Cardiovascular disease (CVD) is the No. 1 killer of women in the United States, with more than 316 000 women dying in 2006 of this disease.1 Furthermore, sudden cardiac death increased 10% from 1990 to 2000 in women younger than 35 years old.2 In addition, women experiencing their first myocardial infarction before age 50 years were twice as likely as their male counterparts to have a fatal outcome.3 Of the women who survived the initial infarction, 42% eventually died within 1 year, compared with only 24% of men.2
Women’s awareness of their risk factors for CVD is poor despite national media campaigns by the American Heart Association (AHA) and the National Heart, Blood and Lung Institute. In fact, more than 40% of women in 2004 were unaware of their cardiovascular risk factor (CVRF) status,4,5 and women continue to be more concerned with the risk of breast cancer than heart disease.1,5 The morbidity for women associated with CVD exceeds that associated with breast cancer. In the 2006 mortality data for women, there were more than 316 000 CVD-related deaths (1 in 2.5) compared with the approximately 40 000 deaths secondary to breast cancer (1 in 30).1 Although studies have shown an increase in women’s awareness of CVD as a leading cause of death for women, they do not number it among their leading health concerns.4,5 Therefore, research is needed to help women become knowledgeable about CVD and the CVRFs in an attempt to reduce their risk of developing CVD through heart-healthy behaviors.
The first step in preventing CVD is assessment of cardiovascular risks.5 Major risk factors for CVD have been shown to include advancing age, smoking, diabetes mellitus, dyslipidemia, family history, hypertension, obesity, sedentary lifestyle, and intake of saturated fats and low dietary fiber.6 Because all these factors except age and family history can be prevented and/or treated, CVD in women is partially preventable and treatable with primary and secondary prevention measures. Primary preventive measures target health promotion activities that protect against the development of a disease, whereas secondary preventive measures target early detection, diagnosis, and treatment of a pathological process. Primary prevention for CVD may involve support of healthy-lifestyle choices such as daily moderate intense physical activity (PA), following AHA nutritional guidelines, not smoking and limiting alcohol ingestion to 1 or less serving per day. Secondary prevention methods include the identification of known risk factors for the development of CVD such as hyperlipidemia or obesity, as well as early treatment and modification of these risk factors. Tertiary prevention addresses the goal of preventing further established disease progression, reducing the associated disability.
Coronary artery disease develops over time, with most CAD in women being diagnosed around the time of menopause.1 A woman’s risk for developing CVD dramatically increases with more CVRFs. Proactive primary and preventive care in the years prior to menopause, through knowledge of CVRFs and their implications, would reduce the incidence of CVD. Indeed, modifying CVRFs in men and women has been linked to reducing CVD by as much as 31%.7 These data support the need to identify and prevent CVD as early as possible, including reinforcing the goal to educate women about CVRFs and to conduct research that promotes heart-healthy behaviors in women.5 Strategies focusing on prioritizing and addressing women’s risk factors for CVD before they are menopausal, through primary and secondary preventive methods, may prove the most efficacious way of combating the CVD epidemic.
Individuals will pursue a certain behavior or goal based on the physiological and psychological variables experienced as a “push” to action.8 In other words, they initiate or persist in behaviors they believe will lead to a desired outcome. These behaviors and actions reflect an individual’s beliefs and values, which are in turn determined by psychological mediators and social factors.9,10 Motivation prompts an individual to keep moving toward a goal, thus attempting to satisfy an innate psychological need.11 Choosing behaviors to meet these needs can be characterized as nurturing self, because the result is a sense of achievement or accomplishment for supporting growth, health, and life. Indeed, actions are influenced by the interaction of personal characteristics of cognition and biological and environmental factors.12 Behaviors are purposeful. Understanding the value attributed to specific behaviors can help researchers and clinicians develop interventions to support continuation of those behaviors.13 Health behavior, health promotion, behavior change, and health-promoting lifestyles are important factors that undergirded this study through their direct links with the Nemcek Wellness Model (NWM),14 the conceptual framework for this research.
The NWM was developed based on self-nurturance research, explanatory models of health behavior, and life satisfaction.6,8,15,16 This model integrates aspects of established health behavior theories such as the Health Promotion Model17 and Health Belief Model18 and uses a systems approach to predicting wellness behaviors. Self-nurturance refers to “the health promotional choices made by the individual.”19(p241) It implies that the self is responsible for the basic decision to exhibit healthy behaviors (ie nutritious dietary intake, regular exercise) and express distaste for self-destructive behaviors (ie smoking or excessive alcohol ingestion). Health is enhanced through nurturance, an engagement in those activities that holistically nourish the individual. Self refers to the separate and distinct individual and, when in conjunction with an activity, implies a process that is primarily the responsibility of that individual. The decision to implement that activity is ultimately the choice of the individual.15
Assessing self-nurturance as it relates to women’s knowledge of CVRFs and levels of physical and mental wellness may assist in developing interventions that support the pursuit of a heart-healthy way of life. Healthy outcomes in the midyears and older years of life are dependent on proactive, preventive lifestyles started earlier in life and on a supportive healthcare environment. Thus, this study was designed to evaluate women’s CVRFs and their participation in preventive heart-healthy activities during the premenopausal years (ages 35–55 years). The elements of the NWM explored in this research were the relationships among knowledge of CVRFs, self-nurturance, and heart-healthy behaviors20 (engaging in PA, eating a heart-healthy diet, not smoking, and moderating alcohol intake).
Explanatory models of health behaviors and life satisfaction also propose that desired health outcomes depend on one’s knowledge about a desired outcome, such as heart health, and about the choices of heart-healthy behaviors, needed to attain the outcome.17 Participation in these choices may be moderated by self-nurturance.14 Therefore, this study was conducted to investigate knowledge of CVRFs, level of self-nurturance, and performance of heart-healthy behaviors in 35- to 55-year-old women. The heart-healthy behaviors assessed were (1) participating in moderate-intensity PA at least 5 days per week, (2) following a heart-healthy diet, (3) having 1 or fewer alcoholic beverages per day, and (4) not smoking.
A cross-sectional survey design with multivariate analysis was used to examine the relationships among women’s knowledge of CVRFs, level of self-nurturance, and heart-healthy behaviors. A power analysis was calculated for each of the instruments utilized in this study. The projected sample size for the analyses ranged from 21 to 107 when calculated for the inference for means comparing 2 independent samples (α level = .05, effect size = 0.50) and was 97 for the linear regressions with 6 independent variables (α level = .05, effect size = 0.15) with a statistical power of 0.8. For this study, a sample of 107 completed surveys was needed to obtain the desired power.
To assess the survey instruments for ambiguity and estimate the time necessary to complete the survey, the instrument packet was given to 7 women, aged 35- to 55 years. The pilot participants were asked to review and complete the survey questionnaire and comment on their perceptions of the instruments. Respondents indicated that the survey was easy to read and understand. They were able to complete the survey in 15 to 20 minutes. No changes or modifications were suggested by these participants.
Sample and Setting
Participants were recruited after the study was approved by the author’s institutional review board. Premenopausal women from medically underserved areas/populations or health professional shortage areas21 in a New England state were recruited by venue-based time-space convenience sampling. This method is designed to access hard-to-reach populations by recruiting them at times and places where they congregate.22 Therefore, women were recruited from a chain discount department store in northern Connecticut on weekday evenings and weekend days. This store was chosen as a sampling venue because it is frequented by 83% of US households to buy personal and home items, 55% of its shoppers list incomes of less than $39 000, and 46% of its shoppers list incomes of $50 000.23 Half of its shoppers are women 44 years or younger.
Women were eligible for the study if they met these criteria: (1) between 35 and 55 years old, (2) physiologically able to menstruate, (3) menstruated within the previous 6 months, and (4) able to read English. Women were excluded if they reported a previous myocardial infarction or cerebral vascular accident. Women meeting the criteria and indicating an interest were given a survey packet, which included a letter of introduction, explanation of the research project, and the survey. A returned completed survey packet implied the participant’s consent to participate. All participants returning a survey were entered into a drawing held at the completion of study enrollment. The drawing prize was a “Caring for Your Heart-Healthy” basket (approximate value, $175) that contained Go Red for Women items (canvas bag, lunch tote, travel mug, eating healthy grocery list), pharmacy gift certificate, heart-healthy cookbook, iPod shuffle-Red, and pedometer.
Data were collected from January through February 2009. Women entering the venue were approached and asked to participate in the survey. The participants had the opportunity to complete the survey on-site or to take the survey home and return it in an addressed, stamped envelope. The packet included the descriptive letter of introduction, estimating the time for completion of the survey as approximately 15 to 20 minutes, and the survey. The survey included a demographic information form, Heart Disease Facts Questionnaire (HDFQ-2),24 Self-Nurturance Survey (SNS),19 International Physical Activity Questionnaire (IPAQ),25 PrimeScreen26 instrument, and a numbered raffle ticket.
Recruitment sessions were held until at least 107 surveys had been returned as this number was the threshold identified by the power analysis to meet the established power (0.8) for this study. Once the threshold had been reached, no further venue recruitment sessions were held, but mail returned surveys were accepted for 1 month after the last recruitment session. A total of 258 surveys were distributed during the venue-based recruitment sessions to interested women entering the venue. One hundred forty-seven surveys were returned either directly to the venue or via US mail, for a response rate of 57%. Of the 147 returned surveys, 11 mail returned surveys were excluded because the participants did not meet the inclusion criteria of having a menstrual cycle in the last 6 months and they were not using a contraceptive method that prohibited menstruation.
Knowledge of Cardiovascular Disease Risk Factors
This variable was measured using the 25-item HDFQ-2,24 which measures knowledge of heart disease risk. The HDFQ-2 has 9 domains: age, sex, smoking status, glycemic control, cholesterol levels, blood pressure, PA, weight, and knowledge about CVD. The response options include true, false, and “I don’t know.” Scores are calculated by adding the number of items answered correctly and multiplying by 4; scores can range between 0% and 100%. The higher the participant’s score, the greater the knowledge of CVD.24 In this study, reliability of the HDFQ-2 (Kuder-Richardson coefficient) was 0.85 (Table 1).
This variable was measured using the 29-item SNS,19 which assesses self-chosen thoughts and feelings as well as health-fostering behaviors. The SNS includes statements concerning health promotional behaviors (eg, “I eat right”) or attitudes (eg, “I forgive myself if I have done something wrong”). Each item is rated from 1 (“not at all true”) to 5 (“extremely true”). Scores are calculated by the mean score of the 29 items. Higher scores indicate higher levels of self-nurturance.19 In the present study, the SNS reliability (Cronbach α) was .92 (Table 1).
This variable was measured using the 7-item IPAQ, which assesses PA in individuals 15 to 65 years old in developed and developing countries.25 The 7-item self-administered short form used for this study measures 4 types of activities: sedentary, moderate-intense, vigorously intense, and walking activities. The instrument provides the individual’s score in 4 domains of PA: leisure time PA, domestic PA, work-related PA, and transportation-associated PA.
Based on a 7-day recall, a total score is calculated by adding the duration and frequency of PA in each domain. The domains cannot be estimated separately. Weekly estimate of total PA is calculated by weighting the reported minutes per week in each PA domain. The last question of the survey is the indicator variable of sedentary time and is not included in the total PA score.27 The reliability and validity of the IPAQ have been extensively tested in 12 countries including the United States,25,27 with test-retest Spearman rank-order correlation coefficients ranging from 0.65 to 0.88. Test-retest reliability of the IPAQ for this study was 0.65 to 0.88 (Table 1).
Heart-Healthy Dietary Intake
This variable was measured using the 18-item PrimeScreen, which assesses adults’ average frequency of consuming specific food groups in primary care settings.26 The PrimeScreen does not measure total dietary intake. Five response options are given for the frequency of consumption and specifically assessed intake of fruits, vegetables, whole grains, fish, red meat, low- and whole-fat dairy items, and saturated and unsaturated fats. Each category of food is given a positive or negative value based on the level of consumption. The total score is calculated by adding each value to give a summary variable. Scores of 35 to 42 indicate excellent intake of a nutritionally healthy diet, 16 to 34 indicates a good dietary intake, and 1 to 15 indicates less than the recommended intake of nutritionally healthy foods and nutrients.26 In this study, the threshold for heart-healthy eating was set at a PrimeScreen score of 16, and scores were dichotomized for analysis into 16 or greater (heart healthy) or less than 16 (not heart healthy).
The test-retest reliability of the PrimeScreen was assessed by administering it 2 times and determining Spearman correlation coefficients of food groups and 13 selected nutrients.26 Spearman coefficients ranged from 0.50 for other vegetables to 0.87 for added salt, the mean r was 0.70; correlations of each food group and selected nutrients ranged from 0.36 for other vegetables to 0.82 for whole eggs, with a mean r of 0.61. Internal consistency reliability in this study (Cronbach α) was .60 (Table 1).
This variable was assessed by asking one question: How hard is it to pay for the very basics like food, housing, medical care and heating?28 Responses were 0 (not hard at all), 1 (somewhat hard), or 2 (very hard), with higher scores reflecting more financial strain. For data analysis, responses were dichotomized into 0 (“not hard at all”) versus 1 (“somewhat hard” and “very hard”).
Smoking Status and Alcohol Intake
These variables were assessed by questions on the demographic section of the survey. Participants were categorized as current everyday smokers, current some-day smokers, former smokers, or never smoked. For data analysis, this variable was dichotomized into current smoker and nonsmoker. Alcohol intake was categorized as abstains from alcohol, drinks 1 to 3 alcoholic beverages per week, drinks not more than 1 alcoholic beverage per day, or drinks more than 2 alcoholic beverages per day. For data analysis, this variable was dichotomized to drinks more than 2 alcoholic beverages per day versus drinks less than 1 alcoholic beverage per day.
Sample characteristics were analyzed by descriptive statistics. Demographic variables of age, financial strain, and education level were dichotomized for data modeling. Relationships between scores for knowledge, self-nurturance, and heart-healthy dietary intake; nonsmoking status; and PA were analyzed by Pearson product-moment correlation coefficients if variables were normally distributed and by the Spearman rank order correlation for those not normally distributed.
Logistic regression was used to assess relationships between CVRF knowledge levels and self-nurturance levels with the dichotomized variables of nonsmoking status, PA, and heart-healthy dietary intake. Significance level was set at .05 for each test.
Differences in knowledge levels by sociodemographic variables (age, education, and financial strain) were assessed using the nonparametric Mann-Whitney U test. Unadjusted associations between each independent demographic variable (age, college education, and financial strain) and each heart-healthy behavior (recommended PA, recommended heart-healthy dietary intake, and nonsmoking status) were analyzed by χ2 test.
Differences between PA (minutes per week) and heart-healthy dietary intake score by age group, educational category, and financial strain category were analyzed by Student t test. Differences in nonsmoking status by age group (median split categories), educational category, and financial strain category were analyzed by χ2 test. Binary logistic regression was used to evaluate heart-healthy behaviors as dichotomized variables with the set of continuous demographic variables. This model was then expanded by including scores for CVRF knowledge and self-nurturance with the 4 demographic variables to evaluate if either former variable contributed any explanatory power after accounting for the latter variables.
Most participants were white (94.9%) and married (80.1%), with a mean age of 45.2 years (Table 2). The participants’ racial makeup reflected available demographic information for their county.23 Most study participants did not smoke (80.1%); the majority rarely or never drank alcohol (57.4%) and did not experience financial strain (70.6%). About half of the participants had post–high school education (50.9%), and only 2.2% had not completed high school.
Participants’ mean HDFQ-2 score was 78.93 (SD, 17.93), indicating high knowledge of CVRFs. Their mean self-nurturance score was 3.41 (SD, 0.60). Participants’ PrimeScreen mean score was 8.46 (SD, 0.56), but most women (85%) did not meet the recommended intake for a heart-healthy diet (score ≥16).
Heart-healthy behaviors included nonsmoking status, less than 1 alcoholic beverage intake per day, at least 30 minutes of PA per day, and a PrimeScreen nutritional intake score 16 or greater. Most participants (80%) were either former or nonsmokers, and 20% smoked some days or every day (Table 3). Participants were generally active, with 63.2% reporting at least 30 minutes of exercise each day. Alcoholic beverage consumption was removed from the analysis because only 1 participant reported consuming more than 1 drink per day. Race and ethnicity were also eliminated from the analysis because only 3 participants self-identified as being from racial or ethnic minorities.
Relationships between self-nurturance and the behavior variables of PA and heart-healthy dietary intake were assessed by Pearson correlations for continuous variables. A moderate correlation was found between self-nurturance and heart-healthy dietary intake (r = 0.331, P < .05) but not between self-nurturance and PA (r = 0.067, P < .05).
Differences in CVRF knowledge level by age, education, and financial strain categories as well as by heart-healthy behaviors were assessed by nonparametric Mann-Whitney U analysis. Knowledge level did not differ significantly by age group, separated by median split into 46 years or younger, or older than 46 years (Z = −0.652, P = .52), but differed significantly by education (Z = −2.55, P = .01) and financial strain (Z = −2.08, P = .04). Higher knowledge scores were associated with less financial strain and college education. Cardiovascular risk factor knowledge did not differ with any heart-healthy behavior examined, that is, by smoking status (Z = −1.59, P = .112), PA category (Z = −1.83, P = .067), or heart-healthy dietary intake (Z = −1.76, P = .079) (Table 4).
Differences in heart-healthy behaviors by demographic characteristics were analyzed by χ2 analysis. Physical activity did not differ significantly by age group (≤46 or >46 years) (χ21 [n = 136] = 0.012, P = .91) or by educational category (no college education or at least some college education) (χ21 [n = 136] = 0.609, P = .44). However, PA differed significantly by financial strain category (χ21 [n = 136] = 4.270, P = .04). Participants who self-rated as having less financial strain were more likely to participate in at least 30 minutes of daily PA.
Heart-healthy dietary intake was not significantly different by age group (χ21 [n = 136] = 1.021, P = .31) or by financial strain category (χ21 [n = 136] = 2.737, P = .10), but was significantly different by educational category (χ21 [n = 136] = 4.427, P = .04). Participants with no college education were less likely to follow a heart-healthy diet. Nonsmoking status did not differ significantly by age group (χ21 [n = 136] = 0.984, P = .32), financial strain category (χ21 [n = 136] = 0.943, P = .33), or educational category (χ21 [n = 136] = 1.003, P = .32).
Differences in self-nurturance levels by heart-healthy behaviors (Table 5) and demographic variables were evaluated by t test for continuous variables and χ2 test for categorical variables. Self-nurturance differed significantly by heart-healthy dietary intake (t = −3.08, P = .002), educational category (t = −4.06, P = .000), and financial strain category (t = 3.41, P = .001). Self-nurturance did not differ significantly by PA (t = −4.19 P = .676) or nonsmoking status (t = 0.385, P = .70).
The relationships between heart-healthy behaviors and levels of CVRF knowledge and self-nurturance were assessed by logistic regression analysis. Each heart-healthy behavior (PA, heart-healthy dietary intake, and nonsmoking status) was first modeled with demographic variables (age, financial strain, and college education). A second model was generated for each heart-healthy behavior as a dependent variable, with mean knowledge and self-nurturance scores as additional independent variables.
In the first model, women’s nonsmoking behavior was not predicted by age group (B = 0.48, P = .29), financial strain category (B = 0.34, P = .50), or educational category (B = −0.48, P = .34). Heart-healthy dietary intake was not predicted by age group (B = 0.72, P = .16), financial strain category (B = 0.68, P = .33), or educational category (B = −1.13, P = .07). Physical activity levels were not predicted by age group (B = 0.054, P = .89) or educational category (B = −0.021, P = .96). However, PA was predicted by the perceived “hard” financial strain category (B = 0.82, P = .05), indicating that women with greater financial strain were less likely to engage in the recommended level of heart-healthy PA.
The first logistic regression model was expanded to assess the relationships between heart-healthy behaviors as binary outcomes with demographic variables and with the covariates of CVRF knowledge and self-nurturance levels. In this second model, self-nurturance predicted smoking status (B = 1.19, P = .01) and a heart-healthy diet (B = 1.19, P = .01). Women’s nutritional intake was not predicted by age group (B = 0.90, P = .10), financial strain category (B = 0.41, P = 58), educational category (B = −0.66, P = .32), or knowledge of CVRFs (B = 0.11, P = .16). Physical activity level was not predicted by age group (B = −0.03, P = .97), educational category (B = 0.10, P = .83), or financial strain category (B = 0.79, P = .07), nor was PA level predicted by adding self-nurturance (B = −0.09, P = .79) or CVRF knowledge (B = 0.07, P = .10).
Results of this study demonstrated that this sample of premenopausal women was highly knowledgeable about CVRFs, with higher levels of knowledge than previously reported.29,30 This improved knowledge in our sample may be related to the increased media attention and national educational programs5 (eg, AHA: Go Red for Women© or National Heart, Blood and Lung Institute: National Women’s Heart Health Education Initiative and The Heart Truth) aimed at increasing women’s awareness of CVRFs.
However, these women’s CVRF knowledge did not correlate with or predict heart-healthy behaviors, possibly because of the relatively low variability in their knowledge scores. Knowledge did differ by education level and financial strain, consistent with previous reports that women’s knowledge was higher in populations with higher educational levels29,31 and lower in women experiencing more financial strain.18,19,31,32 These results suggest that heart-healthy behaviors are not related to women’s CVRF knowledge, as previously reported.33
This lack of association between CVRF knowledge and behavioral outcomes may be due to the failure of health behavior theory to account for the human dynamic.34 One indication of this dynamic has been conceptualized as an “intention-behavior gap”35; that is, individuals’ health-behavior intention to act is not translated into action. Knowledge levels may help individuals to set goals or identify steps toward reaching their goals, but it does not necessarily motivate them to change behavior. Thus, increased knowledge of CVRFs may increase women’s awareness but not heart-healthy behaviors.
The findings from this research indicate that most participants adopted heart-healthy behaviors, including not smoking, limiting alcohol intake, and engaging in some PA. In contrast, only 15% of the women described a diet that met the heart-healthy recommendations. The analyses revealed that knowledge did not predict heart-healthy behaviors. Only self-nurturance and financial strain had predictive value in the final model, with self-nurturance predicting more PA, better diet, and nonsmoking status. In addition, less financial strain predicted more PA.
The sample’s mean self-nurturance score (3.41) was consistent with those reported for registered nurses in studies of life and career satisfaction (3.5)16 and of work environment and self-nurturance (3.41).19 The analysis also found that self-nurturance was moderately correlated with heart-healthy dietary intake; that is, participants with higher self-nurturance scores were more likely to report eating a heart-healthy diet. However, no relationship was found between self-nurturance and the heart-healthy behaviors of PA or nonsmoking status. This finding was unexpected because it was anticipated that moderate to moderately high mean scores for self-nurturance would support behaviors that promote health and caring for self. It was also found that self-nurturance differed by education and financial strain; women who did not attend college and those with more financial strain had lower self-nurturance scores.
Self-nurturance is a process. In choosing to engage in a wellness behavior, individuals must first have the knowledge and rationale to act, the ability to identify self as a separate entity, and the capacity to implement the action.15 Self-nurturance differed by education and financial strain. Women who had not attended college and perceived more financial strain had lower self-nurturance scores, suggesting that self-nurturance may be socioeconomically biased concept. Further exploration of this concept is needed to determine whether it is laden with socioeconomic or cultural bias.
Financial strain predicted PA and was related to knowledge of CVRFs and self-nurturance levels. Participants with less financial strain were more likely to engage in the recommended 30 minutes of PA per day. These findings suggest that economic constraints predict less PA, consistent with previous reports that PA was predicted by education16,18 and socioeconomic status.19,20,36
Only 20% of our sample smoked some days or every day, comparable to the smoking rate for women in the same age group statewide (18.5%).26 In our study, fewer women younger than 46 years reported smoking (16.9%) than those older than 46 years (23.7%). Although not statistically significant, this finding suggests the need to assess smoking practices among all age groups.
For this study, the threshold for unhealthy alcohol intake was more than 1 alcoholic beverage per day. Because only 1 participant drank at this level (.07% prevalence), the effect of knowledge and self-nurturance on alcohol consumption could not be evaluated. This finding contrasts with statewide surveillance data for behavioral risk factors37 that 6% of the female population reported consuming at least 1 alcoholic beverage per day. This discrepancy might be due to social desirability bias, despite participants’ anonymity being ensured.
Select variables from the NWM were evaluated in this research. The variable of self-nurturance provided some predictive power. However, the model was less useful for explaining the influence of CVRF knowledge on any of the measured heart-healthy behaviors. No reports were found that used the NWM to evaluate heart-healthy behaviors. Therefore, no comparisons can be made with other studies. Further evaluation of self-nurturance as a health promotion–related concept may be useful. However, this model appears to be inadequate for explaining the complex factors that predict heart-healthy behaviors in women.
This study had several limitations. First, the study sample included mostly white women, despite attempts to actively recruit women from diverse racial or ethnic backgrounds. Second, the study sample included mostly well-educated women with minimal financial strain. Although venue-based time-space sampling22 was used to access low-income and minority women, they were more likely to decline participation in the study. The actual choice of venue for the venue-based time-space sampling may have been a limitation in accessing a heterogeneous study population. It did limit access to the underrepresented populations of Hispanics or African-Americans. Third, the reliability of the PrimeScreen nutritional intake scale was lower than anticipated in this study population, increasing the potential of random error. Finally, the model (NWM) chosen to conceptually guide this study has not been previously used to explain heart-healthy behaviors, which may partially explain the limited explanatory power of self-nurturance in this study.
Additional information that was not assessed but may have assisted in informing the researcher about other contributing factors to the women’s lack of heart-healthy behaviors was as follows: women’s height and weight to calculate their body mass index, where they obtained their healthcare (local or outside the immediate area), the economic demands beyond that measured by the financial strain measure, and the influence of any cultural bias. These factors may have given additional insight to study findings.
This study demonstrated that our sample of premenopausal women was quite knowledgeable about CVRFs. Most women participated in heart-healthy behaviors; that is, they did not smoke or drink minimal alcohol and participated in some PA. Participants also exhibited a moderate level of self-nurturance, which was moderately correlated with heart-healthy dietary intake. We found that heart-healthy behaviors were not predicted by knowledge of CVRFs. However, PA, heart-healthy diet, and nonsmoking status were predicted by self-nurturance, and PA was predicted by less financial strain.
This study is the first to examine self-nurturance as a concept for promoting heart-healthy behaviors. Results suggest that women are becoming more knowledgeable about CVRFs, but that increase is not being translated into behaviors that would sustain heart health. Further research is needed on how best to translate knowledge of CVRFs into behavior change. This study was conceptually guided by the NWM, but its key concepts (knowledge and self-nurturance) provided minimal explanatory power for heart-healthy behaviors. Only self-nurturance was related to or predicted some heart-healthy behaviors. Finally, the difference in self-nurturance by financial strain and education needs to be studied further to determine if self-nurturance is biased toward more affluent socioeconomic groups.
Implications for Practice and Research
Findings from this study suggest that interventions to promote self-nurturance improve some heart-healthy behaviors among premenopausal women. Counseling patients about risk reduction has been based on a paradigm of fear to motivate steps to prevent disease or change behavior patterns.11,35 Our results suggest that primary care providers consider a new paradigm that includes self-nurturance. They could focus on the positive aspects of behavior change to improve healthy eating, increase PA, stop smoking, and moderate drinking. Providers could also use our study results to tailor self-nurturing interventions to reduce CVRFs for women with greater financial strain and lower education levels.
Future research should explore self-nurturance as an important health-promotion concept. In addition, more studies are needed on venue-based time-space sampling, which is designed to target women from lower socioeconomic groups (based on the numbers frequenting a particular venue).22,23 However, our results suggest that this approach may be of limited value for accessing vulnerable populations, such as low-income women and minorities. Finally, studies that explore the “intention-behavior gap”9 may be useful for bridging the chasm between knowing what steps to take and implementing a recommended behavior change.
The author thanks Carol Bova, PhD, ANP-BC, for her guidance in the development and implementation of this research project.