Prevalence of discordance between femoral and lumbar bone mineral density among older adults in a community-based setting

Abstract

BACKGROUND:

T-score measurement via dual-energy X-ray absorptiometry (DXA) is the gold standard for assessing and classifying the bone mineral density status of patients as normal, osteopenic, or osteoporotic according to the World Health Organization criteria. However, the diagnostic accuracy may be affected by the skeletal site selected for DXA.

OBJECTIVES:

Estimate the prevalence of femoral and lumbar BMD discordance in a community-based setting in Riyadh, Saudi Arabia.

DESIGN:

Cross-sectional

SETTING:

Polyclinics at a tertiary care center

PATIENTS AND METHODS:

This study included all patients aged ≥60 years who visited the Department of Family Medicine and underwent DXA screening between 2016 and 2022.

MAIN OUTCOME MEASURES:

Discordance was defined as a difference in BMD status between two skeletal sites. Minor discordance occurs when adjacent sites have different diagnoses; i.e., one site exhibits osteoporosis and the other exhibits osteopenia. In contrast, major discordance occurs when one site exhibits osteoporosis and the other exhibits normal BMD.

SAMPLE SIZE:

1429 older adults

RESULTS:

The study patients had a median age of 66 years (60–99, minimum-maximum). The prevalence of discordance was 41.6%, with major discordance present in 2.2% of patients and minor discordance in 39.4%. The distribution of discordance did not differ significantly among the sociodemographic factors.

CONCLUSION:

Discordance is prevalent among the Saudi geriatric population. During the analysis of DXA results, physicians should account for discordance when diagnosing and ruling out osteoporosis in high-risk patients.

LIMITATIONS:

All factors influencing discordance were not explored thoroughly; this study mainly focused on older adults. Furthermore, diverse age groups need to be investigated for a more comprehensive understanding of the analyzed factors.

INTRODUCTION

Osteoporosis is a disease characterized by low bone mineral density (BMD), deterioration of bone tissue, and disruption of bone microarchitecture; it can lead to bone loss and increase the risk of fracture.¹ Factors associated with an increased risk of osteoporosis-related fractures include age, sex, steroid intake, vitamin deficiency, and use of glucocorticoid medication.²

BMD is measured using dual-energy X-ray absorptiometry (DXA) and is expressed as grams of minerals (mainly calcium) per area of scanned bone (g/cm²). Femoral and lumbar BMD measurements are used to confirm the diagnosis of osteoporosis and monitor patients receiving treatment. On DXA, the T-score is calculated by dividing an individual's BMD by the standard deviation of BMD of the population.³ DXA results can help classify an individual's BMD status as normal, osteopenic, or osteoporotic according to the T-score.⁴ Although the hip and spine are expected to have consistent T-scores, the BMD status of the two sites often differ.⁵ A discrepancy between T-scores at different skeletal sites in the same patient is known as discordance.⁶

In the diagnosis of osteoporosis, discordance in T-scores between two skeletal sites in the same patient may reflect different conditions (osteoporosis, osteopenia, and normal).⁷ Various studies have analyzed the prevalence and effects of different T-scores on the management of osteoporosis.^7–12 According to a previous study on discordance, clinicians should expect differences in the spine and hip T-scores in at least 40% of the patients examined via DXA. T-score differences can result from various physiological and pathological reasons related to patient factors as well as performance or DXA analysis.¹³ The patients in the abovementioned studies were either referred to the department or visited for osteoporosis evaluations and may differ from the general population in some aspects, such as socioeconomic status, education levels, or prevalence of conditions associated with osteoporosis (smoking and alcohol consumption, vitamin D deficiency, or long-term corticosteroid use).¹⁴

In the context of the Saudi population, obesity has emerged as a potential factor influencing the rate of discordance between femoral and lumbar BMD measurements. The mechanism by which obesity may affect BMD and T-score discordance remains unexplored, but it may be related to biomechanical forces on weight-bearing bones, differential rates of bone turnover, and the presence of technical artifacts on imaging, particularly in lumbar BMD measurements.^14,15

An understanding of the prevalence of and factors contributing to T-score discordance is essential for the accurate assessment and management of osteoporosis. This study aimed to delineate the extent of femoral–lumbar BMD discordance in a community-based setting in Riyadh, Saudi Arabia. By determining the prevalence of and factors underlying discordance, we are providing insights that will help develop future clinical guidelines and therapeutic strategies, ensuring accurate interpretation of DXA results to optimize patient outcomes. Our findings would reinforce not only the importance of a comprehensive understanding of BMD measurements but also the need for heightened clinical vigilance in interpreting DXA results, particularly in populations with a high prevalence of discordance, such as patients with obesity or those with significant risk factors for osteoporosis.

PATIENTS AND METHODS

This cross-sectional study included patients who visited the Department of Family Medicine and various polyclinics affiliated with the King Faisal Specialist Hospital & Research Centre, Riyadh, Saudi Arabia. The data were collected between January 1, 2016 and December 31, 2021. This research was approved by the Research Ethics Committee of King Faisal Specialist Hospital & Research Centre on January 15, 2023 (RAC 2231174). The patients’ personal information was kept strictly confidential to ensure their privacy throughout the study.

Individuals aged ≥60 years who had undergone osteoporosis screening via DXA were included in this study. Patients with secondary causes of osteoporosis, such as liver disease, kidney disease, and malabsorption syndromes, were excluded. Moreover, patients whose medical records contained incomplete DXA reports and missing information were excluded.

DXA, which was used to evaluate the patients, was performed mainly at two sites: the lumbar spine and femoral neck. According to the World Health Organization, a T-score of ≤2.5 standard deviation (SD) at any skeletal site indicates osteoporosis, a T-score between −1 and −2.5 SD indicates osteopenia, and a T-score above −1 is considered normal.³ The extent of discordance was classified as major or minor. Minor discordance occurs when adjacent sites have different diagnoses; i.e., one site exhibits osteoporosis and the other exhibits osteopenia or one site exhibits osteopenia and the other has normal BMD. In contrast, major discordance occurs when one site exhibits osteoporosis and the other has a normal BMD.⁸

Statistical analysis

All statistical analyses were performed using IBM SPSS version 23.0 (Armonk, NY, USA: IBM Corp). The Kolmogorov–Smirnov test was used to evaluate normality of age and body mass index (BMI). Descriptive statistics for continuous variables are reported as median (interquartile range [IQR]), and range whereas those for categorical variables are summarized as frequencies and percentages. Inferential statistics were performed for continuous variables using the nonparametric Kruskal-Wallis test to assess the presence of any statistically significant differences between the discordance categories. To compare categorical variables, we used the chi-square test or the Fisher exact test, as appropriate. The significance level was set at .05, and 95% confidence intervals were calculated for all statistically significant findings.

RESULTS

We reviewed the data of 1429 patients who had a median (IQR) age of 66 (10), range 60–99 years. The majority of patients (74.6%) were females; 53% of patients were obese, 85.7% were married, and 98.7% were nonsmokers (Table 1). Discordance was observed in 595 patients (41.6%): minor discordance in 563 patients (39.4%) and major discordance in 32 patients (2.2%) (Table 2).

Table 1.

Sociodemographic factors (N=1429).

Variables
Age (years)	66 (63–73)
Gender
Female	1066 (74.6)
Male	363 (25.4)
Marital statusa
Single	39 (2.7)
Married	1221 (85.7)
Divorced	46 (3.2)
Widow	118 (8.3)
Current smokingb
No	1338 (98.7)
Yes	18 (1.3)
BMc
Normal weight	201 (14.2)
Underweignt	8 (0.6)
Overweignt	459 (32.3)
Obese	752 (53)

Data are median (25th–75th percentile) for age number (percentage) for categorical data.

^a5 missing cases (0.3%);

^b73 missing cases (5.1%);

^c9 missing cases (0.6%).

Table 2.

Prevalence of discordance and its types.

Discordance type
Minor	563 (39.4)
Normal (lumbar spine) - osteopenia (femoral neck)	243 (17)
Osteopenia (lumbar spine) - normal (femoral neck)	167 (11.7)
Osteopenia (lumbar spine) - osteoporosis (femoral neck)	50 (3.5)
Osteoporosis (lumbar spine) osteopenia (femoral neck)	103 (7.2)
Major	32 (2.2)
Normal (lumbar spine) - osteoporosis (femoral neck)	14 (1)
Osteoporosis (lumbar spine) - normal (femoral neck)	18 (1.3)
Total discordance	595 (41.6)

Data are number (percentage).

The distribution of discordance did not significantly differ among the sociodemographic factors (Table 3). Furthermore, age and BMI were not normally distributed (P=.001). However, the patients who had major discordance were older (higher median age) than those who had concordant DXA results and those with minor discordance.

Table 3.

Relationship between discordance and sociodemographic factors.

Variables	Concordance	Minor discordance	Major discordance	P
Age (years)	66 (10)	67 (10)	69 (12)	.134
Gender
Female	633 (75.9)	410 (72.8)	23 (71.9)	.405
Male	201 (24.1)	153 (27.2)	9 (28.1)
Marital statusa
Single	27 (3.2)	12 (2.1)	0 (0)	.073
Married	715 (86)	483 (86.1)	23 (71.9)
Divorced	27 (3.2)	16 (2.9)	3 (9.4)
Widow	62 (7.5)	50 (8.9)	6 (18.8)
Current smokingb
No	788 (99)	520 (98.5)	30 (93.8)	.055
Yes	8 (1)	8 (1.5)	2 (6.3)
BMIc
Normal weight	104 (12.5)	93 (16.7)	4 (12.5)	.142
Underweight	4 (0.5)	4 (0.7)	0 (0)
Overweight	263 (31.7)	181 (32.4)	15 (46.9)
Obese	459 (55.3)	280 (50.2)	13 (40.6)

Data are median (interquartile range) for age and number (percentage) for categorical data.

^a5 missing cases (0.3%);

^b73 missing cases (5.1%);

^c9 missing cases (0.6%)

DISCUSSION

Our research findings indicated that a substantial proportion of the study patients exhibited discordance in T-scores between femoral and lumbar BMD measurements. Within this subset, most participants had minor discordance.

Several previous studies have shown results comparable to those of the current study. Woodson et al reported a T-score concordance rate of 56%, with a minor discordance rate of 39% and major discordance rate of 5%.⁹ Chan et al reported a minor discordance rate of 30.3% and major discordance rate of 2.3%.¹⁰ Moayyeri et al reported minor and major discordance rates of 38.9% and 2.7%, respectively. Furthermore, Mounach et al documented minor and major discordance rates of 42% and 4%.7,¹¹ Singh et al reported a concordance rate of 42.50% in their study, with 54.15% minor discordance and 3.35% major discordance.¹² Balachandran et al recorded a slightly higher major discordance rate of 6.1%, indicating regional variations in osteoporosis risk factors, such as obesity, which was also identified as a significant risk factor for discordance in their study.¹³ A recently published local study by Alzaid et al reported a major discordance rate of 6.1% and a minor discordance rate of 42.6%.¹⁴

Among our patients with both major and minor discordances, lower BMD was more commonly observed in the lumbar spine (47.7%) than in the femoral neck (42.1%). This could be due to dissimilar purposes. The disparity between the rates of bone disappearance in various parts of the human body could be a major reason.¹⁵ Most females in this study were postmenopausal, and spongy bones (typical of the lumbar area) are well-known to have a higher rate of demineralization in the early postmenopausal state than compact (cortical) bones (typical of the proximal femur).¹⁶ Furthermore, most of the subordinate causes of osteoporosis (such as glucocorticoid overconsumption, hyperthyroidism, hypothyroidism, malabsorption, liver disease, rheumatoid arthritis, obesity, and smoking) initially affect the vertebral column and then the femoral bone.¹⁷ This may lead to a higher prevalence of lumbar osteoporosis. In contrast, weight bearing, a known cause of physiologic disparity, can lead to an increase in bone density, especially in the hip and femoral areas.¹⁸ This may increase the major discordance in T-scores occurring due to an increase in BMI in this study, most notably in the female population.

Numerous factors have been reported in the literature as potential contributors to discordance in T-scores. Our findings did not indicate any significant cause of these differences. This is particularly notable because our study concentrated on older adults, a population inherently vulnerable to more pronounced age-related changes in BMD.

Chan et al reported that advanced age, decreased height, and physical activity levels were significant predictors of discordance in a Malaysian cohort.¹⁰ Similarly, Moayyeri et al revealed that discordance was notably observed in older patients,⁷ and Mounach et al identified age, menopausal status, and obesity as key factors associated with discordance in their study.¹¹ Furthermore, Singh et al indicated BMI, weight, age, and postmenopausal status as crucial predictors of discordance,¹² and Balachandran et al reported that obesity significantly influences the rate of discordance, highlighting the complex interplay of physiological factors that can affect BMD measurements.¹³ These results collectively underline the multifaceted nature of BMD discordance and suggest that various modifiable and nonmodifiable characteristics contribute to the differences in BMD measurements between the lumbar spine and hip.

In addition to age, obesity is related to higher rates of discordance; obesity is prevalent among menopausal and postmenopausal women aged >60 years. Therefore, the associations between BMD and obesity as well as between BMD and the postmenopausal state in women are complex.¹⁹ In patients with increased BMI on lumbar BMD measurements, DXA demonstrates the artifactual involvement of obesity and fat in the abdominal area.^20,21 In addition to artifactual components, increased waist perimeter—an indicator of visceral obesity—was noted to correspond to the extent of osteoporosis.

Discordance can occur due to physiological, pathophysiological, and anatomical causes, and evaluation can be complicated by artifacts and technical issues.⁹ The prevalence and underlying causes of discordance between femoral neck and lumbar spine BMD measurements can vary considerably across different populations. For accurate interpretation of DXA results, particularly in cases of major discordance, it is essential to understand these contributing factors thoroughly and formulate appropriate treatment recommendations. Comprehensive and in-depth research is warranted to elucidate these patterns within the Saudi population.

A significant strength of this study was the inclusion of a large sample size, which potentially represents rates of discordance in the general population. However, this study also had some limitations. Various factors that could affect discordance were not explored or quantified, and because the study primarily focused on older adults, other age groups were not investigated. Addressing these limitations in future research will contribute to a more thorough understanding of discordance across diverse demographics.

In conclusion, discordance between femoral and lumbar BMD measurements is prevalent among the older adult population in Saudi Arabia. This finding highlights the need for clinicians to be acutely aware of discordance when interpreting DXA results and consider a broader range of factors that might influence BMD. Further research aimed at improving DXA accuracy and the interpretation of results could significantly enhance the management of osteoporosis, reducing the burden of this disease on the healthcare system and improving quality of life of older adults.

Funding Statement

None.

CONFLICT OF INTEREST:

None.

DATA AVAILABILITY

The datasets used in this study are available from the corresponding author upon reasonable request.