Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Body composition, energy expenditure and physical activity

Body composition in overweight and obese women postpartum: bioimpedance methods validated by dual energy X-ray absorptiometry and doubly labeled water



Obesity, pregnancy and lactation all affect body composition. Simple methods to estimate body composition are useful in clinical practice and to evaluate interventions. In overweight and obese lactating women, such methods are not fully validated. The objective of this study was to validate the accuracy and precision of bioimpedance spectroscopy (BIS) by Xitron 4200 and 8-electrode multifrequency impedance (multifrequency bioimpedance analysis, MFBIA) by Tanita MC180MA with the reference methods dual energy X-ray absorptiometry (DXA) and doubly labeled water (DLW) for the assessment of body composition in 70 overweight and obese women postpartum.


The LEVA-study (Lifestyle for Effective Weight loss during lactation) consisted of an intervention and follow-up with three assessments at 3, 6 and 15 months postpartum, which made possible the validation of both accuracy and precision. Mean differences between methods were tested by a paired t-test and Bland–Altman plots for systematic bias.


At baseline, BIS and MFBIA underestimated fat mass (FM) by 2.6±2.8 and 8.0±4.2 kg compared with DXA (P<0.001) but without systematic bias. BIS and MFBIA overestimated total body water (TBW) by 2.4±2.2 and 4.4±3.2 kg (P<0.001) compared with DLW, with slight systematic bias by BIS. BIS correctly estimated muscle mass without systematic bias (P>0.05). BIS overestimated changes in TBW (P=0.01) without systematic bias, whereas MFBIA varied greatly and with systematic bias.


BIS underestimates mean FM compared with DXA but can detect mean changes in body composition, although with large limits of agreement. BIS both accurately and precisely estimates muscle mass in overweight and obese women postpartum. MFBIA underestimates FM and overestimates TBW by proprietary equations compared with DXA and DLW.


There is an increasing demand for valid, affordable and user-friendly methods of measuring body composition, especially in overweight or obese individuals. For many women, overweight and obesity result from childbearing. The excessive weight may result from extensive gestational weight gain, postpartum weight retention and/or postpartum weight gain.1 In richer societies, there is a pattern of only modest loss of weight and body fat during lactation,2 especially during the first 2 months postpartum.3 The reproductive period may be a target period for lifestyle intervention, especially because the experience of becoming a parent may be linked to desires for a healthier lifestyle.4 Here, monitoring of body composition is valuable to evaluate compliance to interventions, in addition to weighing, as the proportion of fat and body water changes during pregnancy and lactation.5 Feedback from these measurements can also be used to reinforce changes in diet and physical activity during lifestyle intervention.6, 7

Body composition can be assessed by a variety of methods, with dual energy X-ray absorptiometry (DXA) being frequently used because of its ability to precisely assess both fat and fat-free mass (FFM), with insignificant radiation exposure.8 However, the DXA equipment is expensive and accessible only in body composition laboratories or for bone density measurements. Bioimpedance analysis (BIA) is a simple, portable and easy to use method for the assessment of body composition, which has been favorably validated for use in many populations with normal hydration but less favorably in altered hydration states.9, 10 In the latter case, different impedance approaches have been proposed, such as multifrequency-BIA (MFBIA) and bioelectrical impedance spectroscopy (BIS). We have found BIS derived by Xitron 4200, and performed as stated in the manual,11 to correctly estimate fat mass (FM) and FFM compared with DXA in elderly Swedes12 but to underestimate FFM compared with DXA in cancer patients.13 Total body skeletal muscle mass can also be estimated by BIS with predictive equations based on DXA in elderly people.12

Using Xitron 4000 in a study on 68 lactating South African women with a mean body mass index of 26, Papathakis et al.14 found BIS to underestimate total body water (TBW) compared with dilution methods and thus to overestimate body fat. BIS derived by Xitron 4200 has been reported to correctly estimate TBW compared with dilution methods in early pregnancy and early postpartum but to underestimate TBW in late pregnancy in 21 healthy Swedish women with on average normal body weight.5 Multifrequency bioimpedance analysis by Bodystat 4000 has also been reported to underestimate TBW during lactation and hence overestimate body fat.15

We have found 8-electrode MFBIA by Tanita using proprietary equations unknown to us to underestimate FM and overestimate FFM and to be both biased and imprecise compared with air displacement pletysmography in obese Swedish women.16

Thus, BIS methods have been found both to give fair estimates of TBW postpartum in small well-controlled studies, but to be biased in larger studies, whereas MFBIA methods have been scarcely used.

We recently reported interventional results on body composition from a randomized 12-week trial postpartum on dietary and exercise behavior modification in overweight and obese women.7 The trial included detailed repeated measurements of body composition using both reference methods, as well as more simple methodology. Consequently, the aim of the present study was to evaluate whether bioimpedance methods, by either BIS or MFBIA, can estimate body composition expressed as FM, FFM and muscle mass (skeletal muscle mass, SMM), both cross-sectional and longitudinal, in lactating overweight and obese women compared with DXA. Furthermore, bioimpedance estimations of TBW were validated against doubly labeled water (DLW).



During 2007–2010, 72 lactating women who were overweight or obese were recruited to a randomized controlled trial (LEVA, Livsstil vid Effektiv Viktminskning under Amning; Lifestyle for effective weight loss during lactation) through advertisements at 15 antenatal care clinics in the region of Gothenburg.7 In total, 70 women, 35 overweight and 35 obese (non-smokers, singleton term delivery, birth weight >2500 g, intention to breastfeed for 6 months and no suspected or diagnosed illness on the part of the mother or infant), completed the baseline measurements. The women were randomized to four groups according to a 2 × 2 factorial study design: control group (C group) and dietary behavior modification group (D group) receiving 2.5 h of individual dietary counseling to achieve a 0.5 kg weight loss per week aided by self-weighing and bi-weekly cell phone text messages, exercise group (E group) receiving 2.5 h of individual exercise counseling to achieve 4 × 45 min of moderate intensity exercise per week aided by heart-rate self-monitoring and bi-weekly cell phone text messages, and combined diet and exercise group (DE group) receiving both D and E interventions. The 12-week interventions resulted in a 9% weight loss after the dietary treatment, which was maintained without further intervention for 9 months, whereas exercise treatment did not result in weight loss.7 Weight loss in all groups consisted mainly of FM, and reductions in muscle mass were minimal. Clinical data and body composition from baseline A, after intervention B and 1 year after baseline are presented in Table 1.

Table 1 Body composition values by reference methods and by impedance methods in overweight and obese Swedish women at baseline 8–12-week postpartum, after 12-week intervention and at 1 year after baseline

Ethics statement

The study received full ethical approval by the regional ethics board (registration number 483-06) in Gothenburg, Sweden, and is registered at as NCT01343238. All participants were informed about study protocol and signed an informed consent sheet.

Study design

This study reports on results from body composition measurements at baseline (8–12 weeks postpartum), after 12 weeks of lifestyle intervention (20–24 weeks postpartum) and 1 year from baseline (15 months postpartum) among 70 women who volunteered to participate in the trial.7

Thus, three cross-sectional validations of bioimpedance vs DXA could be performed:

A: baseline (8–12-week postpartum), B: after a 12-week intervention and C: 1 year from baseline. By analyzing the changes between the cross-sectional measurements, two longitudinal validations were carried out, named AB, with changes over 3 months, (including DLW) and AC, with changes over 1 year (without DLW).

Abbreviations: BIS, bioimpedance spectroscopy; DLW, doubly labeled water; DXA, dual X-ray absorptiometry; MFBIA, Multifrequency bioimpedance.


Baseline measurements were all carried out within a few hours at the laboratories of the Department of Internal Medicine and Clinical Nutrition, University of Gothenburg, Sweden, and at the Sahlgrenska University Hospital, Sweden, after an overnight fast. Anthropometric measurements included body weight to the nearest 0.1 kg in light underclothing on an electronic scale (MC-180 MA III, Tanita, Tokyo, Japan) and body height by a wall-mounted stadiometer. The same body height was used for all body composition analyses. In addition to measuring body weight, Tanita MC-180 MA III is an 8-electrode multifrequency bioimpedance device that measures body composition in the upright position by proprietary equations (not presented in the manual), with a manufacturer-reported accuracy of 2%.17 FM (kg) and FFM (kg) were measured by DXA (Lunar Prodigy, GE Lunar Corp, Madison, WI, USA), with an in-house precision of 2.2% for FM and 0.8% for FFM, calculated as the coefficient of variation in duplicate measurements. Total body SMM was derived from appendicular lean soft mass from DXA as described by model 1 by Kim et al.18 FM and FFM, as well as the Cole-Cole parameters ReBIS, RiBIS cell membrane capacitance (McBIS) and critical frequency (FcBIS), were at the same occasions measured by BIS (Xitron Hydra 4200, Xitron Technologies, San Diego, CA, USA), with a in-house precision expressed as coefficient of variation in duplicate measurements of 0.4% for FFM. The subjects rested in the supine position for 5 min before the tetrapolar whole body measurement with electrodes on the dorsal surface of the right hand/wrist and at the right foot/ankle according to the manufacturer’s instructions.11 All measurements were taken in duplicate, and the mean value was used. Extracellular and intracellular water by BIS, ECWBIS and ICWBIS, were calculated from Hanai mixture theory using Xitron software:11, 19

where ρECW is extracellular resistivity (female: 39 Ωcm, male: 40.5 Ωcm), H is body height (cm), W is body weight (kg), D is body density (1.05 kg/l) and KB=4.3 is a shape factor.11

where total body resistivity ρTBW was calculated as

and ρICW is intracellular resistivity (female: 264.9 Ωcm, male: 273.9 Ωcm).

All fits were classified as excellent or good.

The equation used by the BIS proprietary software to predict FFM (FFMBIS) is as follows:

where δECW is the mean density of the ECW and its associated materials (1.106 kg/l), δICW is the mean density of the ICW and its associated materials (1.521 kg/l), ECWBIS is the predicted total extracellular fluid volume (litres) and ICWBIS is the predicted total intracellular fluid volume (litres).11

Phase angle (PaBIS) or Φ in degrees at 50 kHz was derived from BIS as Φ=57.297 × a tan(Xc/R), where Xc is the reactance and R the resistance,20 and SMM was derived from BIS as described by equation (2) by Tengvall et al.12

TBW was measured at baseline and after 12 weeks of intervention by isotope dilution of both D2O and H2O18, separately as ND/1.04 and NO18/1.01 and as their mean, through the DLW technique as described.21 TBW was also recalculated from FFM by bioimpedance methods as FFMBIS × 0.73=TBWBIS0.73 and FFMMFBIA × 0.73=TBWMFBIA0.73 and from DXA as FFM × 0.73=TBWDXA0.73 by convention, given that not all women were lactating. Likewise was FFM recalculated from TBW from DLW as TBW/0.73=FFMDLW0.73.

Statistical analyses

Results are presented as mean and s.d. Normal distribution was tested by one sample Kolmogorov's tests. Differences between BIS, DXA or DLW and between measurements at baseline, 12 weeks and 1 year were assessed by paired t-tests. The performance of BIS and MFBIA in relation to the reference methods DXA and DLW was evaluated with Bland–Altman plots.22 Correlations between methods were expressed as Kendall's concordance correlation but in addition as rPearson as this is used in the Bland–Altman plots' regressions. Systematic bias between methods was defined as significant regression in Bland–Altman plots. All analyses were performed using SPSS version 21.0 (IBM, Somers, NY, USA).


As stated in the original presentation of the study, the participating women were mainly white and on average well educated. There were 70 women completing both DXA and BIS measurements at baseline, whereas 53 also completed MFBIA measurements. DLW analyses were completed at baseline by 67 women and by 64 after intervention. In total, 63 women had body composition assessed at the end of intervention and 52 at the follow-up 1 year after baseline.

Body composition at baseline, after 12-week intervention, at follow-up after 1 year, as well as the changes from baseline to follow-up, is presented in Table 1. After 1 year, FM, body mass index and extracellular fluid decreased, whereas membrane capacitance and phase angle increased.

Differences in the estimation of FM, SMM and TBW by DXA, DLW and bioimpedance methods at baseline (A) are presented in Table 2 and in Figures 1 and 2. Kendall's concordance correlations were 0.76 (BIS) and 0.71 (MFBIA) for the estimation of FM compared with DXA (P<0.01 for both). However, both BIS and even more so MFBIA underestimated FM (Figure 2a and b) (and thus overestimated FFM), with broad limits of agreement, but with no differences in bias between baseline and after intervention or at follow-up (the latter data not shown).

Table 2 Differences between DXA, DLW and bioimpedance methods for the assessment of body composition at baseline in overweight and obese initially lactating women postpartum
Figure 1

Comparison of body composition methods in overweight and obese lactating women at baseline, measurement A, n=53–70. Muscle estimate by Tanita proprietary equation was biased 32.1±4.1 kg compared with DXA (n=52) and therefore not shown. Boxes represent interquartiles and means; whiskers show ranges. BIS (Xitron 4200); MFBIA (Tanita MC-180 MA III).

Figure 2

Bland–Altman plots comparing body composition by BIS and MFBIA compared with DXA and DLW in overweight and obese lactating women at baseline, measurement A, n=49–68. Horizontal dotted line=mean difference (kg). Limits of agreement correspond to±2 s.d. Regression line=difference between reference method minus bioimpedance method as dependent variable, and mean of reference method minus bioimpedance method as independent variable, with rPearson and P-values inserted. (a and b) Differences in FM by BIS and MFBIA compared with DXA. (c) Differences in SMM by BIS compared with DXA. (d and e) Differences in TBW by BIS and MFBIA compared with DLW. (f) Differences in TBW by DXA calculated as FFM*0.73 compared with DLW. BIS (Xitron 4200); MFBIA (Tanita MC-180 MA III).

Muscle mass was correctly estimated by BIS using a predictive equation developed for elderly12 and with no systematic bias related to the average SMM, according to Bland–Altman plots (Figure 2c). Muscle mass expressed from proprietary equations by MFBIA was implausible at 53.4±5.6 kg at baseline, with a bias of 31.1±4.1 kg compared with DXA.

TBW by proprietary equations was overestimated compared with DLW by BIS and even more so by MFBIA (Figures 2d and e). TBW estimated from 73% of FFM slightly improved accuracy by BIS but on the contrary slightly deteriorated accuracy by MFBIA. Most accurate (true) and precise (narrowest limits of agreement) estimation of TBW compared with DLW was obtained by using 73% of FFM as assessed by DXA (Figure 2f).

Differences in the estimation of changes in FM, TBW and SMM by DXA and bioimpedance methods are presented in Table 3 and Figure 3. BIS underestimated changes in FM between baseline and end of intervention (AB) but correctly estimated the overall changes in FM over 1 year, although with broad limits of agreement (Figure 3a). Changes in FM between baseline and end of intervention, as well as between baseline and follow-up, assessed by DXA and BIS were also significantly correlated, rKendall=0.61 and 0.80 (P<0.01 for both) and without systematic bias.

Table 3 Differences between DXA, DLW and BIS (by Xitron and Tanita) to assess changes in body composition (fat mass, total body water and muscle mass), between baseline and after 12-week intervention (AB) and between baseline and 1 year later (AC) in overweight and obese initially lactating women postpartum
Figure 3

Bland–Altman plots comparing differences between DXA, DLW and BIS+MFBIA to assess changes in body composition between baseline and after 12-week intervention (AB) in overweight and obese initially lactating women postpartum. Horizontal dotted line=mean difference (kg). Limits of agreement correspond to±2 s.d. Regression line=difference between reference method minus bioimpedance method as dependent variable, and mean of reference method minus bioimpedance method as independent variable, with rPearson and P-values inserted. (a and b) Differences in FM changes by BIS and MFBIA compared with DXA. (c and d) Differences in TBW changes by BIS and MFBIA compared with DLW. (e and f) Differences in SMM changes by BIS and MFBIA compared with DXA. BIS (Xitron 4200); MFBIA (Tanita MC-180 MA II).

MFBIA showed no significant difference in FM changes compared with DXA, partly due to larger variation, and no systematic bias (Figure 3b),

BIS overestimated changes in TBW compared with DLW but with no systematic bias (Figure 3c). MFBIA did not significantly underestimate TBW but showed systematic bias (Figure 3d).

BIS correctly estimated changes in SMM compared with DXA between baseline and end of intervention (AB) (Figure 3e), although with systematic bias, and slightly overestimated changes over 1 year (AC). MFBIA overestimated changes in SMM compared with DXA, with clear systematic bias and with broad limits of agreement (Figure 3f).


We have found both BIS by Xitron and MFBIA by Tanita to underestimate FM, and hence to overestimate FFM compared with DXA in overweight and obese women postpartum, and also to overestimate TBW compared with DLW. On the other hand, we found BIS to correctly estimate SMM compared with DXA, without systematic bias, whereas MFBIA overestimated SMM with ≈140% compared with DXA. Both BIS and MFBIA could occasionally detect changes in body composition correctly, although for SMM estimation with systematic bias, and especially for MFBIA, with large limits of agreement.

Body composition by impedance methods in overweight and obese subjects

We have previously reported BIS by Xitron 4200 to accurately estimate FM in mostly overweight 75-year-old Swedish citizens and moreover to accurately estimate SMM compared with DXA by Lunar Prodigy.12

We recently reported eight-electrode multifreqency-BIA (MFBIA) by Tanita MC-180 MA III, the same equipment as used in this study, to underestimate FM compared with air displacement plethysmography in obese women, and with considerable imprecision.16 Using the same type of Tanita device in healthy young adults, Leahy et al. noted an underestimation of FM compared with DXA, especially in women, which increased with higher body fatness.23 Using a different type of Tanita, an eight-electrode single-frequency BD 418, in obese and overweight Swedish women, Neovius et al.24 found it to underestimate FM by on average 3.6 kg compared with DXA by Lunar Prodigy, again with increasing bias with higher body fatness. These findings thus support our results in that several Tanita impedance devices underestimate FM in overweight and obese women.

Still others have found different MFBIA devices such as Impedimed SFB7 to give good absolute and relative agreement with DXA measurement in overweight and obese women during weight loss, whereas a single-frequency BIA device (Tanita Ultimate Scale Model 2000) showed poor absolute agreement with overestimation of FM but was still able to detect changes in body composition.25 Thus, different Tanita devices appear to be biased in opposite direction for the estimation of FM, as compared with DXA, but are able to detect changes in body composition in overweight women after weight loss. Again, this latter finding is coherent with our results.

Others have reported eight-electrode multifreqency-BIA by other manufacturers such as Inbody 3.0 to give accurate estimations of TBW and extracellular water compared with dilution methods also in severe obesity,26 thus underlining that different impedance devices perform differently, probably owing to different assumptions in the model algorithms.

Body composition in women postpartum

Validation studies with impedance methods have mostly been performed with BIS devices among women postpartum. During pregnancy and directly postpartum, BIS estimation of total body water (TBW) based on mixture theory using Xitron 4000 software27 has been shown to correlate fairly well with isotope dilution methods in 10 healthy and mostly normal weight women during pregnancy and early postpartum.28 BIS by the same device underestimated TBW compared with dilution and hence overestimated FM in 68 south African women with average body mass index of 26.14 Using dilution methods as references, Löf and Forsum5 found BIS by Xitron 4200, using both resistivity coefficients given by the manufacturer,11 and derived from dilution results, to correctly estimate body water compartments in 21 healthy women 2 weeks postpartum. In 44 Cameroonian lactating women, the use of 3- out of 4-frequency BIA (Bodystat Quadscan 4000) was found to overestimate FM compared with dilution, despite using 12 different predictive equations.15

Thus, BIS by Xitron 4000 has been more frequently tried and found useful, including postpartum, albeit with fairly small numbers of women, which might have reduced the ability of these studies to detect differences between methods. In addition, previous validation studies have included assessment also during pregnancy, thereby making it necessary to use dilution methods instead of DXA for safety reasons, hence reducing the comparability with the present study.

Thus, our present results regarding underestimation of body fat by MFBIA Tanita in overweight and obese women postpartum are in close agreement with our earlier report of non-lactating overweight and obese women,16 but it is hardly possible to compare the performance of other MFBIA devices as the proprietary equations are unknown. The use of algorithms inbuilt into instrument firmware is tempting, but should not be relied upon, as suggested in a recent review.29 Obviously there is a problem with the muscle equation for Tanita MC-180 MA III as the output entitled ‘Total muscle mass’ gives a bias of >30 kg when compared with DXA. As FFM and SMM by MFBIA were on average 56.5 kg and 53.4 kg respectively, it seems probably that Tanita MC-180 MA III calculates SMM as FFM subtracted by an (to us unknown) estimate of bone mass, as BMC by DXA in the same women amounted to 3.0 kg.

Strengths and limitations

We have used accepted reference methods to validate impedance methods in overweight and obese initially lactating women postpartum. In addition, the trial included three repeated measurements per woman. At baseline, all women were lactating; after 12 weeks, 94% were lactating; but after 1 year, only 7% were lactating, thus complicating the longitudinal changes but not affecting the method comparisons. The number of women included was based on a power calculation for the original intervention study and is slightly greater than any other validation study so far reported. The numbers of MFBIA assessments were for logistical reasons lower than the numbers of BIS assessments, thus reducing the power to detect differences in MFBIA compared with DXA and DLW. The TBW by DLW dilution was calculated by regression and not from a 3 h post-exposition concentration as in earlier studies, which might reduce the comparability.

BIS was measured in the supine position and MFBIA in upright position according to the manufacturers' instructions. This introduces systematic bias as the impedance will increase approximately 5 Ohm after 30 min in the supine position.30 Still, the devices are designed to produce estimates of the same variables FM and TBW by their proprietary equations and should thus be comparable.


Bioimpedance method estimations of FM and TBW by proprietary equations by BIS Xitron 4200 slightly underestimate mean FM without systematic bias and overestimate body water with slight systematic bias compared with DXA and DLW in overweight and obese women postpartum but with broad limits of agreement. BIS Xitron 4200 can detect mean changes in body composition and, in addition, both accurately and precisely estimates muscle mass.

Bioimpedance method estimations by proprietary equations by MFBIA Tanita MC-180 MA III grossly underestimate FM without systematic bias compared with DXA, and overestimate body water without systematic bias compared with DLW, although with very broad limits of agreement, and present unplausible estimates of muscle mass.

MFBIA Tanita MC-180 MA III can detect mean changes in body composition but with systematic bias and very broad limits of agreement in body water and muscle mass.

The bioimpedance devices used in this study seem to be too imprecise to be used for individual evaluation, because of large limits of agreement, in both absolute comparisons and in changes over time, compared with reference methods. Still, BIS by Xitron could deliver correct estimate of muscle mass on a group level.

Bioimpedance instruments per se measure electrical impedance parameters. Any other BIS impedance measuring instrument, when correctly calibrated, should provide similar raw, which could then be used in the Xitron algorithms. Thus, the current findings are equally applicable to any BIS device that implements (or the researchers implement) the Xitron/Hanai/Mixture models.


  1. 1

    Gunderson EP . Childbearing and obesity in women: weight before, during, and after pregnancy. Obstet Gynecol Clin North Am 2009; 36: 317–32 ix.

    Article  Google Scholar 

  2. 2

    Winkvist A, Rasmussen KM . Impact of lactation on maternal body weight and body composition. J Mammary Gland Biol Neoplasia 1999; 4: 309–318.

    CAS  Article  Google Scholar 

  3. 3

    Sadurskis A, Kabir N, Wager J, Forsum E . Energy metabolism, body composition, and milk production in healthy Swedish women during lactation. Am J Clin Nutr 1988; 48: 44–49.

    CAS  Article  Google Scholar 

  4. 4

    Amorim AR, Linne YM, Lourenco PM . Diet or exercise, or both, for weight reduction in women after childbirth. Cochrane Database Syst Rev 2007; CD005627.

  5. 5

    Lof M, Forsum E . Evaluation of bioimpedance spectroscopy for measurements of body water distribution in healthy women before, during, and after pregnancy. J Appl Physiol 2004; 96: 967–973.

    Article  Google Scholar 

  6. 6

    Ware LJ, Hurling R, Bataveljic O, Fairley BW, Hurst TL, Murray P et al. Rates and determinants of uptake and use of an internet physical activity and weight management program in office and manufacturing work sites in England: cohort study. J Med Internet Res 2008; 10: e56.

    Article  Google Scholar 

  7. 7

    Bertz F, Brekke HK, Ellegård L, Rasmussen KM, Wennergren M, Winkvist A . Diet and exercise weight-loss trial in lactating overweight and obese women. Am J Clin Nutr 2012; 96: 698–705.

    CAS  Article  Google Scholar 

  8. 8

    Jebb S . Measurement of soft tissue composition by dual energy X-ray absorptiometry. Br J Nutr 1997; 77: 151–163.

    CAS  Article  Google Scholar 

  9. 9

    Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Manuel Gómez J et al. Bioelectrical impedance analysis-part II: utilization in clinical practice. Clin Nutr 2004; 23: 1430–1453.

    Article  Google Scholar 

  10. 10

    Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Gómez JM et al. Bioelectrical impedance analysis—part I: review of principles and methods. Clin Nutr 2004; 23: 1226–1243.

    Article  Google Scholar 

  11. 11

    Xitron Operating Manual Revision 1.01 San Diego, USA, Xitron Technologies, Inc 1997.

  12. 12

    Tengvall M, Ellegård L, Malmros V, Bosaeus N, Lissner L, Bosaeus I . Body composition in the elderly: reference values and bioelectrical impedance spectroscopy to predict total body skeletal muscle mass. Clin Nutr 2009; 28: 52–58.

    Article  Google Scholar 

  13. 13

    Ellegård LH, Ahlén M, Körner U, Lundholm KG, Plank LD, Bosaeus IG . Bioelectric impedance spectroscopy underestimates fat-free mass compared to dual energy X-ray absorptiometry in incurable cancer patients. Eur J Clin Nutr 2009; 63: 794–801.

    Article  Google Scholar 

  14. 14

    Papathakis PC, Rollins NC, Brown KH, Bennish ML, Van Loan MD . Comparison of isotope dilution with bioimpedance spectroscopy and anthropometry for assessment of body composition in asymptomatic HIV-infected and HIV-uninfected breastfeeding mothers. Am J Clin Nutr 2005; 82: 538–546.

    CAS  Article  Google Scholar 

  15. 15

    Medoua GN, Nana ES, Essa'a VJ, Ntsama PM, Matchawe C, Rikong HA et al. Body composition of Cameroonian lactating women determined by anthropometry, bioelectrical impedance, and deuterium dilution. Nutrition 2011; 27: 414–419.

    Article  Google Scholar 

  16. 16

    Bosaeus M, Karlsson T, Holmäng A, Ellegård L . Accuracy of quantitative magnetic resonance and eight-electrode bioelectrical impedance analysis in normal weight and obese women. Clin Nutr 2014; 33: 471–477.

    Article  Google Scholar 

  17. 17

    Anonymous. Multi-frequency body composition analyzer MC-180 MA III instruction manual Tokyo, Japan, Tanita 2008.

  18. 18

    Kim J, Heshka S, Gallagher D, Kotler DP, Mayer L, Albu J et al. Intermuscular adipose tissue-free skeletal muscle mass: estimation by dual-energy X-ray absorptiometry in adults. J Appl Physiol 2004; 97: 655–660.

    Article  Google Scholar 

  19. 19

    Matthie JR . Second generation mixture theory equation for estimating intracellular water using bioimpedance spectroscopy. J Appl Physiol 2005; 99: 780–781.

    Article  Google Scholar 

  20. 20

    Baumgartner RN, Chumlea WC, Roche AF . Bioelectric impedance phase angle and body composition. Am J Clin Nutr 1988; 48: 16–23.

    CAS  Article  Google Scholar 

  21. 21

    Slinde F, Bertz F, Winkvist A, Ellegård L, Olausson H, Brekke HK . Energy expenditure by multisensor armband in overweight and obese lactating women validated by doubly labeled water. Obesity (Silver Spring) 2013; 21: 2231–2235.

    CAS  Article  Google Scholar 

  22. 22

    Bland JM, Altman DG . Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–310.

    CAS  Article  Google Scholar 

  23. 23

    Leahy S, O'Neill C, Sohun R, Jakeman P . A comparison of dual energy X-ray absorptiometry and bioelectrical impedance analysis to measure total and segmental body composition in healthy young adults. Eur J Appl Physiol 2012; 112: 589–595.

    Article  Google Scholar 

  24. 24

    Neovius M, Hemmingsson E, Freyschuss B, Uddén J . Bioelectrical impedance underestimates total and truncal fatness in abdominally obese women. Obesity (Silver Spring) 2006; 14: 1731–1738.

    Article  Google Scholar 

  25. 25

    Thomson R, Brinkworth GD, Buckley JD, Noakes M, Clifton PM . Good agreement between bioelectrical impedance and dual-energy X-ray absorptiometry for estimating changes in body composition during weight loss in overweight young women. Clin Nutr 2007; 26: 771–777.

    Article  Google Scholar 

  26. 26

    Sartorio A, Malavolti M, Agosti F, Marinone PG, Caiti O, Battistini N et al. Body water distribution in severe obesity and its assessment from eight-polar bioelectrical impedance analysis. Eur J Clin Nutr 2005; 59: 155–160.

    CAS  Article  Google Scholar 

  27. 27

    Xitron Operating Manual San Diego, USA, Xitron Technologies, Inc 1993.

  28. 28

    Van Loan MD, Kopp LE, King JC, Wong WW, Mayclin PL . Fluid changes during pregnancy: use of bioimpedance spectroscopy. J Appl Physiol 1995; 78: 1037–1042.

    CAS  Article  Google Scholar 

  29. 29

    Ward LC . Segmental bioelectrical impedance analysis: an update. Curr Opin Clin Nutr Metab Care 2012; 15: 424–429.

    Article  Google Scholar 

  30. 30

    Slinde F, Bark A, Jansson J, Rossander-Hulthén L . Bioelectrical impedance variation in healthy subjects during 12 h in the supine position. Clin Nutr 2003; 22: 153–157.

    CAS  Article  Google Scholar 

Download references


This work was supported by the Swedish Research Council (Project grant: K2009-21091-01-3. PI Anna Winkvist) and the Swedish Council for Working Life and Social Research (Project grant: 2006-0339, PI Anna Winkvist).

Author information



Corresponding author

Correspondence to L Ellegård.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Ellegård, L., Bertz, F., Winkvist, A. et al. Body composition in overweight and obese women postpartum: bioimpedance methods validated by dual energy X-ray absorptiometry and doubly labeled water. Eur J Clin Nutr 70, 1181–1188 (2016).

Download citation

Further reading


Quick links