Abstract
Objective To quantify the burden of osteoporosis and examine the interplay between osteoporosis and various comorbidities as it relates to patient outcomes.
Methods Data from the 2011 Japan National Health and Wellness Survey (NHWS; n = 30 000), an internet health survey fielded to a nationally representative sample of the Japanese population were used. Only women between the ages of 50–90 years were included in the analyses (n = 6950).
Results Compared with matched controls (n = 404), patients with osteoporosis (n = 404) had lower MCS scores (48.94 vs 51.63), PCS scores (45.57 vs 49.12) (all p < 0.05). The presence of osteoporosis was associated with worse patient outcomes among those with hypertension, high cholesterol, and insomnia, among other conditions.
Conclusions The results suggest a significant quality-of-life and economic burden for patients with osteoporosis in Japan. Moreover, in a complex co-morbid environment, the presence of osteoporosis contributes more to patient outcomes than other chronic conditions.
Background
Patients with osteoporosis experience a range of comorbidities and osteoporosis has been linked to poorer health-related quality-of-life as well as economic outcomes such as increased healthcare resource utilization.
This article explored the relationship between osteoporosis and patient outcomes in Japan, which helps to address a recent call to action for more osteoporosis research in Asia.
This study also contributed to understanding the interplay between comorbidities and osteoporosis.
Despite patients with osteoporosis having a range of comorbidities, the results suggest that the poor patient outcomes are predominantly a function of their osteoporosis and not their comorbidity profile.
Introduction
Osteoporosis is a skeletal disorder characterized by reduced bone strength, leading to bone fragility and an increased risk of fracturesCitation1. Most common in post-menopausal women, the condition often results in frequent and acute fractures of the hip, wrists, shoulders, and spine, which can lead to chronic pain and disabilityCitation2–4. Osteoporosis is associated with a variety of risk factors including older age, excessive alcohol consumption, smoking, malnutrition, lack of exercise, endocrine disorders (including prolonged estrogen-deficiency), vitamin D deficiency, and as a by-product of numerous other medical conditions such as rheumatologic disorders and diabetesCitation5–7. Loss of bone mass can also be brought on by certain medications used to treat other disorders, such as glucocorticoidsCitation8,Citation9.
Worldwide, osteoporosis affects one in three women over the age of 50 and is the underlying cause of millions of fractures each yearCitation10,Citation11. These injuries result in tens of billions of dollars in related medical care annually and are responsible for a large portion of disability among elderly populationsCitation10,Citation12. As populations around the world age, osteoporosis is becoming a greater factor in increased morbidity and mortalityCitation13. In Japan, the latest estimates indicate that nearly 15 million people suffer from osteoporosisCitation14–16. The enormity of osteoporosis-related fractures, as well as related morbidity and mortality, make this disease of pressing concern in Japan.
However, there is a lack of research on the effect osteoporosis can have on patients in Japan. Osteoporosis has been linked with a reduced quality-of-life and greater fears of physical harmCitation17–19 along with a significant economic burdenCitation20–22 but, as noted in the call to action of Mithal and KaurCitation23, these associations have been less studied in Asia. One aim of this study is to explore the relationship between osteoporosis and a variety of patient outcomes including quality-of-life, work-related productivity, impairment in daily activities, and healthcare resource use in Japan. An additional aim is to explore the inter-play between osteoporosis and other comorbidities. Patients with osteoporosis have a number of comorbidities, some causally related and some notCitation5–7,Citation24. Prior research has found that the comorbidity profile of patients with osteoporosis can adversely affect their quality-of-lifeCitation24. This study also aims to quantify the burden these comorbidities have on patients with osteoporosis (and the burden osteoporosis has on patients with these comorbidities) to better tease apart the role osteoporosis has on patient outcomes in Japan.
Method
Data source
The data source for this study was the 2011 Japan National Health and Wellness Survey (NHWS; n = 30,000). The NHWS is a self-administered, Internet-based questionnaire from a nationwide sample of adults (aged 18 or older). A random stratified sampling framework was implemented to ensure the overall NHWS sample was comparable to the adult Japanese population with respect to age, sex, and regional distributions.
Potential respondents for this study were identified through the general panel of Lightspeed Research. Members of the Lightspeed Research panel were recruited through opt-in email, co-registration with Lightspeed Research partners, e-newsletter campaigns, banner placements, and other means. All potential panelists had to register with the panel through a unique email address and password and complete an in-depth demographic registration profile. All panel members explicitly agreed to become part of the panel and receive invitations to participate in online surveys. The survey received Institutional Review Board approval and all respondents provided informed consent prior to participating. Only NHWS respondents who were female, without a diagnosis of osteopenia (unless it occurred with a diagnosis of osteoporosis), between the ages of 50–90 were included (n = 6950). In Japan, physicians do not routinely diagnose osteopenia; therefore, it is unclear whether women who report a diagnosis of osteopenia truly have osteopenia or whether they actually have osteoporosis and this was misreported as osteopenia. Due to the uncertainty, respondents who reported a diagnosis of osteopenia without reporting a diagnosis of osteoporosis were excluded.
Measures
Demographics
Age (entered continuously), education (coded as university degree vs all else), household income (coded as <¥3,000,000, ¥3,000,000–¥5,000,000; ¥5,000,000–¥8,000,000; ¥8 000 000 or more; or decline to answer), health insurance (National Health Insurance, Social Insurance, Late Stage Elderly Insurance, other, or no insurance), and employment status (coded currently employed or not currently employed) were reported by respondents.
Health history
Smoking status (current, former, or never smoker), exercise behavior (regular exercise in the past month vs not), and alcohol use (currently consume alcohol vs abstain) were also included. Respondents provided their height and weight, which was converted to a body mass index (BMI) score and then converted to a BMI category: underweight (<18.5 kg/m2), normal weight (18.5–25 kg/m2), overweight (25–30 kg/m2), obese (30+ kg/m2) or decline to answer (for those who did not provide their weight). Self-reported diagnoses were used to calculate a Charlson comorbidity index (CCI), a measure of overall comorbidity burdenCitation25.
Osteoporosis
All respondents were asked whether they had experienced osteoporosis and, if so, whether their condition has been diagnosed by a physician. Only respondents who reported that they had been diagnosed were considered diagnosed and all others were considered not to have osteoporosis. This was purely based on self-report and no clinical information (such as bone mineral density test results) was available.
Comorbidities, family history, and fracture risk
The comorbidities examined in this study included high cholesterol, hypertension, and insomnia, which were selected a priori given their high prevalence as chronic conditions and availability within the NHWS dataset. All of these conditions have been found to be independently associated with poorer quality-of-life and greater societal costsCitation26–30. Additionally, elevated fracture risk and family history with osteoporosis were also included. Respondents with osteoporosis who reported that they had been diagnosed with high cholesterol, hypertension, and insomnia by a physician were considered as having each of these comorbidities. Fracture risk was calculated by inputting clinical risk factors (e.g. alcohol use, smoking behavior, body mass index, etc.) into the Japanese model of the FRAX algorithm at the University of SheffieldCitation31. Based on these risk factors, the algorithm predicted each respondent’s probability of having a fracture in a defined time horizon. High risk was operationalized as having a 15% or greater chance of a major osteoporotic fracture in the next 10 years. All others were considered as low risk. Respondents were asked directly whether they had a family history of osteoporosis.
Health outcomes
Health-related quality-of-life was assessed using the physical (PCS) and mental component summary (MCS) scores from the Short Form-12 version 2 (SF-12v2). Health utility scores (using the SF-6D algorithm) where derived from the SF-12v2 and includedCitation32–34. Work productivity and impairment was assessed using the Work Productivity and Activity Impairment (WPAI) questionnaireCitation35. Four sub-scales (absenteeism, presenteeism, overall work impairment, and activity impairment) were generated in the form of percentages, with higher values indicating greater impairment. These models were run only for those employed (full-time, part-time, or self-employed). Healthcare utilization was defined by the number of healthcare providers seen in the past 6 months, the number of traditional healthcare visits, the number of emergency room (ER) visits, and the number of times hospitalized in the past 6 months.
Statistical analyses
Three sets of comparisons were conducted: (1) patients with osteoporosis were compared with those without osteoporosis; (2) among those with osteoporosis, those with selected comorbidities, fracture risk, and family history were compared with those without those selected comorbidities, fracture risk, and family history; and (3) among those with each selected comorbidity, high facture risk, and family history (separately), those with osteoporosis were compared with those without osteoporosis. All three comparisons were conducted in the same manner, using a propensity score matching method, as described below. A propensity score matching method allows for two groups (e.g., osteoporosis vs no osteoporosis) to be made more comparable with respect to their characteristic differencesCitation36. In observational research, there can be systematic differences between groups that can be related to the outcome of interest (e.g., quality-of-life), thus obscuring true associations. For example, patients with osteoporosis may have a greater comorbidity burden than those without osteoporosis, making it unclear whether osteoporosis or the comorbidity burden is what is responsible for the association with poorer quality-of-life. Using a propensity score matching method, patients from the comparison group (e.g., no osteoporosis) are selected based on their similarity to patients in the group of interest (e.g., osteoporosis). As a result, any differences in the health outcomes can be attributed to what conceptually differentiates the groups (e.g., presence of osteoporosis) rather than other characteristic differences.
This propensity score matching process is conducted in several steps. First, differences between groups (e.g., osteoporosis vs no osteoporosis) were examined among demographics, health history, and health outcomes using chi-square and one-way analysis of variance (ANOVA) tests for categorical and continuous variables, respectively. Statistical significance was considered if p < 0.05.
Next, for each comparison group (e.g., osteoporosis vs no osteoporosis), a logistic regression was run predicting group membership from the following variables: age, education, annual household income, health insurance, smoking status, exercise behavior, alcohol use, BMI category, and the CCI. Propensity score values were saved from this regression. Propensity score values represent the probability of a respondent being a case (e.g., having osteoporosis) given their demographic, health history, and comorbidity data. Using a 1:1 greedy matching algorithmCitation36, each case (e.g., a patient with osteoporosis) was paired with a control (e.g., a patient without osteoporosis) whose propensity score value was identical. This allowed for patients and controls to have the same pattern of data with respect to their demographic, health history, and comorbidities, in order to rule out those variables being responsible for any relationships with patient outcomes. Controls that were not matched were excluded from further analysis. Post-match, differences between the comparison groups (e.g., osteoporosis vs matched controls without osteoporosis) were made with respect to all health outcomes (quality-of-life, work productivity, and healthcare resource use) using one-way ANOVA tests. Statistical significance was considered as p < 0.05.
Results
The association between osteoporosis and health outcomes
presents the demographic and health history information for the entire sample (n = 6950) as well as comparing those with (n = 404) and without (n = 6546) osteoporosis. Compared with patients not diagnosed with osteoporosis, patients with osteoporosis were older (67.91 years old vs 60.70 years old), less likely to be employed (15.84% vs 34.88%), more likely to have an annual household income of <¥3 million (25.50% vs 18.27%), and had a higher CCI (0.27 vs 0.16) (all p < 0.05; see ).
Table 1. Demographic and health history differences between those with and without osteoporosis.
With respect to health outcomes differences between those with and without osteoporosis, patients with osteoporosis had lower PCS mean scores (45.52 vs 50.41), lower health state utilities (0.735 vs 0.783), higher levels of overall work impairment (23.3% vs 14.5%), and higher levels of activity impairment (28.42% vs 17.25%) compared with patients without osteoporosis (all p < 0.05; see ). Additionally, patients with osteoporosis reported more healthcare provider visits (16.68 vs 5.80), ER visits (0.25 vs 0.08), and hospitalizations (1.40 vs 0.41) in the past 6 months (all p < 0.05).
Table 2. Health outcomes differences between those with and without osteoporosis.
After implementing the propensity score match, one patient with osteoporosis did not have a suitable match and was removed from the analyses. The remaining patients with osteoporosis (n = 403) were compared with matched controls without osteoporosis (n = 403; see ). Patients with osteoporosis had lower MCS scores (48.94 vs 51.63), PCS scores (45.57 vs 49.12), and health state utilities (0.736 vs 0.795; all p < 0.05) compared with matched controls. Additionally, patients with osteoporosis had higher levels of overall work impairment (23.3% vs 11.1%), activity impairment (28.26% vs 18.61%), and reported more healthcare provider visits in the past 6 month (16.66 vs 7.92; all p < 0.05) compared with matched controls.
Table 3. Health outcomes differences between those with osteoporosis and matched controls.
The association between comorbidities and health outcomes among those with osteoporosis
Among respondents with osteoporosis, those with and without select comorbidities were generally no different with respect to health outcomes (see ). One exception was that those with high fracture risk reported significantly worse mean health utilities relative to matched controls (0.719 vs 0.763, p < 0.05). The only other differences (presenteeism and overall work impairment being significantly higher among those with comorbid insomnia relative to matched controls) were difficult to interpret, as only one patient had comorbid insomnia and work impairment data.
Table 4. Health outcomes differences between those with select comorbidities and conditions and matched controls among patients with osteoporosis.
The association between osteoporosis and health outcomes across various conditions
The presence of osteoporosis was associated with significant decrements in health outcomes among patients with those same selected comorbidities, high fracture risk, and a family history of osteoporosis (see ). In all cases, the presence of osteoporosis was associated with significant and clinically relevant decrements in PCS scores and health utilities and, in the case of high cholesterol, family history, high fracture risk, and significant and clinically relevant decrements in MCS scores (all p < 0.05).
Table 5. Health outcomes differences between those with osteoporosis and matched controls among patients with select comorbidities and conditions.
Although work-related impairment was no different between those with and without osteoporosis within each comorbidity/condition, levels of activity impairment were significantly higher among those with osteoporosis (all p < 0.05; see ). For patients with insomnia, higher rates of ER visits and hospitalizations were observed among those with osteoporosis (both p < 0.05). Among patients with hypertension, a family history, and a high fracture risk, the presence of osteoporosis was associated with significantly more physician visits (all p < 0.05).
Discussion
Despite osteoporosis increasing in its importance due to the growing elderly population in Japan, there is a lack of research investigating the patient outcomes of this population. Consistent with literature reported from other countriesCitation17–19, patients diagnosed with osteoporosis in Japan reported poorer health-related quality-of-life and higher levels of activity impairment compared with matched controls. Indeed, these health-related quality-of-life effects exceeded cut-offs for clinical significance. Furthermore, consistent with prior researchCitation20–22, patients diagnosed with osteoporosis demonstrated a higher economic burden, as indicated by having greater overall work impairment (among those still employed), and more healthcare provider visits in the past 6 months than matched controls.
Although these associations, like prior studies, incorporated comorbidities as confounding variables, it is difficult to properly account for the role of comorbidities in influencing patient outcomes because of the complexity of properly measuring their presence and severity. This is a particularly important issue with osteoporosis, in which the comorbidity burden is substantial.
However, our additional analyses helped to address this by examining the incremental effect of comorbidities on patients with osteoporosis as well as examining the incremental effect of osteoporosis on patients with select conditions. Our results suggest that the addition of specific comorbidities/conditions (e.g., hypertension, high cholesterol, high fracture risk) generally had little effect on the health outcomes of patients with osteoporosis. The presence of hypertension, or high cholesterol, or insomnia was not associated with poorer outcomes. Conversely, the presence of osteoporosis had a significant effect on health-related quality-of-life, activity impairment, and healthcare resource use among those with the same selected comorbidities/conditions. These results suggest that the poor health outcomes observed among patients with osteoporosis is primarily a function of the osteoporosis rather than other chronic conditions. For a patient with osteoporosis, the addition of hypertension or high cholesterol does not make an appreciable difference on their patient outcomes. Conversely, for a patient with hypertension, the addition of osteoporosis has a substantial influence on those very same outcomes.
Limitations
The NHWS data is cross-sectional. As such, causal relationships between diagnoses, fractures, treatments, and health outcomes cannot be assumed. Additionally, all data are patient reported, so no verification of osteoporosis or risk status could be conducted. Indeed, it is quite possible that respondents who did not report a diagnosis of osteoporosis may, in fact, have undiagnosed osteoporosis. We were also limited by the data collected in the NHWS. For example, information on fractures was not available, nor was a comprehensive listing of comorbidities. As a result, our analyses on the impact of osteoporosis was not able to account for the number of fractures which undoubtedly would affect health outcomes. Similarly, there may be other comorbidities which could affect the health outcomes of patients with osteoporosis but were not assessed. Finally, although the NHWS is broadly representative of the Japanese adult population, it is unclear the extent to which the female 50–90 years old population (and those with osteoporosis within this sub-group) are representative of the larger population.
Conclusions
In summary, results from this study suggest a significant health-related quality-of-life and economic burden for patients with osteoporosis in Japan. Moreover, in a complex comorbid environment, the presence of osteoporosis contributes more to patient outcomes than other chronic conditions. Although the current study utilizes cross-sectional data and, thus, causal relationships between diagnoses and health outcomes cannot be assumed, these results add to a better understanding of the disease burden of osteoporosis in Japan. Indeed, in light of these findings and the increasing prevalence of osteoporosis, greater attention to the proper management of osteoporosis is warranted.
Transparency
Declaration of funding
This study was funded by Pfizer Japan, Inc.
Declaration of financial/other relationships
LAY and IK are employees of Pfizer Japan, Inc., which purchased access to the National Health and Wellness Survey dataset, funded the analysis, and funded the preparation of this manuscript. MDB is an employee of Kantar Health, which conducts the NHWS, and was the recipient of funding from Pfizer Japan, Inc. JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.
Acknowledgments
The authors would like to acknowledge the contributions of Dr Megan Shen, who assisted in the literature review and provided editorial support. Dr Shen is a paid consultant to Kantar Health.
Related Research Data
References
- NIH Consensus Development Panel on Osteoporosis Prevention, Diagnosis, and Therapy. Osteoporosis prevention, diagnosis, and therapy. JAMA 2001;285:785-95
- Seeman E. Pathogenesis of bone fragility in women and men. Lancet 2002;359:1841-50
- Braithwaite RS, Col NF, Wong JB. Estimating hip fracture morbidity, mortality and costs. J Am Geriatr Soc 2003;51:364-70
- Center JR, Nguyen TV, Schneider D, et al. Mortality after all major types of osteoporotic fracture in men and women: an observational study. Lancet 1999;353:878-82
- Sampson HW. Alcohol and other factors affecting osteoporosis risk in women. Alcohol Res Health 2002;24:292-8
- Hannan MT, Felson DT, Dawson-Hughes B, et al. Risk factors for longitudinal bone loss in elderly men and women: The Framingham Osteoporosis Study. J Bone Miner Res 2000;15:710-20
- Waugh EJ, Lam M-A, Hawker GA, et al. Risk factors for low bone mass in healthy 40–60 year old women: a systematic review of the literature. Osteoporos Int 2009;20:1-21
- Boling EP. Secondary osteoporosis:u disease and the risk for glucocorticoid-induced osteoporosis. Clin Ther 2004;26:1-14
- Grossman JM, Gordon R, Ranganath VK, et al. American College of Rheumatology 2010 recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis. Arthritis Care Res 2010;62:1515-26
- Johnell O, Kanis JA. An estimate of the worldwide prevalence and disability associated with osteoporotic fractures. Osteoporos Int 2006;17:1726-33
- Johnell O, Kanis J. Epidemiology of osteoporotic fractures. Osteoporos Int 2005;16:S3-7
- Harvey N, Dennison E, Cooper C. Osteoporosis: impact on health and economics. Nat Rev Rheumatol 2010;6:99-105
- Rachner TD, Khosla S, Hofbauer LC. Osteoporosis: now and the future. Lancet 2011;377:1276-87
- Iki M. [Epidemiology of osteoporosis in Japan]. Clin Calcium 2012;22:797-803
- Tatsuno I, Terano T, Nakamura M, et al. Lifestyle and osteoporosis in middle-aged and elderly women: Chiba bone survey. Endocr J 2013;60:643-50
- Population trends in Japan. Tokyo, Japan: Ministry of Health Labour and Welfare in Japan, 2011
- Martin AR, Sornay-Rendu E, Chandler JM, et al. The impact of osteoporosis on quality-of-life: the OFELY cohort. Bone 2002;31:32-6
- Cook DJ, Guyatt GH, Adachi JD, et al. Quality of life issues in women with vertebral fractures due to osteoporosis. Arthritis Rheum 1993;36:750-6
- Borgström F, Lekander I, Ivergård M, et al. The International Costs and Utilities Related to Osteoporotic Fractures Study (ICUROS)—quality of life during the first 4 months after fracture. Osteoporos Int 2013;24:811-23
- Svedbom A, Ivergård M, Hernlund E, et al. Epidemiology and economic burden of osteoporosis in Switzerland. Arch Osteoporos 2014;9:1-8
- Brown P, McNeill R, Leung W, et al. Current and future economic burden of osteoporosis in New Zealand. Appl Health Econ Health Policy 2011;9:111-23
- Becker DJ, Kilgore ML, Morrisey MA. The Societal Burden of Osteoporosis. Curr Rheumatol Rep 2010;12:186-91
- Mithal A, Kaur P. Osteoporosis in Asia: a call to action. Curr Osteoporos Rep 2012;10:245-7
- Guillemin F, Martinez L, Calvert M, et al. Fear of falling, fracture history, and comorbidities are associated with health-related quality of life among European and US women with osteoporosis in a large international study. Osteoporos Int 2013;24:3001-10
- Charlson M, Pompei P, Ales K, et al. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 1987;40:373-83
- Daley M, Morin CM, LeBlanc M, et al. Insomnia and its relationship to health-care utilization, work absenteeism, productivity and accidents. Sleep Med 2009;10:427-38
- Bolge SC, Doan JF, Kannan H, et al. Association of insomnia with quality of life, work productivity, and activity impairment. Qual Life Res 2009;18:415-22
- Trevisol DJ, Moreira LB, Kerkhoff A, et al. Health-related quality of life and hypertension: a systematic review and meta-analysis of observational studies. J Hypertens 2011;29:179-88
- Wang G, Fang J, Ayala C. Hypertension-associated hospitalizations and costs in the United States, 1979-2006. Blood Press 2014;23:126-33
- Kannan H, Thompson S, Bolge SC. Economic and humanistic outcomes associated with comorbid type-2 diabetes, high cholesterol, and hypertension among individuals who are overweight or obese. J Occup Environ Med 2008;50:542-9
- Fujiwara S, Nakamura T, Orimo H, et al. Development and application of a Japanese model of the WHO fracture risk assessment tool (FRAXTM). Osteoporos Int 2008;19:429-35
- Ware J, Kosinkski M, Keller S. A 12-item short-form health survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996;34:220-33
- Fukuhara S, Bito S, Green J, et al. Translation, adaptation, and validation of the SF-36 Health Survey for Use in Japan. J Clin Epidemiol. 1998;51:1037-44
- Fukuhara S, Ware JE, Jr, Kosinski M, et al. Psychometric and clinical tests of validity of the Japanese SF-36 Health Survey. J Clin Epidemiol 1998;51:1045-53
- Reilly M, Zbrozek A, Dukes E. The validity and reproducibility of a work productivity and activity impairment instrument. Pharmacoeconomics 1993;4:353-65
- Parsons L. Reducing bias in a propensity score matched-pair sample using greedy matching techniques. SAS Conference Proceedings. Long Beach, CA, 2011. p. 214–26