 months or longer indicates that a long lactation may also have a longlasting effect on trabecular bone. The findings of long-lasting decreases in cortical bone in women with extended lactation are especially interesting, considering that Swedish women in general have a long duration of lactation (21). Whether these decreases in vBMD, microstructural and dimensional parameters persist and may implicate increased risk of fracture in later life is unknown. However, Melton et al (24) showed among postmenopausal women that cortical and trabecular vBMD, cortical thickness and cortical area were major determinants of fracture risk. Thus, further investigations with follow-up periods longer than 18 months are needed to elucidate whether women with extended lactation fully recover their skeletal health after weaning. Recently Sawo et al (15) showed among Gambian women that successive periods of long lactation were not associated with persistent decreases in aBMD, which points toward the observation that a full recovery takes place also after an extended lactation. One limitation of our study was the difficulty in recruiting formula-feeding women or women with a very short duration of lactation (ⱕ3.9 mo). Another limitation was the study dropouts, which were mainly due to new pregnancies and which especially resulted in incomplete follow-up data for the group lactating 0 –3.9 months at 18 months postpartum. Furthermore, the lack of prepregnancy values for bone parameters made it impossible to evaluate to what degree the women had recovered to their prepregnancy bone mineral density values after lactation. An additional limitation is that only measurements of the ultradistal tibia from the HR-pQCT were reported in this study. The percentage of women in this study who were lactating at 4 months postpartum was 88%, compared with 76% in the whole country (21). Earlier studies have shown that women with a higher level of education breast-feed for a longer period than women with a lower level of education (32). In our study, 80% of the postpartum women and 71% of the controls had studied for 3 years or more at a university, which is a higher proportion than in the general population of women at the same age (37%) (33). The mean age in postpartum women was 32.9 years, which is slightly higher than comparable national data (34). The mean self-reported prepregnancy body weight in postpartum women was 64.1 kg and the mean measured body weight in controls at baseline was 63.4 kg, which are both lower than the values at the same age in the general population of 67.0 kg (35). Thus, the women in this study weigh less and have a higher education level than the general population, why the interpretation of the results has


to be handled with care. Also, these women are possibly more health conscious, since they have actively chosen to participate in a longitudinal study. In conclusion, decreased cortical vBMD and thickness and trabecular thickness at ultradistal tibia postpartum were found in women with a long lactation. At 18 months postpartum, the cortical vBMD and trabecular thickness were still lower than baseline in women lactating 9 months or longer. Longer follow-up periods are needed to confirm whether these changes could potentially lead to an increased risk of fracture in later life.

Acknowledgments We thank all the volunteers. We also thank research nurse Ulrika Hjertonsson and Daniel Sundh for performing the bone measurements, laboratory assistant Elisabeth Gramatkovski and research nurse Anna Folino for help with the data collection and Dr Ann Laskey for valuable comments. Address all correspondence and requests for reprints to: Petra Brembeck, MS, Department of Internal Medicine and Clinical Nutrition, Sahlgrenska Academy, University of Gothenburg, Box 459, 405 30 Gothenburg, Sweden. E-mail: [email protected]. This work was supported by The Swedish Research Council Formas (Grants 2007–398 and 2009 –1504), The Graduate School Environment and Health, The Swedish Nutrition Foundation, Willhelm and Martina Lundgren Foundation, Magnus Wiberg Foundation, Fredrik and Ingrid Thuring Foundation, Olof Johannisson Foundation, The Swedish Society of Medicine, Sahlgrenska University Hospital Foundation, and Gustaf V’s and Queen Victoria’s Freemason Foundation. Disclosure Summary: The authors have nothing to disclose.

References 1. Laskey MA, Prentice A, Hanratty LA, et al. Bone changes after 3 mo of lactation: influence of calcium intake, breast-milk output, and vitamin D-receptor genotype. Am J Clin Nutr. 1998;67:685– 692. 2. Laskey MA, Prentice A, Shaw J, et al. Breast-milk calcium concentrations during prolonged lactation in British and rural Gambian mothers. Acta Paediatr Scand. 1990;79:507–512. 3. Sowers M, Corton G, Shapiro B, et al. Changes in bone density with lactation. JAMA. 1993;269:3130 –3135. 4. Karlsson C, Obrant KJ, Karlsson M. Pregnancy and lactation confer reversible bone loss in humans. Osteoporos Int. 2001;12:828 – 834. 5. Moller UK, Vieth Streym S, Mosekilde L, Rejnmark L. Changes in bone mineral density and body composition during pregnancy and postpartum. A controlled cohort study. Osteoporos Int. 2012;23: 1213–1223. 6. Drinkwater BL, Chesnut CH 3rd. Bone density changes during pregnancy and lactation in active women: a longitudinal study. Bone Miner. 1991;14:153–160. 7. Holmberg-Marttila D, Sievanen H, Tuimala R. Changes in bone mineral density during pregnancy and postpartum: prospective data on five women. Osteoporos Int. 1999;10:41– 46. 8. Karlsson MK, Ahlborg HG, Karlsson C. Maternity and bone mineral density. Acta Orthop. 2005;76:2–13.


doi: 10.1210/jc.2014-2825

9. Laskey MA, Prentice A. Bone mineral changes during and after lactation. Obstet Gynecol. 1999;94:608 – 615. 10. More C, Bettembuk P, Bhattoa HP, Balogh A. The effects of pregnancy and lactation on bone mineral density. Osteoporos Int. 2001; 12:732–737. 11. Hopkinson JM, Butte NF, Ellis K, Smith EO. Lactation delays postpartum bone mineral accretion and temporarily alters its regional distribution in women. J Nutr. 2000;130:777–783. 12. Krebs NF, Reidinger CJ, Robertson AD, Brenner M. Bone mineral density changes during lactation: maternal, dietary, and biochemical correlates. Am J Clin Nutr. 1997;65:1738 –1746. 13. Polatti F, Capuzzo E, Viazzo F, Colleoni R, Klersy C. Bone mineral changes during and after lactation. Obstet Gynecol. 1999;94:52– 56. 14. Kolthoff N, Eiken P, Kristensen B, Nielsen SP. Bone mineral changes during pregnancy and lactation: a longitudinal cohort study. Clin Sci (Lond). 1998;94:405– 412. 15. Sawo Y, Jarjou LM, Goldberg GR, Laskey MA, Prentice A. Bone mineral changes after lactation in Gambian women accustomed to a low calcium intake. Eur J Clin Nutr. 2013;67:1142–1146. 16. Affinito P, Tommaselli GA, di Carlo C, Guida F, Nappi C. Changes in bone mineral density and calcium metabolism in breastfeeding women: a one year follow-up study. J Clin Endocrinol Metab. 1996; 81:2314 –2318. 17. Tsvetov G, Levy S, Benbassat C, Shraga-Slutzky I, Hirsch D. Influence of number of deliveries and total breast-feeding time on bone mineral density in premenopausal and young postmenopausal women. Maturitas. 2014;77:249 –254. 18. Bjornerem A, Ahmed LA, Jorgensen L, Stormer J, Joakimsen RM. Breastfeeding protects against hip fracture in postmenopausal women: the Tromso study. J Bone Miner Res. 2011;26:2843–2850. 19. Wiklund PK, Xu L, Wang Q, et al. Lactation is associated with greater maternal bone size and bone strength later in life. Osteoporos Int. 2012;23:1939 –1945. 20. Sowers M, Randolph J, Shapiro B, Jannausch M. A prospective study of bone density and pregnancy after an extended period of lactation with bone loss. Obstet Gynecol. 1995;85:285–289. 21. The National Board of Health and Welfare. Breast-Feeding and Smoking Habits Among Parents of Infants Born in 2011. Stockholm: National Board of Health and Welfare; 2013. 22. Olausson H, Goldberg GR, Laskey MA, Schoenmakers I, Jarjou LM, Prentice A. Calcium economy in human pregnancy and lactation. Nutr Res Rev. 2012;25:40 – 67.


543

23. Kalkwarf HJ, Specker BL. Bone mineral loss during lactation and recovery after weaning. Obstet Gynecol. 1995;86:26 –32. 24. Melton LJ 3rd, Riggs BL, van Lenthe GH, et al. Contribution of in vivo structural measurements and load/strength ratios to the determination of forearm fracture risk in postmenopausal women. J Bone Miner Res. 2007;22:1442–1448. 25. World Health Organization. Declaration of Helsinki: ethical principles for medical research involving human subjects. Bull World Health Organ. 2001;79:373–374. 26. Rudang R, Darelid A, Nilsson M, Mellstrom D, Ohlsson C, Lorentzon M. X-ray-verified fractures are associated with finite element analysis-derived bone strength and trabecular microstructure in young adult men. J Bone Miner Res. 2013;28:2305–2316. 27. Nilsson M, Ohlsson C, Sundh D, Mellstrom D, Lorentzon M. Association of physical activity with trabecular microstructure and cortical bone at distal tibia and radius in young adult men. J Clin Endocrinol Metab. 2010;95:2917–2926. 28. Burghardt AJ, Pialat JB, Kazakia GJ, et al. Multicenter precision of cortical and trabecular bone quality measures assessed by high-resolution peripheral quantitative computed tomography. J Bone Miner Res. 2013;28:524 –536. 29. Cole TJ. Sympercents: symmetric percentage differences on the 100 log (e) scale simplify the presentation of log transformed data. Stat Med. 2000;19:3109 –3125. 30. Olausson H, Laskey MA, Goldberg GR, Prentice A. Changes in bone mineral status and bone size during pregnancy and the influences of body weight and calcium intake. Am J Clin Nutr. 2008;88:1032– 1039. 31. Bjornerem A, Ghasem-Zadeh A, Bui M, et al. Remodeling markers are associated with larger intracortical surface area but smaller trabecular surface area: a twin study. Bone. 2011;49:1125–1130. 32. Flacking R, Nyqvist KH, Ewald U. Effects of socioeconomic status on breastfeeding duration in mothers of preterm and term infants. Eur J Public Health. 2007;17:579 –584. 33. Statistics Sweden. Educational attainment of the population 2010. http://www.scb.se/statistik/UF/UF0506/2010A01B/UF0506_ 2010A01B_SM_UF37SM1101.pdf. Accessed April 23, 2013. 34. Statistics Sweden. Äldre mammor vanligare förr. http://www.scb.se/ statistik/_publikationer/LE0001_2013K01_TI_02_A05TI1301.pdf. Accessed April 23, 2013. 35. Statistics Sweden. BMI, vikt och längd: medelvärden, 1988 –1989, 2008 –2011. http://www.scb.se/sv_/Hitta-statistik/Statistik-efteramne/Levnadsforhallanden/Levnadsforhallanden/Undersokningarnaav-levnadsforhallanden-ULFSILC/12202/12209/. Accessed April 23, 2013.


Relationship between intima-media thickness and bone mineral density in postmenopausal women: a cross-sectional study.

Effect of prior oophorectomy on changes in bone mineral density and carotid artery intima-media thickness in postmenopausal women.

Combined cortical thickness and bone density determination by photon absorptiometry.

Muscle Deficits in Rheumatoid Arthritis Contribute to Inferior Cortical Bone Structure and Trabecular Bone Mineral Density.

Cortical thickness and brain volumetric analysis in body dysmorphic disorder.

Is the cortical thickness index a valid parameter to assess bone mineral density in geriatric patients with hip fractures?

Longitudinal changes in cortical thickness in autism and typical development.

Mean wall thickness of trabecular bone packets in the human iliac crest: changes with age.

Bone mineral density and mammographic density in Mexican women.

Ischemic heart disease is associated with lower cortical volumetric bone mineral density of distal radius.

Bone mineral density and abstention-induced changes in bone and mineral metabolism in noncirrhotic male alcoholics.

Inverse Correlation at the Hip Between Areal Bone Mineral Density Measured by Dual-Energy X-ray Absorptiometry and Cortical Volumetric Bone Mineral Density Measured by Quantitative Computed Tomography.

The effect of restriction of dietary calcium on trabecular and cortical bone mineral density in the rats.

Adult Tph2 knockout mice without brain serotonin have moderately elevated spine trabecular bone but moderately low cortical bone thickness.

Heritable changes in regional cortical thickness with age.

Bone mineral density and anorexia nervosa in women.

Coffee consumption and bone mineral density in korean premenopausal women.

Central Adiposity and Cortical Thickness in Midlife.

Iliac trabecular bone formation predicts radial trabecular bone density changes in type 1 osteoporosis.

Evaluation of endometrial thickness and bone mineral density based on CYP2D6 polymorphisms in Turkish breast cancer patients receiving tamoxifen treatment.

Trends in trabecular architecture and bone mineral density distribution in 152 individuals aged 30-90 years.

Comparison of mandibular cortical thickness and QCT-derived bone mineral density (BMD) in survivors of childhood acute lymphoblastic leukemia: a retrospective study.

The role of inter-prosthetic distance, cortical thickness and bone mineral density in the development of inter-prosthetic fractures of the femur: a biomechanical cadaver study.

Effect of concomitant vitamin D deficiency or insufficiency on lumbar spine volumetric bone mineral density and trabecular bone score in primary hyperparathyroidism.