Abstract
Background:
Men differ from women in their cardiovascular mortality and morbidity globally, possibly due to their varying dietary and lifestyle behavior and usage of medications in dyslipidemia control.
Objectives:
The study aimed to compare the gender differences among community-dwelling Asians in their dietary habits, physical activity, their perception and use of lipid-lowering medications towards achieving their LDL-Cholesterol goals.
Methods:
A cross-sectional study was conducted, which included patients with physician-diagnosed dyslipidemia, aged 31–80 years in two public primary care clinics in Singapore. They were administered a questionnaire survey on their diet, exercise and lipid-lowering medication. Their latest laboratory fasting lipid tests, retrieved from their electronic health records, defined their treatment goals based on adapted Framingham risk scores.
Results:
Amongst 1093 patients, 61.9% were female. Fewer men (65.9%) achieved LDL-C treatment goals compared with women (75.2%). Comparing with their counterpart, more women were willing to change their diet (75.8% vs. 46.2%) and exercise (78.4% vs. 70.9%) to achieve their LDL treatment goals. However, men who were treated with lipid medication were more likely to reach LDL-C treatment goals.
Conclusion:
Fewer Asian men in Singapore attained their LDL-C treatment goals than women, which appeared to be associated with greater reluctance to embark on dietary and lifestyle changes. Gender-specific interventions should be considered to address these differences.
Background
Dyslipidemia underpins the development of atherosclerotic vascular diseases and results in major causes of mortality in developed countries globally. 1 In dyslipidemia, persistently elevated serum Low Density Lipoprotein-Cholesterol (LDL-C) is the prime risk for cardio- and cerebrovascular diseases, which often top the cause of death in developed nations. It is managed largely by lifestyle modifications such as dietary control, 2 increased physical activities, 3 and pharmacotherapy. 4
However, gender differences in lipid profiles have been reported in population-based studies. 5 Kolovou et al. reported that women had higher High Density Lipoprotein-Cholesterol (HDL-C) and Total Cholesterol (TC), but lower Triglyceride (TG) and lower TC/HDL-C ratio compared with men. 5 The shorter life expectancy of men compared with women could be attributed to their less favorable lipid profiles. 6
The gender differences in their lipid characteristics may in turn be influenced by variations in their respective lifestyle habits, including dietary and exercise patterns, smoking, alcohol intake and uptake of lipid-lowering therapy (LLT). 7 Early studies alluded that men tended to have lower intakes of fruits, vegetables, dietary fiber, lower intakes of fat and devoted less attention to healthy food choice and healthy eating compared with women. 7 A population-based study in Watari district in Japan revealed that men demonstrated a significantly more unhealthy lifestyle such as late dinner, skipping breakfast, smoking, heavy alcohol drinking or lack of regular exercise, across all age groups compared with women. 8 On the other hand, Rodriguez et al., in their study of an American managed care population, revealed that treatment-naïve women were less likely to commence LLT than men, regardless of their LDL-C levels. 9 They also showed that women initially treated with LLT were significantly less likely to continue with their LLT or to receive LLT dose titration to achieve optimal LDL-C control level. 9 Thus gender differences exist in their lifestyle behavior and attitude towards LLT in ethnically distinct populations.
Some 61% of the world population lives in the Asia Pacific region alone. 10 Globally there are more men than women, with average gender ratio of 102 men to 100 women, and an even higher ratio in South-Central, West and East Asia. 10 In contrast, the ratio is lower at 99 in South-east Asia. Singapore, at the center of this region, has an Asian population with an even lower ratio of 965 males per 1000 females in 2015. 11 Ischemic heart and cerebrovascular diseases remain the top five causes of their mortality in 2014. 12 High total blood cholesterol was more common in males (18.3%) than in females (16.5%) amongst its multi-ethnic Asian citizens who have inherent Eastern social values but receive a Western style of education, and have been extensively exposed to East–West mix-culture. 13 Singapore is thus an ideal microcosm to provide an insight into the lifestyle behavior and factors associated with the gender differences in their dyslipidemia status.
This study primarily aimed to determine the gender differences in lifestyle behavior, such as dietary pattern, lifestyle habits and LLT, which would influence dyslipidemia control status. Understanding these gender variations on this important metabolic disease will yield invaluable information which may be relevant to other rapidly Westernized Asian communities, and facilitate the development of gender-specific interventions to reduce the impact of this major cardiovascular risk factor.
Methods
Subjects
The cross-sectional study was conducted in public primary care clinics (polyclinics) from October 2013 to September 2014. The recruited patients had physician-diagnosed dyslipidemia in the past 6 months prior to study enrolment, as verified from their electronic health records. They included patients of Chinese, Malay or Indian ethnic origins, aged between 31 and 80 years, and were Singaporeans or Permanent Residents.
Patients with cognitive, hearing and/or visual impairment, who were incapable of providing written informed consent, were excluded.
Study sites
Primary healthcare services in Singapore are delivered by both public polyclinics and private general practitioner clinics. Due to the provision of healthcare subsidies and privileges, 45% of the local Asian population consult polyclinics for the management of their chronic diseases, including dyslipidemia. 14
The study was carried out in two typical adjacent polyclinics in Tampines and Pasir Ris housing estates in north-eastern Singapore. As the clinical practice sites of the investigators, both polyclinics served about 258,810 multi-ethnic Asian residents (updated as in June 2014), who are socio-demographically similar to the rest of the Singapore population.
Recruitment of subjects
Two trained research assistants (RAs) were assigned to screen patients for eligibility criteria for study enrolment, one at each study site. They approached potential patients in the consultation waiting areas and enquired if their medical conditions included lipid disorders; this was then verified using their electronic health records. Once the eligibility criteria were fulfilled, the RA explained the study using a participant information sheet (approved by the institution review board), and clarified doubts and queries before obtaining the patient’s written informed consent. Participation in the study was voluntary. This recruitment method was carried out one subject after another, in a consecutive case-encounter fashion.
The RA administered the questionnaire to consented patients face to face, retrieved their clinical and laboratory data from their electronic medical records and subsequently entered the data into a spreadsheet. An independent data management officer audited the data and rectified any errors before handing over the dataset to the biostatistician.
Questionnaire
The questionnaire was assembled by the investigators as there was no suitable instrument to meet the objectives of the original study. The questionnaire covered demographic characteristics, patients’ perceptions on diet, exercise and medication. The laboratory investigation reports were retrieved from the electronic health records. The questionnaire was pilot-tested on several subjects and staff; any poor congruency was rectified before finalizing the eventual questionnaire for use in this study.
In this sub-study, the investigators defined the following variables according to their respective domains.
Sample size calculation
For a conservative sample size, the proportion of those who reach their lipid goals is assumed to be 0.5, hence the sample size required was 380 based on a precision of 5% and 95% confidence level. A disproportionate sample was collected to ensure that smaller ethnic groups were not underrepresented while comparing the inter-ethnic group differences. The sample size was increased threefold for in-depth analysis by each ethnic group, reaching a target 1140 patients for the study.
Definition of outcome (LDL-C treatment goals)
The outcome of the study is the proportion of males and females reaching LDL-C goals. LDL-C treatment goals are defined using the following criteria of the three risk groups, adapted and simplified from the Framingham coronary heart disease risk score.
High risk (HR)—patients with dyslipidemia and any of the following existing vascular diseases (Type II Diabetes, Ischemic Heart Disease, Stroke, Renal Disease, Peripheral Vascular Disease)
Medium (Intermediate) risk (MR)—patients with dyslipidemia and hypertension without vascular complications
Low risk (LR)—Patients with dyslipidemia without other co-morbidities
According to the local Ministry of Health Clinical Practice Guidelines for the management of lipids disorders (2006), the recommended LDL-C goal level for the High Risk Group is <100 mg/dl (2.6 mmol/l), <130 mg/dl (3.4 mmol/l) for the Intermediate Risk Group, and <160 mg/dl (4.1 mmol/l) for the Low Risk Group. 15
Statistical analysis
Demographics of the study population were reported as frequency and percentage for categorical variables, while mean and standard deviation were reported for continuous variables. Missing data were excluded from the analysis, if any. The outcome LDL goal achievement and other categorical variables were tested using Chi-square test or Fisher’s Exact test for categorical variables and independent t-test for continuous variables. Factors on demographics, perception, exercise and medication with a p<0.2 were entered into a logistic regression to identify the adjusted odds of factors associated with lipid control. A p-value of <0.05 was considered statistically significant. All analyses were performed using SPSS version 22.0.
Ethics approval and consent to participate
The study received approval from the SingHealth Centralized Institutional Review Board (CIRB reference 2013/630/E). Informed consent was obtained from all study participants. Participation in the study was voluntary.
Results
The final study population comprised 1093 patients, after excluding 47 patients who did not satisfy eligibility criteria or with incomplete laboratory data. Men constituted 38.1% of the study population (Table 1). Men (65.9%) were less likely to achieve LDL-C treatment goal compared with women (75.2%). More of the older women, who were aged 60 years and above, achieved their LDL goals compared with younger women (79.0% vs. 70.5%).
Demographic characteristic, co-morbidities and lipid profiles of the study population in association with their LDL-C treatment goals a .
Low-Density Lipoprotein-Cholesterol (LDL-C) levels.
Goals achieved refers to LDL ≤160 mg/dl for the low-risk group, LDL ≤130 mg/dl for the medium-risk group, LDL ≤100 mg/dl for the high-risk group.
Goals not achieved refers to LDL >160 mg/dl for the low-risk group, LDL >130 mg/dl for the medium-risk group, LDL >100 mg/dl for the high-risk group.
Men of Chinese ethnicity (75.8%) were significantly more likely to achieve LDL-C treatment goals, compared with those of other ethnic groups (62.6% and 58.3% for Malays and Indians, respectively). Men with lower educational level, i.e. secondary education or below (68.6% vs. 56.4%) were more likely to achieve LDL-C control.
Table 2 shows that patients of either gender currently on LLT were more likely to achieve better LDL-C control. However, more women were willing to change diet (75.8% vs. 46.2%) and to exercise at least 150 minutes per week (78.4% vs. 70.9%) than men. Amongst women, those who perceived LLT to be expensive were less likely to achieve their LDL treatment goal.
Gender differences in dietary habits, exercise profiles and perceptions to lipid-lowering therapy in association with their LDL-C treatment goals. a
Low-Density Lipoprotein-Cholesterol (LDL-C) levels.
Goals achieved refers to LDL ≤160 mg/dl for the low-risk group, LDL ≤130 mg/dl for the medium-risk group, LDL ≤100 mg/dl for the high-risk group.
Goals not achieved refers to LDL >160 mg/dl for the low-risk group, LDL >130 mg/dl for the medium-risk group, LDL >100 mg/dl for the high-risk group.
Logistic regression analyses in Table 3 and 4 indicate that ethnicity was associated with LDL-C goal attainment for both genders. Men of Chinese ethnicity were significantly more likely to control their LDL-C, compared with Malay (OR=0.49, 95% CI=0.28–0.85) and Indian (OR=0.43, 95% CI=0.26–0.73) men. Women of Chinese ethnic group were more likely to achieve LDL-C control compared with Malay women (OR=0.52, 95% CI=0.33–0.82). Men with secondary education or lower were significantly more likely to achieve their LDL goals (OR=1.80, 95% CI=1.08–3.00) than men who had higher education.
Factors associated with LDL-C treatment goals amongst men using logistic regression.
Factors associated with LDL-C treatment goals amongst women using logistic regression.
Discussion
The multi-ethnic Asian population in Singapore has the fourth longest longevity in the world, with the life expectancy of men at 80 years and women reaching 85 years. 16 The local National Health Surveys in 1992, 1998, 2004 and 2010 revealed that women consistently had more favorable lipid profiles compared with men. From 2004 to 2010, their mean LDL-C declined from 3.4 to 3.2mmol/l but mean HDL remained at 1.5mmol/l. 17 Unlike the general population in these national surveys, this study population with dyslipidemia had higher risks of atherosclerotic vascular diseases.
The results showed gender differences in demographic, clinical status, lifestyle behavior and LLT uptake and perceptions, which provided an insight into the less favorable lipid profiles of men. The LDL goal achievement in this study showed differences in ethnic groups, which had been reported by several earlier studies which alluded to the ethnic differences in cardiovascular diseases.18,19 Lee et al. also found that Asian Indians were at greatest risk of coronary heart disease, compared with Chinese and Malays. 19
Fewer men with higher education achieved their LDL-C treatment goals. These men would have easier access to the internet, where biased information could have adversely affected their attitude towards medical treatment of dyslipidemia. Such a hypothesis requires further evaluation, in which a qualitative research method would be preferable to gather in-depth perspectives of this observation.
Likewise, Asian men in this study were noted to be less willing to change their diet to control cholesterol compared with women. Women could be more weight and health-conscious, and with a higher proportion of them being home-makers (but classified as “unemployed”), they were in a better position to make decision on home cooking and food choices for themselves. As more men were employees, their occupation often required them to eat food prepared by vendors during their work, where food options may be limited and restrict their selection of healthy food choices.
Besides diet, women were more likely to achieve their LDL treatment goals if they exercised more than 150 minutes per week. A review article also suggested that a combined intervention of physical activity alongside pharmacological treatment facilitated the attainment of favorable cardiovascular health outcomes. 20 Men were more likely to have more restricted hours for their leisure and sport participation due to their employment. However, work–life balance varies between individuals and women could be similarly affected due to family and work commitments. Further controlled study would be needed to test the effect of exercise on LDL treatment goals for the two genders, with consideration given to their biological and lifestyle differences. Moreover, this study did not monitor their actual physical activity as the data collected were based solely on self-reported information.
Unlike many Western populations, the prevalence of smoking and alcoholism amongst women were much lower. Compared with women, there were far more male smokers. No female alcoholic drinker was recruited in this study. Smoking is known to interfere with the effectiveness of LLT and is a surrogate indicator of an unhealthy lifestyle. Excessive alcohol intake would raise the serum TG levels of the consumers. Nonetheless, there has been rising prevalence of female smokers over the years. 17 More effective gender-centric public education programs are needed to retard or even reverse this unhealthy trend and await evaluation in future research.
Over 90% of the study subjects were on LLT. Both men and women had better lipid profiles than the general population, regardless of their gender variations in diet and their perceptions of the lifestyle measures of the disease. The high prevalent use of LLT reflects on the access and quality of the local fee-for-service primary healthcare service. The polyclinics provided subsidized consultation fees and medications, including LLT such as statins, to the general population. Similarly, the MONICA study in northern Sweden has shown that the increasing use of LLT from 1994 to 2014 has benefited both genders by narrowing the gap of their lipid profiles, with overall reduction of their cholesterol levels. 21
Our study showed the cost of the medication was also associated with the achievement of LDL-C goals. During the recruitment phase of this study, atorvastatin was not a subsidized medication. Hence, some patients, especially women, reported that the lipid-lowering medication was costly. Such perception could lead to reduced adherence, and hindered their attainment of LDL-C treatment goals. Treatment expenditure is a potential hurdle which could interfere with optimal dyslipidemia management in a fee-for-service primary healthcare system. Physicians should proactively raise the treatment cost as an issue for discussion during consultation and make provisions for affected patients to obtain financial assistance whenever necessary to overcome this barrier.
Limitations
This study recruited fewer men than women, as proportionately fewer men sought consultation for chronic diseases in the polyclinics. Furthermore, the original objective of the study was to determine ethnic differences in achievement of optimal dyslipidemia management goals. 22 In addition, the study was conducted at two polyclinics, where healthcare services are subsidized to any walk-in patients. This differs from the private general practitioner practices, where subsidized care at only selected clinics is accorded to those who qualify after means testing of their household income. For these reasons, caution has to be exercised in generalization of the results, although the populations in these two estates are not dissimilar to other residents located in other regions of the island. The self-reported method in obtaining the data could be subjected to recall bias, but such an approach is widely used in routine clinical practice in view of the time and resource constraints. Enhanced methodology will be deployed in future studies to determine the answers to several of the observations in this study.
Conclusion
Gender differences in lipid status were found amongst patients with dyslipidemia, associated with fewer men achieving their LDL-C treatment goals in the context of less favorable dietary habits. The majority of the study population were on LLT, which appeared to optimize their dyslipidemia management. Whilst gender differences were noted in their attitude and perceptions, exercise or physical activity did not seem to influence their attainment of LDL-C treatment goals.
Footnotes
Acknowledgements
The authors are grateful to the Singapore Heart Foundation for the sponsorship of this Lipid HEALTH study and appreciate the assistance of Ng CE and Xu HY in executing the questionnaire survey, as well as the generous support rendered by the polyclinic staff to optimize subject recruitment at the two study sites. Due to confidentiality issues, individual level data is not available in public repository. However, interested researchers may write in for data request. TNC designed the study protocol, GSCP, GCC, KKH were involved in supervision of during subject recruitment and ensuring smooth execution of the study. KYLE was involved in data analysis and interpretation of data. TNC drafted the manuscript while all authors reviewed and approved the final manuscript before submission.
Declaration of Conflicting Interests
None declared.
Funding
This work was supported by the Singapore Heart Foundation [grant number PRG2012/07].