Increased medical costs in elders with the metabolic syndrome are most evident with hospitalization of men
Abstract
Background: Little is known about health care costs associated with the metabolic syndrome (MetS). We have assessed annualized healthcare costs and health outcomes for both genders in different health care settings among representative Taiwanese elders.
Methods: The Nutrition and Health Survey in Taiwan (1999-2000) provided 1378 individuals aged 65 years or older with known MetS status. NAHSIT files were linked to National Health Insurance records (1999-2006). T-tests and multiple regression models were used to assess expenditures in total and in six services: inpatient, ambulatory care, dental care, traditional Chinese medicine, emergency and contracted pharmacy. The Cox model was used to assess gender effect on all-cause mortality and cardiovascular disease (CVD)
mortality while logistic regression was used for that on CVD hospitalization. The five MetS component costs were evaluated by multiple regressions.
Results: The MetS affected 29% of men and 48% of women. After full adjustment, those with MetS had 1.30 (95% Confidence Interval [CI] 1.11-1.52), men had 1.43 (95% CI 1.20-1.70) and women had 1.19 (95% CI 0.93-1.52) times higher costs than those without MetS. Over no MetS, MetS costs were increased 2.94 (95% CI 1.23-7.10)-fold for inpatient care and 1.30 (95% CI 1.12-1.52)-fold for ambulatory care for men, while ambulatory MetS costs were 1.28 (95% CI 1.05-1.57)-fold for women. The MetS was associated with higher risk of CVD hospitalization in men (adjusted odds ratio [OR] = 1.76, 95% CI 1.20-2.58) but not in women (OR=1.08, 95% CI 0.67-1.75). Among those with the MetS, all-cause and CVD mortality were comparable between men and women. Of the MetS components, low HDL-cholesterol had the greatest impact on costs, more so in men (2.23-fold) than women (1.58-1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
fold)
Conclusions: With the MetS, service costs were greater overall, significantly for men, but not women, evident for men with hospitalization, but not women. At the same time,
cardiovascular and all-cause mortalities were not significantly different by gender in regards to the MetS in Taiwanese elders.
Key words: health care costs, gender, metabolic syndrome, HDL-cholesterol, ethics 26 27 28 29 30 31
Introduction
The metabolic syndrome (MetS) is a cluster of health risk factors which are each and together associated with an increased incidence of a range of chronic diseases like diabetes,
cardiovascular and cerebrovascular disease, and, consequently, all-cause and cardiovascular mortality in a variety of socio-economic settings.1-3 In a US NHANES adult cohort, MetS prevalence was 35% for NCEP and 39% for IDF definitions.2,4 A European study of over-30 year-old adults indicated that 27-35% of men and 20%-34% of women had the MetS.5 In Taiwan, the incidence of MetS during a 4-year follow up of subjects who received health check-ups was 12.7% (17.5% for men and 8.3% for women) according to NCEP criteria; 6 and a nationwide survey showed the age-standardized prevalence of MetS to be 15.7% by the modified NCEP criteria, and 14.3% by the IDF criteria.7 Taiwan has an increasing prevalence and incidence of the MetS in an ageing population,6 but with gender differences dependent on menopause8 and obstructive sleep apnoea9 with consequences for cardiovascular disease, both coronary heart disease and stroke10 and all-cause and cardiovascular mortality, the latter with the higher population attributable risks (PAR).8
There is concern about the costs of the MetS with cardiometabolic risk factor clusters costing some $80 billion a year in the US11 and excess attributable annual medical costs ranging from $2000-$6000 per person.11-15 Incremental medical costs during a ten-year period, which result from the development of diabetes, acute myocardial infarction (MI) or stroke, in an adult cohort with the MetS, have been estimated to be US $6200, $2200 and $1300.16 These costs would be tripled, 25 times, and almost 30 times, respectively, if these events were present at baseline. In addition, the MetS leads to extra economic costs due to productivity loss among working populations.17,18 It, therefore, has global disease relevance with an associated economic burden.
There are few studies which explore gender in the occurrence and consequential morbidity 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
and mortality of the MetS.19-20 Studies examining the risk of premature death or fatal events caused by the MetS in women and men have been conflicting,21, 22 especially in the elderly,21, 23-26 which requires an appreciation of the PAR by gender as well.8 Little is known about the related health care costs. Relevant to the MetS, some studies indicated that elderly women are less frequently treated than their male counterparts with congestive heart disease,
hypertension and diabetes.27-30 In available studies of health costs for the MetS, gender has generally been treated as a confounder and not developed the gender pathway in its own right which might link determinants to the MetS and its sequelae, Thus, the attendant utilization and costs of health care and health consequences, have been overlooked.
In a cohort of elders, we have sought to identify any gender differences in the MetS and related costs in settings which can be identified through the universal health care system in Taiwan. We seek to answer three questions: firstly, whether the MetS imposes medical costs both in men and women over and above those without the MetS (non-MetS); secondly, whether there are cost differences between genders with the MetS similar to the non-MetS; thirdly, what the MetS attributable costs are for each gender. In addition, we document differences in health care utilization31-34 and in health outcomes, particularly cardiovascular disease (CVD)28,35-41 and all-cause mortality,8, 21-25 which may relate to gender disparity.42
Methods
Study population
The Elderly Nutrition and Health Survey in Taiwan, a nationally representative sample of the elderly aged 65 and over in 1999 and 2000 formed the cohort. The detailed description of this survey can be found elsewhere.43 A questionnaire was administered, a health assessment conducted, and fasting blood taken from 1422 elders. We excluded 42 persons with missing values and 2 who had equivocal death status, leaving 1378 eligible. These were linked to the 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
National Health Insurance (NHI) claim database. Death was ascertained from the National Death Registry between 1999 and 2008. All subjects signed informed consent. The ethics committees of the National Health Research Institutes and of Academia Sinica approved the study.
The metabolic syndrome criteria
Participants who met at least three of the following revised-NCEP criteria6 were considered to have the MetS: 1) waist circumference (WC) 90 cm for men and 80 cm for women, 2) fasting triglycerides >150 mg/dL, 3) blood pressure ≥130/85 mmHg or current use of antihypertensive medication, 4) HDL cholesterol <40 mg/dL for men and <50 mg/dL for women or current use of medication, and 5) fasting glucose (FG) >100 mg/dL or current use of diabetic medication.
Medical expenditure ascertainment
The NHI claim data for the participants were collected between their interview date and Dec 31st 2006, as the endpoint for follow up, except for date of death before Dec 31 2006 or the date when the person withdrew from NHI before Dec 31, 2006. Payments were summed from NHI claims for each participant with a yearly discount rate at 3% and with yearly core Consumer Price Indices adjustment. The annualized costs, as 1999 New Taiwan Dollars (NTD), were generated by summed costs divided by observational years, periods between interview date and the end of follow-up, death date, or date of withdraw from NHI coverage. Annual utilization rates for six areas of health care and their categorical costs were
calculated. These were inpatient care, ambulatory care, dental care, traditional Chinese medicine, emergency room and contracted pharmacy (for approved prescription drugs). 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106
Statistical analysis
Data were expressed as frequencies and percentages for categorical variables and means ± SD for continuous variables. Chi-square test and t-test were used for assessing differences in covariates among participants with statistical significance as p<0.05. In addition, to answer the first and second question, log-linear regression models were employed because of highly right-skewed distributions of medical costs.44 In log-linear regressions, one dollar was
assigned to those who did not use any health care so to avoid missing values in logarithmic transformations. The coefficients, when exponentiated, estimate the ratio of the annualized costs of two compared counterpart groups.
A ratio of one indicates no cost discrepancy between MetS status for the first question, or between men and women for the second. To explicitly examine gender difference of the effects on costs due to the MetS, an interaction term between the MetS status and gender was added in the models based on overall sample. Furthermore, we performed gender-specific log-linear models, based on men or women sub-samples only, to ascertain differential effects of the MetS on costs by gender. To satisfactorily answer the second question, we used log-linear models to compare annualized costs between men and women within MetS
participants. These log-linear models were developed for annualized total costs and six categorical medical costs. Covariates for adjustments included demographics, socio-economic variables, health behaviours, modified Charlson Comorbidity Index scores (according to the 1999 NHI claims) and physician-to-population ratio of residence as an indicator of access to health care. The significant level for cost ratio reference to 1.00 was taken as p<0.05.
To answer the third question, we calculated MetS attributable costs in genders using the recycled prediction approach.45 Here, attributable cost was the mean difference between the costs, assuming, on the one hand, that everyone experienced the MetS and, on the other, that 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131
the same individuals did not experience the MetS. We estimated these costs according to models which were corresponded and developed beforehand. For prevention and clinical resource management, the added medical costs when a pre-existent MetS component
develops into the complete MetS may be ascertained. We approached this through evaluation of the total cost ratios for the MetS to the non-MetS participants, based on five non-exclusive subgroups where participants were each progressively restricted to one of five MetS
components. Consequently, five auxiliary models for these subgroups in each gender were assessed. The non-MetS group was the reference in these models.
To understand gender differences in health outcomes of the MetS, logistic regression was used to assess occurrence of hospital admissions due to cardiovascular disease (CVD). We carried out survival analysis and used Cox models to assess gender effects on
cardiovascular disease (CVD) and all-cause and mortality among MetS participants. Cases with CVD admission or causes of death were defined by ICD-9 codes 390-459. The same covariates as before were used as adjustments. SAS statistical software (version 9.1.3) was used for all analyses and SUDAAN (version 10.0) was used to adjust for the design effect of sampling. Statistical significance in these regressions was considered as p<0.05.
Results
For Taiwanese elders, MetS prevalence was 29% in men and 48% in women. It was more in younger elderly men and men with relatively higher education and so were former smokers in women. (Table 1)
Table 2 provides the values for the 5 components of the Mets and BMI by gender and MetS status. Significant higher values for FG, TG, BP and WC and lower values for HDL-C indicate the nature of the MetS representing a cluster of cardiometabolic risk factors in both genders. Values for TG, SBP and HDL-C were higher and for DBP and WC lower in women, 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
whether MetS or not. Women and men have a similar level of FG and BMI when they develop the MetS.
Table 3 shows the annual utilization and costs (in NTD) by MetS and gender. Compared to the non-MetS, elders with the MetS tended to access more ambulatory visits, both for men (28.4 vs 24.8) and women (30.2 vs 25.1), more hospital days in men (12.5 vs 9.3) but less hospital days in women (5.9 vs 6.7). Nearly NTD 18700 more medical costs (~USD 623) were incurred by MetS elders than by the non-MetS in one year (p<0.01). A man with the MetS consumed NTD 106000 (~USD 3600) per year compared to NTD 79600 (~USD 2750) by a non-MetS person and cost difference is NTD 26300(p<0.001). In women, annual costs for the MetS and the non-MetS were NTD 76300 and 54300, respectively; the cost difference is NTD 22000. In every heath care category, there was more expenditure in MetS patients regardless of gender. Inpatient and ambulatory care represented the greatest NHI medical costs. An additional NTD 16600 for inpatient (p<0.05) and NTD 8600 for ambulatory care (p=0.001) were used by men with MetS, with a corresponding NTD 14800 and 6700 (p<0.01) for women. The excess costs in ambulatory care were significant for both genders (p<0.01). Within MetS participants, men significantly received more health care than did women regarding hospital admission (gender difference=0.35 times/per year, p<0.01) and hospital days (6.64 days/per year, p<0.05) but less in pharmacy services (-1.4, p<0.05). In general, there were more medical costs spent in men than women among MetS participants. The cost differences achieve significance in total costs (NTD 29600, p=0.001), inpatient care (NTD 23300, p=0.002), ambulatory care (NTD 5700, p<0.05) and emergency visit (NTD 400, p=0.018).
Multivariable analysis revealed cost ratios of the MetS to the non-MetS (Table 4). The MetS were 1.30 times the non-MetS in total costs (p<0.01), 1.73 times in inpatient costs (p<0.05) and 1.30 times in ambulatory care costs (p<0.001). The MetS and gender interaction 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181
was significant for total costs, which indicates that a gender difference might exist in medical costs due to the MetS. In men, MetS expenditure reached 1.43 times the non-MetS ones in total costs (p<0.001). A cost ratio for inpatient costs reached 2.94 (p<0.05) in men whereas only 1.17 in women. In ambulatory care, the ratios were comparable for men (1.30, p<0.01) and women (1.28, p<0.05). Among MetS participants, there were more costs in women than men in ambulatory care, dental care, traditional Chinese medications and pharmacy services, but less in inpatient care and ER utilisation, though all of these measures were not significant. Based on these cost models, the costs attributable to the MetS in each year were NTD 15800 for men and NTD 5900 for women (Figure 1), with a difference of NTD 9900 (p<0.001). The greatest difference between genders was found in inpatient care: the MetS created NTD 9400 additional cost for men whereas it did only NTD 700 for women; the cost difference is NTD 8800 (p<0.001). In ambulatory care, attributable cost was NTD 4900 in men and NTD 4300 in women. The cost difference between genders was NTD 600 (p<0.01).
Survival analysis shows that women with the MetS had less risk than men with MetS in all-cause mortality (adjusted Hazard Ratio, HR=0.70, 95% confidence interval (C.I.): 0.44-1.11) and CVD mortality (adjusted HR=0.48, 95% C.I.:0.20-1.14) (Table 5). The MetS were associated with increased risk of CVD -caused hospitalization in men (adjusted OR=1.76, 95% C.I.=1.20-2.58) while it was not in women (adjusted OR=1.08, 95% C.I.=0.67-1.75). The adjusted odds ratio (OR) of CVD hospitalization was 0.49 (95% C.I.=0.26-0.92) in women with MetS compared to men with MetS.
Figure 2 shows the gender-specific total cost ratios of MetS to non-MetS for subgroups with different MetS components, namely central obesity (defined as WC), elevated fasting glucose (FG), elevated blood pressure (BP), dyslipidemia (triglycerides TG), and
dyslipidemia HDL-Cholesterol (HDL-C), to reflect the impact of a particular MetS component on medical expenditure. In men, MetS-FG raises the ratio to 1.45 (p<0.001), 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206
MetS-BP to the ratio of 1.46 (p<0.001) and MetS-HDL-C to the ratio of 2.23 (p<0.001), which were all higher than the general cost ratio, 1.43, estimated for all participant men (Table 4), In women, the most cost-accentuating component was HDL-C (1.58, p<0.05) whereas the general cost ratio was 1.19.
Discussion
Gender differences in the costs of the metabolic syndrome
We found that those with the MetS consume more annualized heath care resources in a universal health insurance system than their MetS-free counterparts, but principally in ambulatory care and inpatient settings. Elderly men use more additional health care resources than do elderly women when they develop the MetS. Among health care settings, the MetS leads to 28%-30% incremental costs in ambulatory in men and women, respectively, while inpatient care costs are almost tripled in men (Table 4). This resulted in more than twice attributable costs of the MetS in elderly men than those in women (Figure 1). Among MetS participants, however, total medical costs allocated to men were greater than women (Table 4). Analysis based on CVD hospital admissions and mortalities showed that women with the MetS may receive less CVD health care, but their health outcomes were not significantly different, even better than their counterparts in men (Table 5).
In general, women have been found to use more health care than men.31-34, 46 However, the gender gap decreases along with increased age or even reverses in elders.34,46 Women may receive less treatment for circulatory diseases. Studies in different settings have showed that, compared to men, women have less drugs dispensed while hospitalized,36-37,47 longer time to wait cardiac examination,41 less diagnostic examinations,38,39 are less managed with
percutaneous transluminal coronary angioplastyor coronary artery bypass surgery.27,47 A study of acute MI also found similar results when severity was taken into account.35 In the 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231
current study, there was no increase in use of inpatient care among MetS women while there was among MetS men (Table 4). Among those with the MetS, hospitalized women had 50% less spending than hospitalized men in angioplasty and PTCA and 40% less in overall CVD treatments (Figure 3), though overall inpatient care costs did not differ significantly between men and women. These findings are similar to those in previous CVD studies.
Health care utilisation for CVD treatment have been found to be less for women than for men, in part because of higher risk of aggressive cardiovascular interventions such as PTCA and bypass, possibly because women may be older than men when interventions are
considered,48 have smaller arteries,49 or more comorbidities.50-52 On the other hand, women’s maybe more likely to be interpreted as emotional problems.53 We found that development of the MetS did not increase rates of hospitalization for CVD in women (adjusted OR=1.08, 95%C.I.:0.67-1.75), while it did in men (adjusted OR=1.76, 95%C.I.:1.20-2.58) Overall, among those with MetS, less women than men received treatments for CVD and therefore less relevant medical costs were spent. However, women with the MetS might have benefitted from lower all-cause and CVD mortality risk. Accordingly, fewer inpatient care resources would have been allocated to women with the MetS than to men who did, but with no gender difference in health outcomes, which suggests that less use of health care resources among women was not necessarily inappropriate.40 This is consistent with previous studies which explored the relationship between interventions for heart disease and health outcomes.
When we examined the annual hospitalization costs for the principal cardiometabolic diseases other than coronary heart disease (CHD) in those with the MetS, namely cancer, diabetes mellitus, kidney disease, hypertension stroke, and ‘other’, there were apparent gender differentials in different directions. But the only significant gender differentials were for stroke and ‘other’, with men the gender with the greater costs (Figure 4). Women with hypertension appeared to consume more hospital costs for hypertension than men, but this 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256
was not significant (Figure 4). Nevertheless, Basu et al45 found that women generally have more medical expenditure associated with hypertension than men. Cutler and colleagues54 reported that hypertensive women had better awareness and more treatment than hypertensive men. Both studies were representative non-institutionalized US populations and found gender differences to be reduced or reversed in old age.
Competing risks between different diseases might be another explanation for the gender differences in costs which we have identified. For example, we identified the top 20 causes for hospital admissions for men and women in our study and found that, among MetS women, there were relatively lower proportions with pneumonia and influenza (<1%) and COPD (4%) compared with non-MetS women being 10% and 6% respectively (data not shown). In men, these two rates were nearly equivalent. It is possible that the MetS might be associated with less respiratory illness in older women, although its significance for immune function and inflammatory disease is not settled.55
It is plausible that the apparent relative dominance of lowered HDL-C in the
determination of health system costs among men and women with the MetS relates to the consequences of these risk factors for the costly outcome of multi-system disease (Figure 2). This will include a range of complications of both diabetes and hypertension, more so when these risks (as fasting glucose and BP in men) occur together and probably synergistically. The hypertension cost factor in women, in the presence of lower HDL-cholesterol (usually higher in women than men) might be expected to favour adverse cardiovascular outcomes, with their attendant cost. Nevertheless, women may not realise the benefits of this knowledge or its use in management for gender-discriminatory reasons (see above).
Recent studies have indicated that the metabolic syndrome does not necessarily predict greater mortality in later life.23-26 This was our finding as well; and the gender differences these same mortalities were also not significant once adjustments were made. However, the 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281
costs of the underlying causes of death as they relate to the MetS are generally higher for men. Presumably, with the MetS, men’s ultimate relative parity with women in achieving similar survival is a reflection of the fiscal investment in their health, but we cannot be sure. We need to explore further the basis of this relatively greater cost of men to the health care system when they have the MetS. One clue we have is that the fasting glucose, BP and HDL-C components of the MetS account for a good deal of the excess cost in men when they develop the MetS. Each of these components is amenable to changes in personal behaviours, especially cigarette smoking, physical inactivity and nutritionally unsatisfactory diets as reported elsewhere for this cohort.56-57 It may be possible to be selectively targeted given the heterogeneity of the MetS in this population6 , but, with advancing years it is probably far more practical, affordable and sustainable to recruit several tactics in modest and manageable ways.
Limitations
Our study has only measured the MetS at baseline. Those without the MetS at baseline might have developed it later. If they did, they may have incurred costs similar to the base-line entrants, so reducing the differential between affected and non-affected individuals. Our qualitative conclusions would not be changed. The corollary might be that those regarded as having the MetS at baseline, might have recovered from it – again reducing the differential between the MetS affected and non-affected patient costs.
Ours is an elderly population and the cohort study lasted for 8 years, during which various age-related health phenomena may have been operative and confounded our findings in ways for which we have been unable to adjust. Inadequate coverage of the Taiwanese population and the services provided by the NHI system is unlikely, since there is currently an enrolment which exceeds 98% of the population. Our findings ought, therefore, be 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306
generalisable to Taiwanese elderly at large, provided the limits of characterising demographic sub-groups are acknowledged. Perhaps more important is how relevant the findings are to elders in comparable universal health care systems; although we cannot be sure, we consider that, in comparable systems, awareness of the impact of age, gender and the MetS on costs for ambulatory and inpatient care which we have identified should help health planners. We observe that all costs calculated were based on claims to NHI from contract health care providers. This should have avoided bias which occurs with self-reports or from individual or household interview questionnaires.
We present limited information as to how differences in costs relate to health outcomes. Moreover, our analysis is in regard to annualised costs. Although those for women are less than men, the differential is largely determined by inpatient costs which tend to occur
towards the end of life, and may be displaced to this period with longevity. Of interest, in our study, even with longer life expectancies, women with the MetS do not have an aggregate use of resources greater than men (data not shown). But the health economic and policy questions raised require more active investigation.
Conclusions
Elders with the MetS, compared with those who do not, have increased health care costs which are incurred in a universal health care system in Taiwan. Men with the MetS use more health care resources on an annualised basis, especially through hospitalization. There were no gender differences in health outcomes between women and men with the metabolic syndrome in later life.
Conflict of Interest
No author had a conflict of interest. 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331
Figure 1: Attributable cost to the MetS by gender. § denotes a significant gender difference of costs examined by t-test, p<0.001.
Figure 2: Annual cost ratios of MetS to non-MetS within 5 component subgroups estimated by multivariable log-linear regression. Models adjusted for age, marital status, work status, education, ethnicity, perceived financial status, exercise, smoking, alcohol drinking, CCI scores, regions and physician-population ratio. *p<0·05, ** p<0·01, ***p<0·001.
Figure 3: Annualized hospitalization costs of MetS individuals for CVD stent and other CVD procedures.
Figure 4: Annualized hospitalization costs for MetS individuals with an underlying cause of admission in cancer, diabetes (DM), kidney disease (KD), hypertension (HT), stroke or others. T-test for costs in gender differences. *p<0.05 and **p<0.01.
332 333 334 335 336 337 338 339 340 341 342 343
References
1. Gami AS, Witt BJ, Howard DE, et al. Metabolic Syndrome and Risk of Incident Cardiovascular Events and Death: a Systematic Review and Meta-Analysis of Longitudinal Studies. J Am Coll Cardiol. 2007;49(4):403-414.
2. Wahlqvist ML, Chang HY, Hsu CC, et al. Is impaired energy regulation the core of the metabolic syndrome in various ethnic groups of the USA and Taiwan? BMC Endocr Disord. 2010;10
3. Wu SH, Liu Z, Ho SC. Metabolic syndrome and all-cause mortality: a meta-analysis of prospective cohort studies. Eur J Epidemiol. 2010;25, 375-384.
4. Ford ES. Prevalence of the metabolic syndrome defined by the International Diabetes Federation among adults in the U.S. Diabetes Care. 2005;28:2745-2749.
5. Qiao Q. Comparison of different definitions of the metabolic syndrome in relation to cardiovascular mortality in European men and women. Diabetologia. 2006;49:12: 2837-2846.
6. Yang FY, Wahlqvist ML, Lee MS. Body mass index (BMI) as a major factor in the incidence of the metabolic syndrome and its constituents in unaffected Taiwanese from 1998 to 2002. Asia Pac J Clin Nutr. 2008;17:339-351.
7. Hwang LC, Bai CH, Chen CJ. Prevalence of obesity and metabolic syndrome in Taiwan. J Formos Med Assoc. 2006; 105(8): 626-635.
8. Wang WS, Wahlqvist ML, Hsu CC, Chang HY, Chang WC, Chen CC. Age- and gender-specific population attributable risks of metabolic disorders on all-cause and cardiovascu-lar mortality in Taiwan. BMC Public Health. 2012;12:111.
9. Ting H, Huang RJ, Lo HS, et al. Vanished gender differences of cardiometabolic risk fac-tors after matching the apnea hypopnea index at postmenopausal age. Gender Med. 2012;9: 9-20. 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369
10. Chien KL, Hsu HC, Sung FC, Su TC, Chen MF, Lee YT. Metabolic syndrome as a risk factor for coronary heart disease and stroke: an 11-year prospective cohort in Taiwan community. Atherosclerosis. 2007;194(1):214-221.
11. Sullivan PW, Ghushchyan V, Wyatt HR, et al. The medical cost of cardiometabolic risk factor clusters in the United States. Obesity. 2007;15(12):3150-3158.
12. Boudreau D, Malone D, Raebel, M, et al. Health care utilization and costs by metabolic Syndrome Risk Factors. Metab Syndr Relat D. 2009; 7:305-314.
13. Fitch K, Pyenson B, Iwasaki K. Metabolic syndrome and employer sponsored medical benefits: an actuarial analysis. Value Health. 2007;10(1).
14. Curtis LH, Hammill BG, Bethel MA, et al. Costs of the metabolic syndrome in elderly individuals: Findings from the Cardiovascular Health Study. Diabetes Care. 2007; 30(10): 2553-2558.
15. Beaton SJ, Robinson SB, Worley AV, et al. Cardiometabolic risk and health care utilization and cost for Hispanic and Non-Hispanic Women. Popul Health Manag. 2009;12(4): 177-183.
16. Caro J, Brien J, O JA, et al. Economic burden and risk of cardiovascular disease and diabetes in patients with different cardiometabolic risk profiles. Value Health. 2007;10: S12-S20.
17. Schultz AB, Edington DW. The Association between changes in metabolic syndrome and changes in cost in a workplace population. J Occup Environ Med. 2009; 51(7): 771-779. 18. Schultz AB, Edington DW. Analysis of the association between metabolic syndrome and
disease in a workplace population over time. Value Health. 2010;13(2): 258-264. 19. Regitz-Zagrosek V, Lehmkuhl E, Mahmoodzadeh S. Gender aspects of the role of the
metabolic syndrome as a risk factor for cardiovascular disease. Gender Med. 2007;4: 162-177. 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394
20. Njelekela MA, Mpembeni R, Muhihi A, et al. Gender-related differences in the
prevalence of cardiovascular disease risk factors and their correlates in urban Tanzania. BMC Cardiovasc Disord. 2009;9.
21. Forti P, Pirazzoli GL, Maltoni B, et al. Metabolic syndrome and all-cause mortality in older men and women. Eur J Clin Invest. 2012.
22. Sidorenkov O, Nilssen O, Grjibovski AM.. Metabolic syndrome in Russian adults: associated factors and mortality from cardiovascular diseases and all causes. BMC Public Health. 2010;10.
23. Salminen M, Kuoppamäki M, Vahlberg T, Räihä I, Irjala, K, Kivelä, S. The metabolic syndrome defined by modified International Diabetes Federation criteria and mortality: A 9-year follow-up of the aged in Finland. Diabetes Metab. 2010;36:437-442.
24. Petersen JL, Yow E, Aljaroudi W, Shaw LK, Goyal A, Mcguire DK, et al. Metabolic syndrome is not associated with increased mortality or cardiovascular risk in nondiabetic patients with a new diagnosis of coronary artery disease. Circulation. Cardiovascular Quality and Outcomes.2010;3:165-172.
25. Thomas F, Pannier B, Benetos A, Vischer UM. The impact of the metabolic syndrome - But not of hypertension - On all-cause mortality disappears in the elderly. J Hypertens. 2011;29, 663-668.
26. Hildrum B, Mykletun A, Dahl AA, Midthjell K. Metabolic syndrome and risk of mortal-ity in middle-aged versus elderly individuals: The Nord-Trøndelag Health Study
(HUNT). Diabetologia. 2009;52, 583-590.
27. Sheppard R, Behlouli H, Richard H, Pilote L. Effect of gender on treatment, resource uti-lization, and outcomes in congestive heart failure in Quebec, Canada. Am J Cardiol. 2005;95: 955-959. 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418
28. Gu Q, Burt VL, Paulose-Ram R, Dillon CF. Gender differences in hypertension treat-ment, drug utilization patterns, and blood pressure control among US adults with hyper-tension: Data from the National Health and Nutrition Examination Survey 1999-2004. Am J Hypertens. 2008;21: 789-798.
29. Daugherty SL, Masoudi FA, Ho PM, et al. Age-dependent gender differences in hyper-tension management. J Hypertens. 2011;29(5): 1005-1011.
30. Vimalananda VG, Miller DR, Palnati M, Christiansen CL, Fincke BG. Gender disparities in lipid-lowering therapy among veterans with diabetes. Women Health Iss. 2011;21: 4: S176-S181.
31. Briscoe ME. Why do people go to the doctor? Sex differences in the correlates of GP consultation. Soc Sci Med. 1982; 25:507-513.
32. Green CA, Pope CR. Gender, psychosocial factors and the use of medical services: a longitudinal analysis. Soc Sci Med. 1999; 48:1363-1372.
33. Ladwig KH, Marten-Mittag B, Formanek B, et al. Gender differences of symptom reporting and medical health care utilization in the German population. Eur J Epidemiol. 2000;16:511-518.
34. Redondo-Sendino A, Guallar-Castillón P, Banegas JR, et al. Gender differences in the uti-lization of health-care services among the older adult population of Spain. BMC Public Health. 2006;6:155.
35. Weitzman S, Cooper L, Charmbless L, et al. Gender, racial, and geographic differences in the performance of cardiac diagnostic and therapeutic procedures for hospitalized acute myocardial infarction in four states. Am J Cardiol. 1997;79(6):722-726.
36. Wilkinson P, Laji K, Ranjadayalan K, Parsons L, Timmis AD. Acute myocardial infarction in women: survival analysis in first six months. BMJ Clinical Research Ed. 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442
1994; 309: 566-569.
37. Clarke KW, Gray D, Keating NA, Hampton JR. Do women with acute myocardial infarction receive the same treatment as men? BMJ Clinical Research Ed. 1994; 309: 563-566.
38. Crilly M, Bundred P, Hu X, Leckey L, Johnstone F. Gender differences in the clinical management of patients with angina pectoris: a cross-sectional survey in primary care. BMC Health Serv Res. 2007;7:142.
39. Tobin JN, Wassertheil-Smoller S, Wexler JP, Steingart RM, Budner N, Lense L, Wachspress J. Sex bias in considering coronary bypass surgery. Annals of Internal Medicine. 1987;107:19-25.
40. Vaccarino V. Ischemic heart disease in women: many questions, few facts. Circ Cardiovasc Qual Outcomes. 2010; 3(2):111-115.
41. Heston TF, Lewis LM. Gender Bias in the evaluation and management of acute nontraumatic chest pain. Fam Practice Res J. 1992;12: 383-389.
42. Milcent C, Dormont B, Durand-Zaleski I, Steg PG. Gender differences in hospital mortal-ity and use of percutaneous coronary intervention in acute myocardial infarction: mi-crosimulation analysis of the 1999 nationwide French hospitals database. Circulation. 2007;115: 833-839.
43. Pan WH, Hung YT, Shaw NS, et al. Elderly Nutrition and Health Survey in Taiwan (1999-2000): research design, methodology and content. Asia Pac J Clin Nutr. 2005;14:203–210.
44. Diehr P, Yanez D, Ash A, et al. Methods for analyzing health care utilization and costs. Annu Rev Public Health. 1999; 20:125-144.
45. Basu R, Franzini L, Lairson DR, et al. Gender disparities in medical expenditures attributable to hypertension in the United States. Women's Health Iss. 2010; 20(2): 114-443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467
125.
46. Verbrugge LM. Gender and health: an update on hypotheses and evidence. J Health Soc Behav. 1985;26:156-182.
47. Chandra NC, Ziegelstein RC, Rogers WJ, et al. Observations of the treatment of women in the United States with myocardial infarction: a report from the National Registry of Myocardial Infarction-I. Arch Intern Med. 1998;158:981-988.
48. Fransoo RR, Martens PJ, Prior HJ, Burland E, Château D, Katz A. Age difference ex-plains gender difference in cardiac intervention rates after acute myocardial infarction. Healthc Policy. 2010;6, 88-103.
49. Ramström J, Lund O, Cadavid E, Thuren J, Oxelbark S, Henze A. Multiarterial coronary artery bypass grafting with special reference to small vessel disease and results in women. European Heart Journal. 1993;14, 634-639.
50. Weintraub WS, Wenger NK, Jones EL, Craver JM, Guyton RA. Changing clinical char-acteristics of coronary surgery patients. Differences between men and women. Circula-tion. 1993;88:79-86.
51. O'Connor GT, Morton JR, Diehl MJ, Olmstead EM, Coffin LH, Levy DG, et al. Differ-ences between men and women in hospital mortality associated with coronary artery by-pass graft surgery. The Northern New England Cardiovascular Disease Study Group. Cir-culation. 1993;88:2104-2110.
52. Hannan EL, Bernard HR, Kilburn HC Jr, O'donnell JF. Gender differences in mortality rates for coronary artery bypass surgery. Am Heart J. 1992;123:866-872.
53. King KB, Clark PC, & Hicks GL Jr. Patterns of referral and recovery in women and men undergoing coronary artery bypass grafting. Am J Cardiol. 1992;69:179-182.
468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490
54.Cutler JA, Sorlie PD, Wolz M, Thom T, Fields LE, Roccella EJ. Trends in hypertension prevalence, awareness, treatment, and control rates in United States adults between 1988-1994 and 1999-2004. Hypertension. 2008;52(5):818-827.
55. Kim MH, Lee HS, Park HJ, Kim WY. Risk factors associated with metabolic syndrome in Korean elderly. Ann Nutr Metab. 2007;51:533-540.
56. Lee MS, Su HH, Lee MZ, Wahlqvist ML, Huang YC. A simple food quality index pre-dicts mortality in Elderly Taiwanese. J Nutr Health Aging.2011;15: 815-821.
57. Lee MS, Chen RCY, Chang YH, Huang YC, Wahlqvist ML. Physical function mitigates the adverse effects of being thin on mortality in a free-living older Taiwanese cohort. J Nutr Health Aging.2012 (accepted).
491 492 493 494 495 496 497 498 499 500
