Site-specific bone mineral density by age strata in older women with proximal femur fractures

Abstract

[Purpose] Limited research has been conducted comparing bone mineral density by anatomical site and age strata in older women following proximal femur fractures. Therefore, in this study, we examined the site-specific bone mineral density and age strata in a population of older women with proximal femur factures. [Participants and Methods] We categorized 68 older women with postoperative proximal femur fractures into three groups based on age: 65–79 years (Group I), 80–89 years (Group II), and ≥90 years (Group III). The bone mineral density of four sites (femoral neck, trochanter, intertrochanteric region, and Ward’s triangle) was measured in the nonfractured femur using dual-energy x-ray absorptiometry. [Results] The young adult mean of the total proximal femur was significantly lower in Groups II and III than that in Group I by 13.8% and 20.8%, respectively. The young adult mean of the change rate of Ward’s triangle was the highest in the four sites and significantly lower in Groups II and III than that in Group I by 24.3% and 31.6%, respectively. [Conclusion] The BMD decline in older women might be more pronounced in the anatomical regions with a higher trabecular bone proportions.

INTRODUCTION

In postmenopausal women, bone resorption is accompanied by decreased estrogen levels and a progressively increasing fracture risk^{1, 2)}. The disruption of bone metabolism due to menopause is pronounced starting at 50 years of age³⁾. Women aged 60–69 years lose approximately 0.6% of their bone mass per year, those 70–79 lose approximately 1.1% per year, and those 80 years and over lose approximately 2.1% per year⁴⁾. Furthermore, decreased bone mineral density (BMD) in older individuals with proximal femur fractures can be accelerated by decreased physical activity and poor nutritional intake^{5, 6)}.

The decline in femoral BMD varies by anatomical region and primarily occurs in the trabecular bone, which has a high metabolic turnover rate¹⁾. The proportion of trabecular bone in the femur is approximately 25% in the neck and 50% in the trochanter and intertrochanteric region⁷⁾. Ravn et al. reported that the decrease in BMD was up to 20% in the femoral neck, trochanter, and intertrochanteric region within 20 years of menopause⁸⁾. Other studies have reported that Ward’s triangle, which contains a high proportion of trabecular bone, experiences a lifetime BMD loss of up to 32.68%⁹⁾, reaching this level within 20 years of menopause⁸⁾. Few studies have comprehensively compared BMD by anatomical site and age strata in older women following proximal femur fractures. As older women retain a high risk of a second fracture^{5, 6)}, clarifying age-related differences in femoral BMD by anatomical site will facilitate the optimization of secondary fracture prevention strategies tailored to specific age groups.

Therefore, in this study, we aimed to examine site-specific BMD by age stratum in older women with proximal femur fractures. The findings will contribute to advances in preventive medicine for osteoporotic fractures.

PARTICIPANTS AND METHODS

We evaluated 64 older women admitted to Matsuda Hospital in Sendai city between April 2023 and October 2025 following femoral or trochanteric fractures. The exclusion criteria included fractures caused by mechanisms other than falls injuries resulting from traffic accidents and missing results. Rehabilitation of the participants began after surgery and was conducted three times daily for 60 minutes per session. The length of hospital stay was up to 150 days postoperatively. The participants were prescribed physical and occupational therapies, and conducted joint range-of-motion and muscle-strengthening exercises, gait training, and activities of daily living training. The participants were classified into three groups based on age: 65–79 years (Group I), 80–89 years (Group II), and ≥90 years (Group III). Study approval was granted by the University Research Ethics Committee (Approval No. 25043).

We examined patient characteristics including age, height, weight, fracture type (femoral neck fracture or femoral trochanteric fracture), and surgical procedure (internal fixation or hemiarthroplasty). Body mass index (BMI) was calculated as weight divided by height squared (kg/m²).

The Charlson Comorbidity Index (CCI) was used to assess comorbidities. The CCI assesses the risk of short-term mortality based on 19 comorbidities and age adjustment; a higher total score indicates a higher risk of mortality¹⁰⁾. Each comorbidity was assigned a score of 1–6, based on its associated mortality risk. For patients aged 50 years or older, 1–4 points were added to the total score¹⁰⁾.

We screened the participants for dementia using the Revised Hasegawa Dementia Scale (HDS-R)¹¹⁾. The HDS-R consists of nine items: age (1 point), orientation (6 points), memory (9 points), calculation (2 points), back recitation (2 points), object recall (5 points), and word recall (5 points). A score of 20 or less was considered indicative of a high likelihood of dementia.

We measured the BMD using dual-energy x-ray absorptiometry (Horizon Ci, Hologic Inc., Marlborough, MA, USA). Prior to measurement, a clinical laboratory technologist instructed the patient to change to a hospital gown and remove any metallic objects. The patients were positioned in supine with the hips in internal rotation. The measurements were completed in approximately 10 minutes. The measurement sites were the femoral neck, trochanter, intertrochanteric region, and Ward’s triangle in the nonfractured femur. The femoral BMD was measured postoperatively in all patients and the young adult mean (YAM) was calculated.

We conducted all statistical analyses using R Commander. One-way analysis of variance (ANOVA) was used for between-group comparisons, and calculated eta-squared (η²) as a measure of effect size. For significant differences, we performed the Scheffé’s test as a post-hoc test. The χ² test was used to assess the associations with differences in fracture type. Statistical significance was set at p<0.05.

RESULTS

The 64 participants were categorized into groups based on age: Group I consisted of 22 participants; Group II, 25; and Group III, 17 (Table 1). Height was significantly lower in Group III (3.6%) than that in Group I (p<0.05). Body weight was significantly lower in Group III than that in Group I by 15.4% (p<0.05). The CCI was significantly higher in Groups II and III than that in Group I by 32.5% (p<0.01) and 27.5% (p<0.05), respectively. However, the groups showed no significant differences in BMI. The χ² test revealed that the associations between age strata and fracture type were significantly skewed (Table 2).

Table 1. Patient characteristics.

	I (n=22)	II (n=25)	III (n=17)	p-value1)
Height (cm)	153.6 ± 6.21*b	151.4 ± 4.91	148.1 ± 5.61*b	<0.05
Weight (kg)	52.9 ± 10.11*b	46.8 ± 7.63	44.7 ± 7.62*b	<0.01
BMI (kg/m2)	22.5 ± 3.92	20.4 ± 3.38	20.3 ± 2.90	0.087
CCI	4.0 ± 0.98**a*b	5.3 ± 1.40**a	5.1 ± 0.99*b	<0.001
HDS-R	25.6 ± 5.57*b	20.0 ± 8.99	18.2 ± 9.01*b	<0.05

Mean ± SD, *p<0.05, **p<0.01. ^a: I vs. II, ^b: I vs. III. ¹⁾: One way analysis of variance. BMI: body mass index; CCI: Charlson Comorbidity Index; HDS-R: Revised-Hasegawa Dementia Scale; SD: standard deviation.

Table 2. Association between fracture type and age strata.

	I (n=22)	II (n=25)	III (n=17)	χ2 (df)	p-value
Neck (%)	81.8%	76.0%	23.5%	16.7 (2)	<0.001
Trochanter (%)	18.2%	24.0%	76.5%

df: degrees of freedom.

Table 3 shows the results of the comparison of the percentage of the YAM between the groups. The total proximal femur percentage of the YAM was significantly lower in Groups II and III than that in Group I by 13.8% (p<0.05) and 20.8% (p<0.01), respectively. The femoral neck percentage was significantly lower in Groups II and III than that in Group I by 14.2% and 19.3%, respectively (all p<0.05). The femoral trochanter percentage was significantly lower in Groups II and III than that in Group I by 13.0% (p<0.05) and 22.3% (p<0.001), respectively. The femoral intertrochanteric region percentage was lower in Groups II and III than that in Group I by 15.1% (p<0.05) and 21.6% (p<0.01), respectively. The Ward’s triangle was significantly lower in Groups II and III than that in Group I by 24.3% and 31.3%, respectively (all p<0.05) (Table 3).

Table 3. The percentage of young adult mean (%).

	I (n=22)	II (n=25)	III (n=17)	p-value1)	η2
Total	72.5 ± 14.10*a, **b	62.5 ± 11.06*a	57.4 ± 12.00**b	<0.01	0.20
Neck	66.7 ± 14.50*ab	57.2 ± 9.68*a	53.8 ± 14.24*b	<0.01	0.16
Trochanter	73.6 ± 12.79*a, ***b	64.0 ± 11.05*a	57.2 ± 11.51***b	<0.001	0.24
Intertrochanteric	72.1 ± 13.41*a, **b	61.2 ± 12.61*a	56.5 ± 12.60**b	<0.001	0.20
Ward’s triangle	46.0 ± 16.87*ab	34.8 ± 15.29*a	31.6 ± 9.81*b	<0.01	0.15

Mean ± SD, *p<0.05, **p<0.01, ***p<0.001. ^a: I vs. II, ^b: I vs. III. ¹⁾: One way analysis of variance SD: standard deviation; η²: eta-squared.

DISCUSSION

The BMD decline in women becomes apparent at age 50 and accelerates after age 65^{2, 12)}. In Japan, the YAM is used as an indicator of osteoporosis; a YAM of less than 70% is consider as a state of high fracture risk¹³⁾. In our study, the total YAM values were 72.5%, 62.4%, and 57.4% in Groups I, II, and III, respectively. Notably, Groups II and III did not reach 70%. These findings suggest that the femoral BMD declines with age, even in female patients following a proximal femoral fracture. This decline is particularly pronounced in older patients. In addition, individuals aged 80 years or older commonly have comorbidities, such as diabetes and renal failure¹⁴⁾, making them prone to a deteriorating general condition. We found that the rate of change in the CCI was significantly higher in Groups II and III than that in Group I. Following a proximal femoral fracture, delayed mobilization¹⁵⁾ may lead to reduced physical activity and weight loss, which could accelerate the progression of BMD decline⁵⁾. Thus, our results suggest that comorbidities contributed to the deterioration of the general condition of the older patients, thereby delaying early postoperative mobilization. These findings emphasize the importance of promoting safe and early mobilization while ensuring an environment that is conducive to fall prevention.

The proportion of trabecular bone in the femur varies by anatomical region, and age-related BMD decline exhibits a region-specific pattern⁸⁾. In particular, Ward’s triangle displays the highest proportion of trabecular bone and is highly susceptible to the effects of age-related decline in bone metabolism¹⁵⁾. Our results showed that the magnitude of the BMD decline in Ward’s triangle was 30%, the decline in the other analyzed regions was approximately 20%. Furthermore, the rate of BMD decline in the neck which has the lowest proportion of trabecular bone was lower than that in the trochanter and intertrochanteric regions. These findings suggest that in very older women, BMD decline may be more pronounced in regions with a higher proportion of trabecular bone^{5, 6, 8)}.

The following limitations must be considered when interpreting the results of this study¹⁶⁾. However, lifestyle factors (e.g., smoking history, alcohol consumption, nutritional status, and physical activity) are associated with declined BMD. We did not analyze these factors and did not directly assess the risk of secondary fractures in our study. Future studies should account for the effects of comorbidities and lifestyle habits on femoral BMD and comprehensively evaluate. The results of our study highlight the importance of understanding anatomical region-specific characteristics of femoral BMD when considering exercise therapy and lifestyle guidance for individuals of advanced ages. Safety considerations and fall prevention must be emphasized in rehabilitation and/or wellness clinics.

Funding and Conflict of interest

None.