Steihaug et al.

BMC Geriatrics (2018) 18:65

https://doi.org/10.1186/s12877-018-0755-x

RESEARCH ARTICLE Open Access

Does sarcopenia predict change in mobility

after hip fracture? a multicenter
observational study with one-year follow-
up
Ole Martin Steihaug1 , Clara Gram Gjesdal2,3, Bård Bogen1,4, Målfrid Holen Kristoffersen5, Gunhild Lien6,
Karl Ove Hufthammer7 and Anette Hylen Ranhoff1,2,8*

Abstract
Background: Patients with hip fracture frequently have sarcopenia and are at great risk of loss of mobility. We have
investigated if sarcopenia predicts change in mobility after hip fracture.
Methods: This is a prospective, multicenter observational study with one-year follow-up. Patients with hip fracture
who were community-living and capable of walking before the fracture were included at three hospitals in Norway
(2011–2013). The primary outcome of the study was change in mobility, measured by the New Mobility Score
(NMS). Sarcopenia was determined postoperatively by anthropometry, grip strength, and NMS.
Results: We included 282 participants and sarcopenia status was determined in 201, of whom 38% (77/201) had
sarcopenia, 66% (128/194) had low muscle mass, 52% (116/222) had low grip strength and 8% (20/244) had low
pre-fracture mobility (NMS < 5). Sarcopenia did not predict change in mobility (effect 0.2 points; 95% CI –0.5 to 0.9,
P = 0.6), but it was associated with having lower mobility at one-year (NMS 5.8 (SD 2.3) vs. 6.8 (SD 2.2), P = 0.003),
becoming a resident of a nursing home (odds ratio 3.2, 95% CI 0.9 to 12.4, P = 0.048), and the combined endpoint
of becoming a resident of a skilled nursing home or death (odds ratio 3.6, 95% CI 1.2 to 12.2, P = 0.02).
Conclusions: Sarcopenia did not predict change in mobility in the year after hip fracture.
Keywords: Activities of daily living, Hip fractures, Independent living, Mobility limitation, Skilled nursing facilities,
Sarcopenia

Background low physical performance [4]. Previous studies on sarcope-

A hip fracture is associated with severe and persisting mo- nia in patients with hip fracture have been cross-sectional,
bility impairment in more than half of patients [1]. For the single-center, have included few participants or have had
last 30 years, a substantial effort has been made to under- short follow-ups [5–10]. The three components of EWG-
stand the condition of sarcopenia, and several definitions SOP sarcopenia have different associations with mobility
have been proposed [2]. Sarcopenia has recently been rec- after hip fracture. Physical performance and mobility are
ognized as an independent condition with its own ICD-10 strong determinants of mobility after hip fracture [11, 12].
code [3]. One of the most widely used definitions is by the Muscle strength is a somewhat weaker predictor [13, 14],
European Working Group on Sarcopenia in Older Persons whereas the studies on muscle mass have been inconclusive
(EWGSOP): low muscle mass with low muscle strength or [15]. Our primary hypothesis is that sarcopenia, determined
by methods suitable for bed-side use, predicts change in
mobility in the year after hip fracture and therefore that sar-
* Correspondence: [email protected]
1
Kavli Research Centre for Geriatrics and Dementia, Haraldsplass Deaconess copenia status is useful for determining prognosis and is a
Hospital, Bergen, Norway possible cause of mobility impairment. Further, we aim to
2
Department of Clinical Science, University of Bergen, Bergen, Norway describe the associations of sarcopenia and the individual
Full list of author information is available at the end of the article

© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Steihaug et al. BMC Geriatrics (2018) 18:65 Page 2 of 10

components of sarcopenia (muscle mass, grip strength and mobility, as described by the EWGSOP [4]. Total body
mobility) and adverse clinical outcomes in the year after muscle mass was determined by anthropometry by the
hip fracture: change in activities of daily living, reoperations method of Heymsfield et al. using height, arm circumfer-
for hip fracture, all-cause hospitalization, fractures, becom- ence and triceps skinfold [20]. Arm circumference was
ing a resident of a nursing home or death. measured on the right arm using a non-elastic tape at
the mid-point of the acromion and olecranon process,
Methods and triceps skinfold was measured on the posterior as-
Study design pect of the same arm at the same level using a skinfold
We conducted a prospective observational study of sar- caliper (Harpenden, Baty International, Great Britain).
copenia in patients with acute hip fracture with follow- Height was measured by a wall mounted stadiometer, or
up at three months and one year, conducted at three if the patients was unable to stand self-reported height
Norwegian hospitals in 2011–2013. was used. If the participant was unable to stand or re-
port their height, the length from heel to crown was
Participants measured while lying in bed. In cases with missing value
Participants were included while in hospital in the postop- on height at baseline, height measured at follow-up was
erative phase. Eligible participants were aged ≥65 years, used. The values for total body muscle mass were trans-
able to give informed consent as judged by experienced formed to appendicular lean mass (ALM) using model 1
clinicians, were living in the community, and were ambu- described by Kim et al. [21]. The cut-points for low
latory before the fracture. Patients who were unstable such muscle mass were ALM ≤7.25 kg/m2 for men and
as with delirium, acute respiratory failure or in severe pain ≤5.67 kg/m2 for women. We chose anthropometry for
were not eligible. Other exclusion criteria were dementia its ease of use at the bed-side in immobile hip fracture
when it made informed consent impossible, remaining life patients. Grip strength was measured with a Jamar Hy-
expectancy of less than three months and bone disease draulic Dynamometer (Sammons Preston, USA) while
other than osteoporosis or osteomalacia. We screened for the patient was sitting in bed or on a chair with the
participants by examining lists of patients admitted for hip elbow flexed, the wrist in the neutral position and with
fracture or staying on the hospital wards. verbal encouragement. Grip strength was measured
three times on each hand with short intervals between
Data collection each attempt while the grip was repositioned. The single
Information was collected by the authors and study best value out of these six measurements was used. Low
personnel by examination, chart review, routine blood grip strength was ≤30 kg for men and ≤20 kg for
tests and by interviews with patients and their caregivers women. Mobility in the two weeks before the hip frac-
from the first postoperative day and until discharge from ture was determined by interview using the New Mobil-
hospital. Weight was measured with the scales on the ity Score (NMS). The NMS is scored 0–9 according to a
hospital wards. We collected the American Society of person’s ability to walk indoors, outdoors, or while shop-
Anesthesiologists (ASA) score, Charlson comorbidity ping [22]. The cut-point for low mobility was chosen as
index [16], Barthel activities of daily living (B-ADL) < 5, as this has been used to predict mortality after hip
score [17], length of the acute care hospital stay, previ- fracture [23]. We used a Danish version of the NMS
ous hip fracture and type of hip fracture. Follow-up was with minimal modifications to Norwegian. Sarcopenia
at 3 months at an outpatient clinic and at one year as a status was determined postoperatively and at follow-up.
telephone interview with the patient or care-giver. Infor-
mation on previous and subsequent hip fractures, and Outcome measures
reoperations for the index hip fracture came from the The primary outcome was change in mobility, calculated
Norwegian Hip Fracture Register [18]. This register as NMS at one year minus the pre-fracture NMS. We
started data collection in 2005 and has coverage on an believe that change in mobility is more relevant than
estimated 90% of all hip fractures in Norway. The regis- mobility for identifying patients who are more likely to
ter has information on reoperations, with an estimated benefit from interventions. We determined mobility pre-
coverage of 65% of hip fractures treated with surgical fracture, at three months, and at one-year. All other ana-
pinning, 68% after hemiarthroplasty and 93% after total lyses were considered exploratory. Other outcome vari-
hip replacement [19]. Mortality data was supplied by the ables at one year were NMS at one year, B-ADL at one
National Population Register, which is complete. year, change in B-ADL, new clinical fractures, new hip
fractures, reoperation for hip fracture, all-cause hospital-
Sarcopenia izations, death, becoming a permanent resident of a
Participants were classified as sarcopenic if they had low skilled nursing home, and the combined endpoint of be-
muscle mass and either low grip strength or impaired coming a permanent resident of a nursing home or
Steihaug et al. BMC Geriatrics (2018) 18:65 Page 3 of 10

death. The combined endpoint was chosen because Results

death and becoming a resident of a nursing home are All patients in hospital with confirmed hip fracture were
competing risks. New clinical fracture was any symp- considered for inclusion if the research staffs at the different
tomatic skeletal fracture reported by the patient. hospitals were present. Some patients were unable to partici-
pate because they were discharged before the two-day con-
Statistical analysis sent process was completed. There was no systematic
We report descriptive data as means with standard devia- recording of the patients who were screened, but not in-
tions or as counts with percentages. To examine the pre- cluded. Figure 1 describes the progress of participants
dictive effect of sarcopenia status on changes in mobility through hip fracture, inclusion in the study and follow-up.
and level of activity of daily living, we used linear regression During the period of inclusion 1592 patients had surgery for
analyses with NMS and B-ADL as response variables and hip fracture and 282 patients were included in the study.
sarcopenia status at baseline (sarcopenic vs. not sarcopenic) Mean age was 79.4 (SD 8.2) years and 76% were female.
and age, sex and BMI as predictors. Age, sex and BMI were Mean BMI was 24.1 (SD 4.3) kg/m2, with a wide range 13.0
included in the models because they are established predic- to 44.7 kg/m 2. See Table 1 for baseline demographics. One
tors of mobility after hip fracture [24] or sarcopenia [25]. patient died during the hospital stay. Participants who had
The relationship with age and BMI was not assumed to be missing data on sarcopenia status during hospitalization
linear and was modelled using restricted cubic splines with had lower pre-fracture NMS and pre-fracture B-ADL and
3 knots, placed at the 10%, 50% and 90% quantiles. We as- were more likely to become a permanent resident of a
sumed that a one-point change in NMS would be clinically skilled nursing home. For 69 participants, height was not
significant. We used Fisher’s exact test for the analysis of assessed during the hospital stay, and for 52 of these, height
sarcopenia associated with new clinical fracture, new hip determined at follow-up was used.
fracture, reoperations, all-cause hospitalization, becoming a
resident of skilled nursing home, and the combined end- Sarcopenia
point of nursing home or death. The association between Sarcopenia status during hospitalization for hip fracture
the separate components of sarcopenia (muscle mass, grip was determined in 201 participants, and 39% (77/201)
strength and mobility) with change in mobility, change in had sarcopenia. Low muscle mass was present in 66%
B-ADL and the combined endpoint of becoming a resident (128/194) of the participants, low grip strength in 52%
of a skilled nursing home or death was analysed using re- (116/222), and 8% had low pre-fracture mobility (19/
gression analysis. Muscle mass, grip strength and mobility 243). One participant did not have muscle mass deter-
were independent continuous variables, and were analyzed mined but had grip strength and NMS above the cut-
separately. Change in mobility and change in B-ADL were points and was considered not-sarcopenic. Figure 2.
continuous, dependent variables and the combined end- Participants with sarcopenia were older, had lower BMI,
point of becoming resident of a nursing home or death was greater ASA score at operation, greater prevalence of
a dichotomous dependent variable. previous hip fracture and pulmonary disease and lower
For the regression analyses, we used multiple imput- B-ADL before the fracture. Grip strength and ALM were
ation (500 imputations), based on predictive mean match- assessed at a median of 4 days after surgery (interquar-
ing, using the ‘aregImpute()’ function in the ‘rms’ R tile range 3 to 6 days) (Fig. 2).
package [26]. The variables used in the imputation models
were the ones included in the regression models, variables Outcomes after one year
highly correlated with these variables and variables ex- Sarcopenia was not associated with change in mobility
pected to explain the missing data mechanism: NMS at at one year in unadjusted or adjusted analyses (e.g., the
baseline, follow-up, and at one year, and change in NMS change in NMS was an additional 0.2 in sarcopenic pa-
from baseline to one year, B-ADL before the hip fracture tients compared to non-sarcopenic patients, 95% CI: –
and change in B-ADL from pre-fracture until one year, 0.5 to 0.9, P = 0.6); see Table 2 for outcomes at one-year.
sarcopenia status at baseline, BMI at baseline, ASA score Sarcopenia status at hospitalization did not predict
during hip fracture surgery, previous hip fracture, serum change in mobility from pre-fracture to 3 months, or
albumin when in hospital, grip strength at follow-up, sex, from 3 months to one-year. Results were not affected by
clinical fractures and hip fractures in the year after admis- imputation of missing values. Mobility was reduced in
sion, becoming a resident of a skilled nursing home, or 54% of participants one year after hip fracture, with a
dying in the following year. All continuous predictors mean NMS of 6.4 (SD 2.2). See Fig. 3 for NMS by sarco-
were modelled linearly in the imputation model. The im- penia status during the year after hip fracture. Figure 4
putation analyses were done in R 3.3.0 [27] and the rest of describes the relationship between specific scores on the
the analyses were done in Stata 14 (Stata Corp., USA). P- NMS pre-fracture and at one-year. Participants with sar-
values ≤0.05 were considered significant. copenia had lower mobility at one-year, NMS 5.8 (SD
Steihaug et al. BMC Geriatrics (2018) 18:65 Page 4 of 10

Temporarily lost
to follow-up
All hip fractures n = 14
n = 1,592
Sarcopenic Follow-up at hospital Follow-up by
at inclusion at 3 months telephone at 1 year
n = 77 n = 59 n = 71

Not eligible, withdrew Permanently lost to Lost to follow-up (n = 2)

after study start, or follow-up (n = 4) Dead: 2
not included for other Dead: 3
reasons Other reasons: 1
n = 1,310

Temporarily lost
to follow-up
Eligible and n = 14
included in study
n = 282 Not sarcopenic Follow-up at hospital Follow-up by
Hospitals: at inclusion at 3 months telephone at 1 year
Haraldsplass: 130 n = 124 n = 110 n = 120
Diakonhjemmet: 79
Haukeland: 73
Lost to follow-up (n = 4)
Dead: 3
Other reasons: 1

Temporarily lost
to follow-up
n = 14

Not evaluated Follow-up at hospital Follow-up by

at inclusion at 3 months telephone at 1 year
n = 81 n = 56 n = 67

Permanently lost to Lost to follow-up (n = 3)

follow-up (n = 11) Dead: 2
Dead: 10 Other reasons: 1
Other reasons: 1

Fig. 1 Overview of patients with hip fracture, patients included in the study and patients returning for follow-up

2.3) vs. 6.8 (SD 2.2), P = 0.003, and greater impairment sarcopenia is not useful in determining prognosis. Fur-
in B-ADL, 16.8 (SD 4.4) vs. 18.6 (SD 2.8), P = 0.001, ther, sarcopenia did not predict change in mobility in
compared to patients without sarcopenia. Sarcopenia analysis adjusted for age, sex and BMI, which indicates
was associated with becoming a permanent resident of a that sarcopenia status is not likely to be causally related
skilled nursing home (OR 3.2, 95% CI: 0.9 to 12.4, P = to developing reduced mobility. We used multiple im-
0.048) and the combined endpoint of becoming a resi- putation to reduce the loss of information associated
dent of a skilled nursing home or death (OR 3.6, 95% with missing values. This approach is considered inferior
CI: 1.2 to 12.3, P = 0.02). to having all the data, but preferable to performing ana-
lysis on complete data. One assumption of multiple im-
Muscle mass, grip strength and mobility putation is that missing values can be estimated by the
Muscle mass or grip strength was not associated with remaining information in the dataset. The results of our
any outcome in adjusted analysis Table 3. In unadjusted analysis were similar when analyzing complete cases and
analysis, grip strength and NMS were associated with a when analyzing datasets with imputed values, indicating
reduced risk of becoming a resident of a nursing home that the results of our analysis are valid even if this as-
or death. The NMS was positively associated with sumption was erroneous.
change in B-ADL in adjusted analysis (estimate 0.2 per Change in mobility was not associated with sarcopenia
point, 95% CI 0.0 to 0.4, P = 0.03). and this was consistent across all the investigated time
periods, from baseline to three months, from baseline to
Discussion one year and from three months to one year. Mobility
The aim of this study was to investigate if sarcopenia from before the hip fracture until one year is character-
predicted change in mobility after hip fracture. We ized by an initial loss of mobility and a subsequent par-
found that sarcopenia status did not predict change in tial recovery. Sarcopenia is not associated with either the
mobility in unadjusted analysis, which indicates that loss of mobility or the recovery, which further supports
Steihaug et al. BMC Geriatrics (2018) 18:65 Page 5 of 10

Table 1 Baseline characteristics of participants by sarcopenia status that sarcopenia is not related to change in mobility. In
Not sarcopenic Sarcopenic P-value contrast to change in mobility, being sarcopenic was as-
Age, years (SD) 77.1 (7.8) n = 124 81.8 (7.6) n = 77 < 0.0001 sociated with having lower mobility pre-fracture, at three
Female, n (%) 95 (77) n = 124 56 (72) n = 77 0.5
months and at one year, compared to not being sarcope-
nic. This is expected, since low mobility is one criteria
Barthel ADL 19.5 (1.1) n = 85 18.7 (1.9) n = 60 0.006
pre-fracture (SD) for sarcopenia. As seen in Fig. 4, pre-fracture mobility is
a determinant of mobility at three months and one year.
Type of hip fracture 0.6
Savino et al. found that grip strength measured in hos-
Neck of femur, 29 (24) n = 123 14 (18) n = 77
not displaced,
pital predicted recovery of walking ability in patients
n (%) with hip fracture [13]. In contrast, our findings indicate
Neck of femur, 46 (37) n = 123 31 (40) n = 77 that neither muscle mass nor grip strength, when ana-
displaced, n (%) lysed as continuous variables, were associated with
Trochanteric, 48 (39) n = 123 32 (42) n = 77 change in mobility. This indicates that the choice of cut-
n (%) points for low muscle mass or low grip strength would
ASA score (SD) 2.3 (0.6) n = 124 2.7 (0.6) n = 77 < 0.001 not have changed our results. We found an association
Previous hip 5 (4) n = 124 9 (12) n = 77 0.039 between mobility pre-fracture and change in activities in
fracture, n (%) daily living, but this was an exploratory analysis and the
Charlson score (SD) 0.9 (1.3) n = 124 1.1 (1.3) n = 77 0.15 effect size was small.
Heart failure, n (%) 7 (6) n = 124 6 (8) n = 77 0.5
Sarcopenia was associated with an increased probability
of becoming a resident of a skilled nursing home (OR 3.2,
Previous myocardial 14 (11) n = 124 9 (12) n = 77 0.9
infarction, n (%) 95% CI 0.9 to 12.4, P = 0.048) and the combined endpoint
of becoming a resident of a nursing home or death (OR
Cerebrovascular 13 (10) n = 124 8 (10) n = 77 0.98
disease, n (%) 3.6, 95% CI 1.2 to 12.3, P = 0.02). This is a clinically relevant
Diabetes 9 (7) n = 124 10 (13) n = 77 0.2
finding but must be interpreted with caution, as it was an
mellitus, n (%) exploratory outcome and we were not able to correct for
Any solid 7 (6) n = 124 8 (10) n = 77 0.2 age, sex or BMI because of the low number of outcomes.
tumor, n (%) Among the participants who had sarcopenia status deter-
Pulmonary 15 (12) n = 124 18 (23) n = 77 0.036 mined, 6 participants died or became permanent residents
disease, n (%) of a nursing home among the not sarcopenic and 12 partic-
Length of hospital 6.8 (2.7) n = 124 9.6 (6.7) n = 77 < 0.001 ipants among those who were sarcopenic. The NMS was
stay, days (SD) chosen as our measure of physical performance because we
Body composition assumed that many participants would be unable to walk at
BMI, kg/m2 (SD) 25.6 (4.2) n = 107 22.1 (3.7) n = 70 < 0.001 inclusion. The NMS is extensively studied as a predictor of
ALM/height2, 6.3 (1.5) n = 111 4.4 (1.0) n = 77 < 0.001
mobility, morbidity, mortality and becoming a resident of a
kg/m2 (SD) nursing home [28–31]. We found a ceiling effect with the
Women 6.1 (1.3) n = 86 4.3 (0.8) n = 56 < 0.001 NMS, with 54% of participants scoring the maximum 9 be-
fore the fracture and 30% at one-year. Possibly because par-
Men 7.0 (1.7) n = 25 4.8 (1.2) n = 21 < 0.001
ticipants with a pre-fracture NMS of 0 or 1 were not
Grip strength
eligible for inclusion. Patients found the NMS easy to
Grip strength, 27.0 (10.3) n = 123 16.5 (6.4) n = 77 < 0.001 understand and scoring was straightforward. Surprisingly,
kg (SD)
we found that 8% of patients had better mobility at one-
Women 22.9 (6.9) n = 94 14.3 (5.0) n = 56 < 0.001 year compared to pre-fracture. For some of the patients this
Men 40.1 (8.3) n = 29 22.3 (5.9) n = 21 < 0.001 was due to illness that started before the fracture, and their
Mobility improvement in mobility after hip fracture was due to reso-
New Mobility 8.0 (1.5) n = 123 7.1 (2.0) n = 74 < 0.001 lution of their illness, rather than successful rehabilitation.
Score (SD) Use of rehabilitation services improves mobility after hip
Women 8.0 (1.6) n = 94 7.1 (2.0) n = 55 0.008 fracture [32, 33]. We did not record what rehabilitation ser-
Men 8.2 (1.4) n = 29 6.8 (2.2) n = 19 0.017 vices the participants received, and it is possible that re-
Baseline characteristics by sarcopenia status (means with standard deviations
habilitation could mediate the effect between sarcopenia
and counts with percentages). P-values for comparison of groups are by the and change in mobility.
Mann–Whitney–Wilcoxon test, except for type of fracture which is by The participants in our study were slightly younger (79.4
chi-squared test. Trochanteric fractures include basocervical femoral neck
fractures and subtrochanteric fractures. Previous hip fracture indicates a vs. 80.0 years) and had a lower mean ASA score (2.5 vs. 2.7)
previous hip fracture, either left or right hip. ALM: Appendicular lean mass, indicating better health compared to patients in the
ADL: Activities of daily living
Norwegian Hip Fracture Register. We did not include
Steihaug et al. BMC Geriatrics (2018) 18:65 Page 6 of 10

Fig. 2 What participants were assessed for muscle mass, grip strength, mobility and sarcopenia

patients from skilled nursing homes or with severe cognitive (95% CI 0.56–0.87) in men [35]. Using anthropometry to
impairment, and our results are not generalizable to those identify low muscle mass instead of DXA can lead to
populations. misclassification of muscle mass status and hence sarco-
Anthropometry is considered a less valid method for penia status. By using anthropometry to determine sar-
determining muscle mass compared to dual-energy X- copenia status we reduced our ability to detect an effect
ray absorptiometry (DXA) or computed tomography of sarcopenia on outcomes. We used anthropometry in
scan [34]. The EWGSOP recommends not using anthro- our study because it is in common use [36], inexpensive,
pometry to determine muscle mass in research but al- and more easily performed on patients with reduced
lows for it in clinical practice [4]. We have previously mobility and acute illness, compared to DXA [37]. Some
investigated how anthropometry compares to DXA in consider objectively measured physical performance su-
identifying low muscle mass and found an area under perior to self-reported mobility, such as the NMS, but
the curve of 0.64 (95% CI 0.54–0.75) in women and 0.72 when the two types of measurement are compared in

Table 2 Outcomes after one year by sarcopenia status

Marginal values Regression (sarcopenic – not sarcopenic)
Not sarcopenic Sarcopenic Unadjusted Adjusted
Mean (SD) No. Mean (SD) No. Estimate 95% CI P Estimate 95% CI P
Change NMS −1.2 (1.8) n = 117 −1.3 (1.9) n = 67 0.0 −0.6 to 0.6 0.9 0.2 −0.5 to 0.9 0.6
Change B-ADL −0.8 (2.4) n = 76 −2.2 (3.9) n = 52 −0.4 −1.2 to 0.3 0.3 −0.3 − 1.1 to 0.6 0.6

Not Sarcopenic Sarcopenic OR 95% CI P

Fracture, n (%) N 7 (6) n = 120 8 (11) n = 71 2.0 (0.6 to 7.0) 0.3
Hip fractures, n (%) N 3 (2) n = 124 3 (4) n = 77 1.6 (0.2 to 12.5) 0.7
Reoperations, n (%) N 7 (6) n = 124 1 (1) n = 77 0.2 (0.0 to 1.8) 0.2
Hospitalization, n (%) N 41 (33) n = 123 24 (33) n = 73 1.0 (0.5 to 1.9) 1.0
Nursing home, n (%) N 5 (4) n = 124 9 (12) n = 77 3.2 (0.9 to 12.4) 0.048
Death, n (%) N 3 (2) n = 124 5 (6) n = 77 2.8 (0.5 to 18.5) 0.3
Death or nursing home, n (%) N 6 (5) n = 124 12 (16) n = 77 3.6 (1.2 to 12.3) 0.02
Outcomes after one year by sarcopenia status (sarcopenic – not sarcopenic). Regression analysis for change in mobility and Barthel ADL from pre-fracture until
one year adjusted for age, sex and BMI with imputation of missing values. ADL: Barthel Activities of daily living. Analysis for fracture, hip fracture, reoperations,
all-cause hospitalization, nursing home, death or nursing home or death by two-sided Fisher’s exact test using available cases. NMS: New mobility score. B-ADL:
Barthel activities of daily living. OR: Odds ratio
Steihaug et al. BMC Geriatrics (2018) 18:65 Page 7 of 10

Fig. 3 New Mobility Score (NMS) during hospitalization, at three months, and at one year, stratified by sarcopenia status during hospitalization.
The horizontal lines show mean NMS scores

Fig. 4 New Mobility Score (NMS) pre-fracture and at one-year follow-up. The first number in each cell is the number of patients with the given
combination of NMS scores. For each row, the percentage values and the cell shadings show the distribution of NMS at follow-up for a given
NMS score at baseline. No patients had a NMS of 1 at baseline, and patients with a NMS score of 0 was excluded from the study. Patients with
the same NMS score at baseline and follow-up are shown in boldface, and any cell to the right of this diagonal indicates an improvement in
the NMS
Steihaug et al. BMC Geriatrics (2018) 18:65 Page 8 of 10

Table 3 Outcomes after one year predicted by muscle mass, grip strength or mobility
Unadjusted Adjusted
β (95% Confidence interval) β (95% Confidence interval)
Change in New Mobility Score at one year
ALM/height2, kg/m2 0.0 (−0.1 to 0.2) p = 0.7 n = 175 0.2 (−0.1 to 0.4) p = 0.2 n = 155
Grip strength, kg 0.0 (−0.0 to 0.0) p = 0.7 n = 193 −0.0 (−0.0 to 0.0) p = 0.7 n = 169
Change in Barthel activities of daily living at one year
ALM/height2, kg/m2 0.0 (−0.2 to 0.3) p = 0.8 n = 121 0.1 (−0.2 to 0.4) p = 0.4 n = 115
Grip strength, kg 0.0 (−0.0 to 0.1) p = 0.07 n = 137 0.0 (0.0 to 0.1) p = 0.1 n = 128
New Mobility Score, point 0.2 (0.0 to 0.4) p = 0.03 n = 148 0.2 (0.0 to 0.4) p = 0.03 n = 130
Death or nursing home at one year
ALM/height2, kg/m2 0.8 (0.6 to 1.2) p = 0.3 n = 194 1.0 (0.6 to 1.7) p = 1.0 n = 170
Grip strength, kg 0.9 (0.9 to 1.0) p = 0.002 n = 222 0.9 (0.9 to 1.0) p = 0.1 n = 186
New Mobility Score, point 0.7 (0.6 to 0.9) p < 0.001 n = 243 0.8 (0.6 to 1.0) p = 0.06 n = 194
Outcomes after one year by muscle mass, grip strength or mobility. Analysis of change in mobility and Barthel activities of daily living by regression and with
imputation of missing values. ALM/height2, grip strength and New Mobility Score are continuous, independent variables. Change in New Mobility Score and
Barthel activities of daily living are continuous dependent variables. n: number of cases without missing values. OR: Odds ratio, ALM: Appendicular lean mass.
Adjusted analysis with age, sex and BMI as covariates

hip fracture patients they have been found to be equally Abbreviations

predictive of outcomes [38]. Future research on sarcope- ALM: Appendicular lean mass; ASA: ASA Physical Status Classification System
score.; B-ADL: Barthel activities of daily living; BMI: Body mass index;
nia in hip fracture patients could explore other methods CI: Confidence interval; DXA: Dual-energy X-ray absorptiometry;
for determining sarcopenia, such as computed tomog- EWGSOP: European Working Group on Sarcopenia in Older Persons;
raphy to directly measure intramuscular adipose tissue NMS: New mobility score; OR: Odds ratio; SD: Standard deviation
[34] or using objective measures of physical performance Acknowledgements
such as the Short Physical Performance Battery [39]. A We thank research nurses Britt Pedersen, Cathrine Sandnes, Mette Irene
randomized controlled study of an intervention targeting Martinsen, Christine Ekrheim and Elin Engh, and physical therapist Sylvia
Sunde for assistance with data collection. We are grateful to associate
sarcopenia status to improve mobility after hip fracture professor Mala Naik for proof-reading the manuscript.
would provide additional insight on the causal relation
between sarcopenia and mobility. Funding
The Western Norway Regional Health Authority, Stavanger, Norway and
The included patients were a minority of all patients Haraldsplass Deaconess Hospital, Bergen, Norway supported the study. The
operated on for hip fracture during the period of financial sponsors had no role in the design and execution of the study,
inclusion. We included postoperative patients who were analysis and interpretation of data, or writing of the manuscript.
frequently bed-bound, receiving opiates for pain relief, Availability of data and materials
with indwelling urinary catheters and while receiving The datasets generated and analyzed during the current study are available
intravenous fluid therapy. We believe there were three from the corresponding author on reasonable request.
main reasons for the low recruitment rate: patients did Authors’ contributions
not fulfill the inclusion criteria, patients were discharged OMS, AHR, BB and CGG designed the study and wrote the study protocol. All
before the consent process could be completed, and authors have contributed to the manuscript, agree on submitting it for
publication, and vouch for the integrity of the data and analysis. OMS and MHK
participants declined to participate because it was too have been responsible for including patients and data collection. OMS
much of a burden. For the patients who did consent to performed all data analyses except the regression analyses that used multiple
participate we found that determining sarcopenia by imputation. AHR has been the project leader. KOH performed the regression
analyses that used multiple imputation and prepared all the figures. All authors
anthropometry, grip strength and the NMS was feasible. read and approved the final manuscript.
The greatest difficulty was in determining the height of
the participants. Ethics approval and consent to participate
The study was approved by the research committees at the participating
hospitals (Haraldsplass Deaconess Hospital and Haukeland University
Conclusion Hospital, Bergen, and Diakonhjemmet Hospital, Oslo, Norway), the Regional
Sarcopenia status determined in postoperative hip frac- Committee on Medical and Health Research Ethics (case 2011/1322/REK
sør-øst B) and was conducted according to the principles of the Declaration
ture patients by anthropometry, grip strength and self- of Helsinki. Participation was by written informed consent with participants
reported mobility did not predict change in mobility in receiving oral and written information on the first day and signed the
the year after hip fracture. Sarcopenia was associated consent form on a subsequent day.
with having lower mobility at one year and a greater risk Consent for publication
of becoming a resident of a nursing home or death. Not applicable.
Steihaug et al. BMC Geriatrics (2018) 18:65 Page 9 of 10

Competing interests 15. Wehren LE, Hawkes WG, Hebel JR, Orwig DL, Magaziner J. Bone mineral
The authors declare that they have no competing interests. density, soft tissue body composition, strength, and functioning after hip
fracture. J Gerontol A Biol Sci Med Sci. 2005;60(1):80–4.
16. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying
Publisher’s Note prognostic comorbidity in longitudinal studies: development and validation.
Springer Nature remains neutral with regard to jurisdictional claims in J Chronic Dis. 1987;40(5):373–83.
published maps and institutional affiliations. 17. Mahoney FI, Barthel DW. Functional evaluation: the Barthel index. Md State
Med J. 1965;14:61–5.
Author details
1 18. Gjertsen JE, Dybvik E, Furnes O, Fevang JM, Havelin LI, Matre K, Engesaeter
Kavli Research Centre for Geriatrics and Dementia, Haraldsplass Deaconess
LB. Improved outcome after hip fracture surgery in Norway. Acta Orthop.
Hospital, Bergen, Norway. 2Department of Clinical Science, University of
2017;88(5):505–11.
Bergen, Bergen, Norway. 3Department of Rheumatology, Haukeland
19. Havelin L.I. FO, Engesæter, L. B.; Fenstad, A. M.; Bartz-Johannessen, C; Dybvik,
University Hospital, Bergen, Norway. 4Western Norway University of Applied
E; Fjeldsgaard, K.; Gundersen, T.. Norwegian National Advisory Unit on
Sciences, Bergen, Norway. 5Department of Orthopedics, Haukeland University
Arthroplasty and Hip Fractures. Annual Report. 2016. In. Bergen, Norway:
Hospital, Bergen, Norway. 6Department of Rheumatology, Diakonhjemmet
Helse Bergen HF, Ortopedisk klinikk, Haukeland universitetssjukehus; 2016:
Hospital, Oslo, Norway. 7Centre for Clinical Research, Haukeland University
305.
Hospital, Bergen, Norway. 8Department of Clinical Science, the Faculty of
20. Heymsfield SB, McManus C, Smith J, Stevens V, Nixon DW. Anthropometric
Medicine and Dentistry, University of Bergen, Postbox 7804, n-5020 Bergen,
measurement of muscle mass: revised equations for calculating bone-free
Norway.
arm muscle area. Am J Clin Nutr. 1982;36(4):680–90.
21. Kim J, Heshka S, Gallagher D, Kotler DP, Mayer L, Albu J, Shen W, Freda PU,
Received: 8 May 2017 Accepted: 27 February 2018
Heymsfield SB. Intermuscular adipose tissue-free skeletal muscle mass:
estimation by dual-energy X-ray absorptiometry in adults. J Appl Physiol
(1985). 2004;97(2):655–60.
References
22. Kristensen MT, Kehlet H. Most patients regain prefracture basic mobility
1. Dyer SM, Crotty M, Fairhall N, Magaziner J, Beaupre LA, Cameron ID,
after hip fracture surgery in a fast-track programme. Dan Med J. 2012;59(6):
Sherrington C. Fragility fracture network rehabilitation research special
A4447.
interest G: a critical review of the long-term disability outcomes following
23. Parker MJ, Palmer CR. A new mobility score for predicting mortality after hip
hip fracture. BMC Geriatr. 2016;16:158.
fracture. J Bone Joint Surg Br. 1993;75(5):797–8.
2. Dawson-Hughes B, Bischoff-Ferrari H. Considerations concerning the
definition of sarcopenia. Osteoporos Int. 2016;27(11):3139–44. 24. Magaziner J, Simonsick EM, Kashner TM, Hebel JR, Kenzora JE. Predictors of
3. Dodds RM, Sayer AA. Sarcopenia, frailty and mortality: the evidence is functional recovery one year following hospital discharge for hip fracture: a
growing. Age Ageing. 2016;45(5):570–1. prospective study. J Gerontol. 1990;45(3):M101–7.
4. Cruz-Jentoft AJ, Baeyens JP, Bauer JM, Boirie Y, Cederholm T, Landi F, Martin 25. Studenski SA, Peters KW, Alley DE, Cawthon PM, McLean RR, Harris TB,
FC, Michel JP, Rolland Y, Schneider SM, et al. Sarcopenia: European Ferrucci L, Guralnik JM, Fragala MS, Kenny AM, et al. The FNIH sarcopenia
consensus on definition and diagnosis: report of the European working project: rationale, study description, conference recommendations, and final
group on sarcopenia in older people. Age Ageing. 2010;39(4):412–23. estimates. J Gerontol A Biol Sci Med Sci. 2014;69(5):547–58.
5. Perez-Zepeda MU, Sgaravatti A, Dent E. Sarcopenia and post-hospital 26. Harrell Jr FE: rms: Regression Modeling Strategies. R package version 4.0–0.
outcomes in older adults: a longitudinal study. Arch Gerontol Geriatr. 2017; City 2013.
69:105–9. 27. R Core Team: R: A language and environment for statistical computing. In.
6. Ho AW, Lee MM, Chan EW, Ng HM, Lee CW, Ng WS, Wong SH. Prevalence Vienna, Austria: R Foundation for Statistical Computing; 2016.
of pre-sarcopenia and sarcopenia in Hong Kong Chinese geriatric patients 28. Kristensen MT, Foss NB, Kehlet H. Timed up and go and new mobility score
with hip fracture and its correlation with different factors. Hong Kong Med as predictors of function six months after hip fracture. Ugeskr Laeger. 2005;
J. 2016;22(1):23–9. 167(35):3297–300.
7. Gonzalez-Montalvo JI, Alarcon T, Gotor P, Queipo R, Velasco R, Hoyos R, 29. Keene GS, Parker MJ, Pryor GA. Mortality and morbidity after hip-fractures. Br
Pardo A, Otero A. Prevalence of sarcopenia in acute hip fracture patients Med J. 1993;307(6914):1248–50.
and its influence on short-term clinical outcome. Geriatr Gerontol Int. 2016; 30. Parker MJ, Palmer CR. Prediction of rehabilitation after hip fracture. Age
16(9):1021–7. Ageing. 1995;24(2):96–8.
8. Di Monaco M, Castiglioni C, De Toma E, Gardin L, Giordano S, Di Monaco R, 31. Foss NB, Kristensen MT, Kehlet H. Prediction of postoperative morbidity,
Tappero R. Presarcopenia and sarcopenia in hip-fracture women: prevalence mortality and rehabilitation in hip fracture patients: the cumulated
and association with ability to function in activities of daily living. Aging ambulation score. Clin Rehabil. 2006;20(8):701–8.
Clin Exp Res. 2015;27(4):465–72. 32. Sylliaas H, Brovold T, Wyller TB, Bergland A. Prolonged strength training in
9. Malafarina V, Uriz-Otano F, Malafarina C, Martinez JA, Zulet MA. Effectiveness older patients after hip fracture: a randomised controlled trial. Age Ageing.
of nutritional supplementation on sarcopenia and recovery in hip fracture 2012;41(2):206–12.
patients. A multi-Centre randomized trial. Maturitas. 2017;101:42–50. 33. Handoll HH, Sherrington C, Mak JC. Interventions for improving mobility
10. Landi F, Calvani R, Ortolani E, Salini S, Martone AM, Santoro L, Santoliquido after hip fracture surgery in adults. Cochrane Database Syst Rev. 2011;3:
A, Sisto A, Picca A, Marzetti E. The association between sarcopenia and CD001704.
functional outcomes among older patients with hip fracture undergoing in- 34. Cesari M, Fielding RA, Pahor M, Goodpaster B, Hellerstein M, van Kan GA,
hospital rehabilitation. Osteoporos Int. 2017:1–8. Anker SD, Rutkove S, Vrijbloed JW, Isaac M, et al. Biomarkers of sarcopenia
11. Kristensen MT. Factors affecting functional prognosis of patients with hip in clinical trials-recommendations from the international working group on
fracture. Eur J Phys Rehabil Med. 2011;47(2):257–64. sarcopenia. J Cachexia Sarcopenia Muscle. 2012;3(3):181–90.
12. Pajulammi HM, Pihlajamaki HK, Luukkaala TH, Nuotio MS. Pre- and 35. Steihaug OM, Gjesdal CG, Bogen B, Ranhoff AH. Identifying low muscle
perioperative predictors of changes in mobility and living arrangements mass in patients with hip fracture: validation of bioelectrical impedance
after hip fracture–a population-based study. Arch Gerontol Geriatr. 2015; analysis and anthropometry compared to dual energy X-ray absorptiometry.
61(2):182–9. J Nutr Health Aging. 2016;20(7):685–90.
13. Savino E, Martini E, Lauretani F, Pioli G, Zagatti AM, Frondini C, Pellicciotti F, 36. Schaap L, Fox B, Henwood T, Bruyère O, Reginster J-Y, Beaudart C, Buckinx
Giordano A, Ferrari A, Nardelli A, et al. Handgrip strength predicts persistent F, Roberts H, Cooper C, Cherubini A. Grip strength measurement: towards a
walking recovery after hip fracture surgery. Am J Med. 2013;126(12):1068– standardized approach in sarcopenia research and practice. Eur Geriatr Med.
75. e1061 2016;7(3):247–55.
14. Visser M, Harris TB, Fox KM, Hawkes W, Hebel JR, Yahiro JY, Michael R, 37. Beaudart C, McCloskey E, Bruyere O, Cesari M, Rolland Y, Rizzoli R, Araujo de
Zimmerman SI, Magaziner J. Change in muscle mass and muscle strength Carvalho I, Amuthavalli Thiyagarajan J, Bautmans I, Bertiere MC, et al.
after a hip fracture: relationship to mobility recovery. J Gerontol A Biol Sci Sarcopenia in daily practice: assessment and management. BMC Geriatr.
Med Sci. 2000;55(8):M434–40. 2016;16(1):170.
Steihaug et al. BMC Geriatrics (2018) 18:65 Page 10 of 10

38. Latham NK, Mehta V, Nguyen AM, Jette AM, Olarsch S, Papanicolaou D,
Chandler J. Performance-based or self-report measures of physical function:
which should be used in clinical trials of hip fracture patients? Arch Phys
Med Rehabil. 2008;89(11):2146–55.
39. Vellas B, Pahor M, Manini T, Rooks D, Guralnik JM, Morley J, Studenski S,
Evans W, Asbrand C, Fariello R, et al. Designing pharmaceutical trials for
sarcopenia in frail older adults: EU/US task force recommendations. J Nutr
Health Aging. 2013;17(7):612–8.

Submit your next manuscript to BioMed Central

and we will help you at every step:
• We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research

Submit your manuscript at

www.biomedcentral.com/submit