Understanding Osteoporosis Prevalence in Hispanic ...

................................................................................ . . . . . . . . . . . . . . . . . . . . . .

PEER REVIEW

Understanding Osteoporosis Prevalence in Hispanic Women KEVIN D. EVANS, Ph.D., R.T.(R)(M)(BD), RDMS, RVS CHRISTOPHER A. TAYLOR, Ph.D., R.D., LD Context Older Hispanic women at risk for osteoporosis have not been studied using an ecological approach to determine the influential factors linked to this disease. Objective To discover if certain ecological factors influence the prevalence of osteoporosis among older Hispanic women. Methods A secondary data analysis was conducted using the National Health and Nutrition Examination Survey data set. The analysis focused on the responses from approximately 413 Hispanic women 60 years and older, a sample that represents 2.2 million U.S. residents. Ecological influence variables, including emotional support, financial support, smoking, housing, milk consumption and self-reported measures of osteoporosis, were chosen to determine if language (read or spoken) indicated a difference among these women. Results Descriptive results and a binary logistic regression model demonstrated the relationship between the independent variables and the self-reported measures of osteoporosis. The first model predicted an increased prevalence of an osteoporosis diagnosis in older Hispanic women who preferred Spanish, had decreased emotional support, had increased financial support, owned their homes and had been milk drinkers. The second model predicted an increased prevalence of self-reported fractures among older Hispanic women who preferred Spanish, had decreased emotional support, had increased financial support and had not smoked. Conclusion Ecological factors can be considered for older Hispanic women in making decisions about screening for disease as well as preventive health education.

here is a great deal of variability in the prevalence of osteoporosis across different populations. The variability is so great that diagnosis and management of this disease is not considered equal across populations.1 Researchers have struggled with a possible genetic explanation for this variability. Osteoporosis is a disease characterized by faulty bone structure or an inadequate amount of bone, resulting in fractures from slight trauma.2 Currently, no gene, gene product or gene polymorphism has been shown to account for the variance in bone mineral density (BMD) across populations.3 One particular group, older Hispanic women, has been difficult to describe relative to the incidence of osteoporosis. An example of this dilemma is the 6% prevalence of osteoporosis at the femoral neck among postmenopausal African American women in the United States vs a prevalence of 14% among Hispanic women.4 The belief is that the best methodology for unraveling the variability across populations would be to conduct research in genetics as well as environmental factors to explain variance in age, gender and race specific to BMD.2

To explore environmental factors specific to a particular age, gender and race, a model was needed to structure the research. Sallis and Owen5 developed such a model using an ecological approach to investigate how people were influenced by their physical and sociocultural surroundings. (See Fig. 1.) Using an ecological

RADIOLOGIC TECHNOLOGY July/August 2006, Vol. 77/No. 6

451

T

Social Environment

Intrapersonal Environment Cultural Environment

Physical Environment Diagnosis of Osteoporosis

Increasing Age of Participant

Fig. 1. Ecological research model for cultural risk factors for osteoporosis.

............................................................................................................. UNDERSTANDING OSTEOPOROSIS

approach provides a multifaceted method for describing risk factors that characterize the prevalence of this disease. A recent challenge was posed by researchers and physicians to develop appropriate criteria for conducting BMD studies for minority populations.6 Thus, this study was undertaken to determine if certain ecological factors influence the prevalence of osteoporosis among older Hispanic women. This kind of information has the potential to refine the screening criteria for bone density studies, raise the quality of preventive health information and focus patient education materials targeted to Hispanic women.

Literature Review Osteoporosis increases the incidence of fragility fractures by loss of the structural content of bone. Fragility fractures result from mechanical forces that would not ordinarily cause fractures in a healthy adult.7 Most of the information on fragility fractures and the subsequent diagnosis of osteoporosis are referenced to groups of white women, while other groups, such as African Americans, American Indians and Hispanics, are used less often to assess the prevalence of osteoporosis. Broad classifications of race and ethnicity are becoming more diverse and complex, which has made it increasingly difficult to quantify population risk factors.8 Recently, culturally sensitive studies of older women’s risk for osteoporosis have been conducted. Some of the important studies from this area are discussed to isolate salient variables, such as social, intrapersonal and physical environmental factors, that might establish a relationship between older Hispanic women’s cultural factors and osteoporosis. Because this body of literature is sparse, research that addressed either Mexican Americans or Hispanics was reviewed for its relevance to the prevalence of osteoporosis among older women who speak Spanish. A retrospective study of postmenopausal women was conducted using the National Osteoporosis Risk Assessment,9 in which data were collected from September 1997 to March 1999, with 12 months of subsequent follow-up data. This risk assessment was conducted with 200 160 postmenopausal ambulatory women who were 50 years and older and had undergone bone densitometry (BD).9 The BD scores were taken from peripheral sites, which represents a limitation when comparing this study to those using spine and hip measurements. Factors that were significantly related to their lower BD scores were Hispanic heritage (participants who preferred Spanish read or spoken), smoking,

452

age and personal history of fracture. These variables could be used in a similar study to show evidence of their influence on osteoporosis. Calcium consumption also is a recognized predictor of osteoporosis. A primary source of dietary calcium is dairy products. Lactase nonpersistence, or loss of enzymes that digest milk sugar (lactose), has been cited as a potential reason for lower milk consumption among certain racial groups and the elderly. Groups that have been identified as having significant levels of lactose intolerance are African Americans, Asian Americans, Hispanic Americans and American Indians.10 In contrast, a study was conducted on the calcium consumption among Hispanic participants living in North America; the results indicated that milk was the primary dietary source of calcium.8 Amerians of recent Mexican descent in this study also had an alternative source of calcium in those cases in which lactase nonopersistence might affect milk consumption levels. Calcium consumption is an important key to bone health, but alternative food sources such as corn tortillas should be considered for studies that want to capture Hispanic women’s total calcium consumption. A small, random digital survey was conducted by telephone with 43 Hispanic women who were 25 to 55 years, to investigate their levels of calcium consumption.11 The results of this study showed that the participants understood the link between body weight and the risk of osteoporosis. However, there was confusion among participants about the roles of calcium consumption and exercise as related to bone health. Although the study was small, it pointed out potential misunderstanding among older Hispanic women about the important role calcium consumption plays in the prevalence of osteoporosis. Smoking also has been investigated among a convenience sample of rural Mexican American women of recent Mexican descent to ascertain their level of understanding about its adverse effects. Interestingly, of the 78 women in the sample, a very low percentage (28%) of total participants associated smoking as a risk factor for osteoporosis.12 Although this study reported more descriptive information about rural Mexican Americans, potential misunderstanding was exposed among other Hispanic women about the risk of smoking and its relationship to osteoporosis. Thus, smoking is a variable worthy of more rigorous study as a risk factor for osteoporosis among older Hispanic women. Self-reported fractures assumed to be caused by osteoporosis were used as a variable for a study of older women

July/August 2006, Vol. 77/No. 6 RADIOLOGIC TECHNOLOGY

............................................................................................................. EVANS, TAYLOR

living within 1 Texas county. The study demonstrated that Mexican American women reported 67 per 100 000 incidences compared with 139 per 100 000 incidences of hip fractures for non-Hispanic white women based on population estimates from the 1980 census data.13 The number of self-reported fragility fractures point to a lower incidence of hip fractures among Hispanic women than non-Hispanic white women; however, it remains important to discover if additional factors might predict fragility fractures among older Hispanic women. This data would have been more meaningful if the recorded bone density data had been correlated to the calculated incidence rates. Acculturation is the cultural learning and behavioral adaptation that takes place among individuals exposed to new cultures.14 This type of learning and adaptation has been studied as a factor that could influence the health practices of Hispanic women. A group of 603 elderly Hispanic women were surveyed from 1984 to 1985; their responses to a questionnaire designed to measure acculturation indicated that, as a group, they had adapted marginally, but were very able to participate in health practices.15 A secondary research analysis of Hispanic Health and Nutrition Examination Survey (HHANES) data16 provided by Mexican American, Cuban American and Puerto Rican adults, ages 20 to 74, investigated acculturation as a significant variable that influences access to medical services. Language was the most predictive indicator of Mexican American participants’ ability to access medical services.16 Unfortunately, data concerning language were only collected for Mexican American participants. A replication of this study could use language as an acculturation variable, as well as age, gender and cultural background, to provide more descriptive information. Acculturation, as measured by language preference, could prove to be an important factor in determining the prevalence of osteoporosis among Hispanic women. More descriptive information about this growing group of older women is needed to better understand which factors influence their incidence of osteoporosis. The literature described here reports a variety of variables that were investigated individually. A holistic approach to determining the prevalence of osteoporosis among older Hispanic women would mean investigating variables both alone and in combination. Investigating variables in concert with each other could provide a new perspective on their combined influence on the development of osteoporosis within this specific group of older women.

RADIOLOGIC TECHNOLOGY July/August 2006, Vol. 77/No. 6

Statement of the Problem The purpose of using an ecological approach was to isolate variables that might help describe the prevalence of osteoporosis among older Hispanic women. An ecological approach has not been used to investigate potential relationships between cultural variables and the incidence of osteoporosis. The research hypotheses for this study were: ■ The cultural environmental factor of language influences the social environment of older Hispanic women (emotional and financial support). ■ The cultural environmental factor of language influences the diagnosis of osteoporosis among older Hispanic women (self-reported fractures or diagnosis). ■ The cultural environmental factor of language influences the intrapersonal environment of older Hispanic women (smoking and lifelong milk consumption). ■ The cultural environmental factor of language influences the physical environment of older Hispanic women (housing arrangement). ■ Intrapersonal and cultural environmental factors are the most significant factors in the diagnosis of osteoporosis in older Hispanic women. The proposed template for this research was adapted from the work of Sallis and Owen.5

Methodology This secondary data analysis used National Health and Nutrition Examination Survey (NHANES) data17 collected from 1999 to 2002. NHANES is a nationally representative probability sample of the entire U.S. population and captures extensive health data from respondents. NHANES was conducted by the National Center for Health Statistics and the Centers for Disease Control and Prevention to assess the health and nutritional status of the civilian, noninstitutionalized population of the United States. Data were collected from a 4-year national probability sample, which was selected using a stratified, multistage probability design. Low-income, elderly and Hispanic participants were oversampled to provide additional data on populations that are typically hard to represent in research studies.17 The NHANES data files provided for public use contain several health measures from a group of subjects restricted to Hispanic women ages 60 and older. For this inquiry, the dependent variable is the participant’s report of osteoporosis, which has been operationally

453

............................................................................................................. UNDERSTANDING OSTEOPOROSIS

defined in 2 ways: reported fractures of the hip, wrist or spine, or a previous diagnosis of osteoporosis. This investigation hypothesized that the prevalence of osteoporosis among this sample of elderly Hispanic women is influenced by a combination of ecological factors such as cultural, social, physical and intrapersonal issues. The independent variables were selected from NHANES to match the ecological model outlined by Sallis and Owen.5 The model depicts cultural environment as an overlapping influence on the intrapersonal, social and physical environment spheres. The model proposes that these spheres of influence have an overarching impact on the diagnosis of osteoporosis. Cultural environment was operationally defined as the NHANES measurement of acculturation. Within the acculturation data, the discrete measure of language read or spoken was used. Acculturation was further subdivided into those participants who preferred Spanish (read or spoken), either English or Spanish (read or spoken), or English (read or spoken). The social environment was operationally defined as the availability and use of emotional and financial support. Participant responses were coded again so that social support was categorized as being either available or unavailable. Individuals indicating that support was available but would not accept it were re-coded to indicate that support was available. The intrapersonal environment was defined by the NHANES measurements of lifestyle behaviors related to osteoporosis. Smoking and lifelong milk consumption were chosen to measure this particular dimension. Smoking status was determined using responses that indicated that the participant had smoked at least 100 cigarettes. To quantify milk intake during the bonebuilding years, a lifetime milk consumption score was computed using the responses about the amount of milk consumed at a series of ages. The categorical responses were coded in ascending amounts (0 = never, 1 = less than one per week, 2 = once per week, 3 = less than once per day but more than once per week, 4 = once per day and 5 = more than once per day). Each categorical response was multiplied by the number of years at each stage of inquiry. The stages of inquiry were from ages 5 to 12 years, 13 to 17 years and 18 to 35 years. A final milk consumption score was calculated by adding each multiplication group and then dividing the score by 35 years, or the age of the oldest measured age of intake. The ultimate score was reduced by 4 to account for not having a question about the stage of life during the first 4 years.18

454

Lastly, physical environment was assessed using housing as a definitive variable. The model’s physical environment was operationally defined as the NHANES measurement of housing characteristics. Within the data collected for housing characteristics, the measures of housing were re-coded as either home ownership or other arrangements.

Statistical Analysis Data analyses were conducted using the Statistical Package for the Social Sciences (SPSS, v13.0, SPSS Inc, Chicago, Ill) to provide adjusted error estimates required for the complex sampling strategy used in data collection. Using the data from the 1999-2002 NHANES allowed a chi-square (␹2) table to be constructed that compared the individual variables and percentages. A binary logistic regression was performed to identify the best predictors of osteoporosis among these older women using both the self-reported fractures and a diagnosis of osteoporosis.

Results Analyses were conducted for Hispanic women 60 years and older. The race/ethnicity of respondents was coded again so that both Mexican American and other Hispanic data were included for statistical comparison. The sample was composed of 413 women, which represents 2.2 million U.S. residents. The average age of respondents was 70 years (± .27 years), which was comparable across the groups composed by language preference. (See Table 1.) The ecological model was tested using operational variables provided by the 1999-2002 NHANES. For the social environment construct, significant differences were seen in the availability of emotional support (P