Diabetes mellitus and functional impairment in Taiwanese older men and women

Contents lists available atScienceDirect

Archives of Gerontology and Geriatrics

Diabetes mellitus and functional impairment in Taiwanese older men and women

Chih-Hsun Wu

_{, Ching-Yu Chen}

_{*, Yin-Chang Wu}

_{, Li-Jen Weng}

_{, Hurng Baai-Shyun}

c_{Department of Family Medicine, College of Medicine, National Taiwan University, No. 1, Sec. 1, Jen-Ai Road, Taipei, 100, Taiwan} d_{Bureau of Health Promotion, Department of Health, 5F, No. 503 Sec. 2, Liming Road, Taichung City, 408, Taiwan.}

a r t i c l e i n f o

Keywords:

Type 2 diabetes mellitus Functional impairments Gender differences Taiwanese older adults

a b s t r a c t

Type 2 diabetes mellitus is strongly related to many kinds of functional impairment, even after adjusting for demographic and comorbid conditions. The current study examined sex differences in the relationships between Type 2 diabetes mellitus and functional impairment in an Asian population sample. Data were obtained from a national survey, the Social Environment and Biomarkers of Aging Study (SEBAS) in Taiwan. A total of 652 older adults aged ≥65 years were included in the study. Pearson’sc2_{-test and multiple logistic regression analysis were used to} examine the relationships between diabetes and functional impairments in older men and women. The reported numbers of impairments were significantly higher in women, in those aged ≥75 years, and in those with diabetes. There were sex and age differences in the relationships between diabetes and functional difficulties. Even after adjustment for age, education, and co-morbid conditions, men with diabetes were about four times more likely to have difficulties related to self-care, and women with diabetes were about two to three times more likely to have difficulties related to higher functioning than their non-diabetic counterparts. Sex differences should be considered when understanding the relationships between diabetes and functional impairments in older adults.

© 2010 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Several cross-sectional and longitudinal studies have found that type 2 diabetes mellitus (DM) is strongly related to many kinds of functional impairments, including problems with mobility, balance, housework, and self-care, in older adults. These relationships remained significant even after adjusting for demographic factors (e.g., age, sex, education, and ethnicity) and common diabetes-related and diabetes-undiabetes-related comorbidities (e.g., Rekeneire et al., 2003; Wray et al., 2005). Gender differences in functioning have been found in many studies involving older adults (Rahman and Liu, 2000; Liang et al., 2008). Although most of the studies focusing on diabetes and functioning partially accounted for the sex effect, only a few further explored the differences in the relationship between diabetes and functional impairment between older men and women (Gregg et al., 2000). Furthermore, it is unclear whether the results are applicable to older Asian people, because most of these studies took place in Western populations (e.g., Maty et al., 2004; Sinclair et al., 2008); Asian studies have been relatively rare (Chou and Chi, 2005).

* Corresponding author. Division of Geriatric Research, Population Health Science Institute, National Health Research Institutes, Room 440, 4F, No. 17 Xu-Zhou Road, Taipei, 100, Taiwan. Tel.: +(886-2)-3393-2198; fax: +(886-2)-2356-3260.

E-mail address: [email protected] (C-Y. Chen).

In 2006, the prevalence of diabetes in Taiwan was around 14.5% in elderly men and 13.9% in elderly women (Chen et al., 2001); people with diabetes had relatively higher mortality rates (Tseng, 2004), and diabetes was ranked the fourth cause of death (Department of Health, Taiwan, 2006). The current study aimed to assess the sex differences in functional impairment related to diabetes in elderly Taiwanese people, and to examine if these relationships remained significant after adjusting for demographic factors and comorbidities.

2. Sample and methods 2.1. Study population

The data used in this study were from the Social Environment and Biomarkers of Aging Study (SEBAS) in Taiwan, which was a random subsample of an ongoing survey, the Taiwan Longitudinal Study on Aging (TLSA). The TLSA began in 1989 with a nationally representative sample of persons 60 years and older. The SEBAS survey procedures were approved by the institutional review boards at the Bureau of Health Promotion of the Department of Health, Taiwan, Princeton University, and Georgetown University, and conformed to the principles embodied in the Declaration of Helsinki. Details of the study are stated elsewhere (Bureau of Health Promotion, Department of Health, Taiwan, 2000). Among the 1,713 randomly selected respondents, 1,497 were interviewed, and

1,023 participated in the physical examination, including physician evaluation and collection of blood and urine samples. The case selection criteria were: (1) older than 65 years old in year 2000; and (2) blood glucose data were available. A total of 652 people were included in this study.

2.2. Diabetes and other medical conditions

The information was gathered from blood tests and by trained personnel who conducted face-to-face interviews. A positive DM status was defined by: (1) fasting plasma glucose AC ≥ 126 mg/dl; or (2) self-reported DM history. The presence of other comorbid medical conditions was assessed with the question, “Do you have <disease>? or has your medical doctor told you that you have <disease>?” for the following medical conditions: high blood pressure, heart disease (not including palpitations), stroke, cataracts, cancer, respiratory problem (including bronchitis, emphysema, pneumonia, lung disease, asthma, or other lower respiratory tract diseases), arthritis or rheumatism, and hip fracture. Depressive mood was evaluated by a 10-item short version of the Center for Epidemiologic Studies Depression Scale (CES-D), and cognitive function was evaluated by selected items from the Short Portable Mental Status Questionnaire (SPMSQ).

2.3. Classification of Functional Impairment

Participants were asked about 19 distinct physical tasks in three functional domains: (1) physical function – grasping or turning things with fingers, raising both arms over the head, lifting 11–12 kg, standing continuously for 15 minutes, squatting, walking 200–300 meters, climbing 2–3 flights of stairs, getting out of bed, and doing heavy housework; (2) higher functioning (IADLs) – personal shopping, managing money, riding bus or train by oneself, doing light housework, and using the telephone; and (3) self-care task (ADLs) – bathing, dressing, eating, moving around indoors, and using the toilet. Each function was assessed by the question, “Without help from another person or special equipment, do you have any difficulty <performing the task> by yourself?”. The responses were categorized into 0 (no difficulty) or 1 (have difficulty). Participants were defined as having functional impairment in a particular domain if they reported having difficulty in any specific task in these three domains. The participants with functional impairment were further differentiated into three mutually exclusive functional impairment statuses: (1) physical function impairment only; (2) higher functioning impairment with/without physical function impairment; and (3) self-care impairment with/without any other impairment. This classification of functional impairment was adopted from two studies in the United States (Maty et al., 2004) and in Hong Kong (Chou and Chi, 2005), and these functional impairment statuses can be considered hierarchical, because a person with difficulty in progressively greater numbers of physical tasks is considered to have a more severe disability.

2.4. Analysis

DM status (present vs. absent) was used as the main independent variable. Sex and age were used as grouping variables while comparing each physical task. Age was used as a grouping variable (the “younger old” group: 65–74 years old and the “older old” group: ≥75 years old) because a previous national older population survey in Taiwan (Wu and Chang, 1997) found that people older than 75 years had significantly higher rates of functional impairment. The main effects of DM status, sex, and age groups on the reported numbers of difficulties were analyzed using the t-test. Then, any self-reported difficulties versus no

difficulties for each task were compared by Pearson’s c2_{-test for}

DM status. Furthermore, participants were classified into the three mutually exclusive functional impairment statuses by their report of disabilities. In order to assess the independent relationships of DM with each of the three functional impairment statuses in men and women, four sequential multiple logistic regression models were constructed. The first model was unadjusted, and the second model was adjusted for age and education. The third model was adjusted for age, education, and medical conditions that are generally related to diabetes (high blood pressure, heart disease, stroke, and cataracts). The fourth model was fully adjusted for age, education, depressive mood, cognitive function, and all comorbid medical conditions that had been assessed in the study. Participants with no functional impairments in any task served as the reference group in all logistic regression analyses.

3. Results

Of the 652 participants, 388 (59.5%) were men and 264 (40.5%) were female. The prevalence of diabetes was significantly lower for men than for women in the whole sample (13.9% vs. 28.4%, c2_{= 20.8, p < 0.001). However, there were no age-group differences}

in the prevalence of diabetes in either men or women. Compared to those without diabetes, men and women with diabetes were more likely to have high blood pressure (men: 51.9% vs. 36.5%; women: 50.7% vs. 36.7%). Women with diabetes were more likely to have respiratory problems (16.0% vs. 7.4%) and hip fractures (4.0% vs. 0.5%). No other significant differences were found (see Table 1). Table 1

Demographic and clinical characteristics of 652 older Taiwanese by sex and diabetes status (%) Parameters Men −DM +DM p = Women −DM +DM p = Number 334 54 189 75 Age, years 65–74 59.0 53.7 0.466 59.8 48.0 0.081 ≥75 41.0 46.3 40.2 52.0 Education, ≤6 years 60.5 50.0 0.146 62.5 80.0 0.629 Comorbid conditions High BP 36.5 51.9 0.032 36.7 50.7 0.037 Heart disease 17.4 18.5 0.836 23.3 28.4 0.389 Stroke 4.2 1.9 0.408 2.6 8.0 0.050 Cataract 33.8 37.0 0.645 40.7 48.6 0.244 Cancer 2.7 0.0 0.222 4.2 2.7 0.548 Resp. problems 17.1 14.8 0.681 7.4 16.0 0.035 Arthritis 14.7 13.0 0.740 23.3 29.3 0.306 Hip fracture 1.5 1.9 0.847 0.5 4.0 0.038 Depressive, CESD >10 13.6 18.5 0.313 19.9 30.7 0.062 SPMSQ ≤7 2.4 3.8 0.562 9.4 17.3 0.659

3.1. Impairments on physical tasks

To get an overview on the functional impairments in the older adults, the main effects of sex, age, and DM status on the reported numbers of difficulties were analyzed first. Women

reported difficulties on an average of 3.84±4.05 tasks, which

was significantly higher (t = 7.85, p < 0.001) than men (1.60±2.77).

People in the older old group reported difficulties on 3.56±4.11

tasks, which was significantly higher (t = 6.35, p < 0.001) than

Table 2

Distribution of difficulties with various tasks by sex, age group, and diabetic status DM Men, 65–74 y %a _{OR (CI)}b _{p =} Women, 65–74 y %a _{OR (CI)}b _{p =} Men, >75 y %a _{OR (CI)}b _{p =} Women, >75 y %a _{OR (CI)}b _{p =} Physical functions Grasping Yes 6.9 3.6 (0.6–20.5) 0.128 8.3 1.6 (0.4–6.8) 0.507 8.0 1.9 (0.4–10.0) 0.442 20.5 1.9 (0.7–5.5) 0.215 No 2.0 5.3 4.4 11.8

Raising arms Yes 10.3 7.5 (1.4–38.9) 0.006 8.3 1.6 (0.4–6.8) 0.507 4.0 2.8 (0.2–32.2) 0.386 20.5 3.0 (1.0–9.4) 0.050

No 1.5 5.3 1.5 7.9

Lifting 11–12 kg Yes 17.2 1.1 (0.4–3.0) 0.892 44.4 1.3 (0.6–2.7) 0.557 44.0 1.8 (0.7–4.2) 0.191 74.4 0.8 (0.3–2.1) 0.695

No 16.2 38.9 30.7 77.6

Standing 15 min Yes 20.7 4.4 (1.5–13.1) 0.004 27.8 3.0 (1.2–7.5) 0.019 16.0 1.3 (0.4–4.4) 0.623 48.7 2.1 (0.9–4.5) 0.072

No 5.6 11.5 12.4 31.6

Squatting Yes 20.7 1.0 (0.4–2.5) 0.938 47.2 1.5 (0.7–3.2) 0.283 36.0 1.2 (0.5–3.0) 0.650 74.4 2.9 (1.2–6.8) 0.012

No 21.3 37.2 31.4 50.0

Walking 200 m Yes 17.2 2.7 (0.9–8.2) 0.066 25.7 1.4 (0.6–3.5) 0.428 32.0 1.7 (0.7–4.3) 0.273 51.3 2.0 (0.9–4.4) 0.077

No 7.1 19.5 21.9 34.2

Climbing stairs Yes 24.1 2.4 (0.9–6.3) 0.065 41.7 2.4 (1.1–5.3) 0.029 28.0 1.2 (0.5–3.2) 0.677 69.2 3.1 (1.4–7.0) 0.006

No 11.7 23.0 24.1 42.1

Getting out of bed Yes 3.4 7.0 (0.4–115) 0.114 5.6 6.6 (0.6–74.9) 0.082 4.0 2.8 (0.2–32.2) 0.386 10.3 2.7 (0.6–12.9) 0.187

No 0.5 0.9 1.5 4.0

Heavy housework Yes 27.6 2.0 (0.8–4.8) 0.135 55.6 2.3 (1.1–4.9) 0.032 37.5 1.2 (0.4–2.9) 0.709 74.4 1.7 (0.7–4.0) 0.227

No 16.2 35.4 33.6 63.2

Higher functioning

Personal shopping Yes 6.9 3.6 (0.6–20.4) 0.128 13.9 4.4 (1.1–17.4) 0.023 16.0 3.1 (0.8–11.1) 0.074 30.8 1.8 (0.7–4.4) 0.186

No 2.0 3.5 5.8 19.7

Managing money Yes 0.0 NA 11.1 1.1 (0.9–1.2) 0.226 4.0 1.0 (0.9–1.1) 0.712 20.5 1.1 (0.9–1.3) 0.143

No 2.0 5.3 5.8 10.5

Traveling by bus Yes 10.3 2.2 (0.6–8.4) 0.255 41.7 2.8 (1.2–6.3) 0.011 20.0 1.5 (0.5–4.3) 0.502 61.5 1.9 (0.9–4.1) 0.116

No 5.1 20.4 14.7 46.1

Light housework Yes 6.9 2.0 (0.4–10.1) 0.390 16.7 3.6 (1.1–11.9) 0.029 20.0 2.4 (0.8–7.4) 0.124 38.5 1.9 (0.8–4.3) 0.134

No 3.6 5.3 9.5 25.0

Using telephone Yes 3.4 1.0 (0.1–8.2) 0.977 22.2 2.6 (1.0–7.2) 0.050 4.0 0.5 (0.1–4.3) 0.547 28.2 0.9 (0.4–2.1) 0.819

No 3.6 9.7 7.3 30.3 Self-care Bathing Yes 3.4 7.0 (0.4–115) 0.114 8.3 5.0 (0.8–31.4) 0.057 4.0 1.9 (0.2–18.6) 0.592 5.1 0.5 (0.1–2.7) 0.440 No 0.5 1.8 2.2 9.2 Dressing Yes 3.4 2.3 (0.2–23.0) 0.463 2.8 1.6 (0.1–18.0) 0.708 4.0 1.9 (0.2–18.6) 0.592 2.6 0.4 (0.0–3.3) 0.359 No 1.5 1.8 2.2 6.6 Eating Yes 0.0 NA 0.0 NA 0.0 NA 2.6 0.6 (0.1–6.4) 0.702 No 0.5 0.0 1.5 3.9

Moving indoors Yes 6.9 14.5 (1.3–165) 0.005 2.8 NA 4.0 5.7 (0.3–93.7) 0.173 7.7 1.5 (0.3–7.1) 0.606

No 0.5 0.0 0.7 5.3

Using toilet Yes 3.4 7.0 (0.4–115) 0.114 5.6 6.6 (0.6–74.9) 0.082 4.0 1.9 (0.2–18.6) 0.592 7.7 1.5 (0.3–7.1) 0.606

No 0.5 0.9 2.2 5.3

a _{Percentage of participants with any difficulty.}

b _{Unadjusted odds ratios (95% confidence interval) for the presence of any difficulty vs. no difficulty in participants with DM vs. without DM.} NA = not available.

Table 3

Unadjusted and adjusted associations of diabetes with any disability by disability status +DM N (%) −DM N (%) Unadjusted Model 1 OR (CI)a p = Adjusted Model 2b OR (CI)a p = Adjusted Model 3c OR (CI)a p = Adjusted Model 4d OR (CI)a p = Men

No Disability 26 (48.1) 182 (54.5) 1.0 (reference) 1.0 (reference) 1.0 (reference) 1.0 (reference) Mobility only 18 (33.3) 104 (31.1) 1.2 (0.6–2.3) 0.561 1.2 (0.6–2.3) 0.580 1.2 (0.6–2.4) 0.570 1.1 (0.5–2.2) 0.837 Higher functioninge _{5 (9.3) 39 (11.7) 0.9 (0.3–2.5) 0.835 0.9 (0.3–2.6) 0.842 1.0 (0.3–2.8) 0.934 0.7 (0.2–2.4) 0.525} Self-caref _{5 (9.3) 9 (2.7) 3.9 (1.2–12.5) 0.023 3.6 (1.1–11.8) 0.034 4.2 (1.2–14.9) 0.027 4.4 (1.1–18.6) 0.035} Disability in any category 28 (51.9) 152 (45.5) 1.3 (0.7–2.3) 0.387 1.3 (0.7–2.3) 0.415 1.3 (0.7–2.4) 0.407 1.1 (0.6–2.2) 0.736

Women

No Disability 10 (13.3) 48 (25.4) 1.0 (reference) 1.0 (reference) 1.0 (reference) 1.0 (reference) Mobility only 17 (22.7) 73 (38.6) 1.1 (0.5–2.6) 0.800 1.1 (0.4–2.7) 0.852 1.0 (0.4–2.6) 0.962 0.8 (0.3–2.0) 0.629 Higher functioninge _{39 (52.0) 58 (30.7) 3.2 (1.5–7.1) 0.004 3.3 (1.4–7.7) 0.006 3.0 (1.2–7.3) 0.014 2.1 (0.9–4.9) 0.086} Self-caref _{9 (12.0) 10 (5.3) 4.3 (1.4–13.4) 0.011 4.6 (1.4–15.3) 0.013 3.5 (1.0–12.2) 0.055 2.1 (0.6–7.8) 0.257} Disability in any category 65 (86.7) 141 (74.6) 2.2 (1.1–4.6) 0.036 2.0 (0.9–4.5) 0.078 1.9 (0.8–4.2) 0.135 1.7 (0.7–4.2) 0.248 a _{Odds ratio (95% confidence interval) for presence of any difficulty vs. no difficulty in participants with DM vs. without DM.} b_{Model 2 adjusted for age and education}

c_{Model 3 adjusted for age, education, and medical conditions potentially related to DM (high blood pressure, heart disease, stroke, and cataracts).} d_{Model 4 adjusted} for age, education, DM-related comorbidities, other medical conditions (cancer, respiratory problem, arthritis or rheumatism, and hip fracture), depressive mood (CESD), and cognitive function (SPMSQ). e_{Higher functioning disability with/without mobility disability.} f _{Self-care disability with/without any other disabilities.}

reported difficulties, on average, on 4.03±4.37 tasks, which was

significantly higher (t = 4.65, p < 0.001) than for those without diabetes (2.13±3.17).

Further explorations on each physical task found that men in the younger old group with diabetes were significantly more likely to report functional impairment in three tasks (raising both arms over the head, standing continuously for 15 minutes, and moving in the house) than those without diabetes. Women in the younger old group with diabetes were significantly more likely to report functional impairment in six tasks (standing continuously for 15 minutes, climbing 2–3 flights of stairs, doing heavy housework, personal shopping, riding bus or train by oneself, and doing light housework) than those without diabetes. Older old women with diabetes were significantly more likely to report functional impairment in two tasks (squatting and climbing 2–3 flights of stairs) than those without diabetes. However, there were no significant differences in the probability of reporting functional impairment between older men with or without diabetes in any of the 19 tasks (Table 2).

3.2. Diabetes and different functional impairment statuses

In general, 51.9% of men with diabetes and 45.5% without diabetes reported functional impairment in at least one task. Men with diabetes were significantly more likely to report functional impairment related to “self-care impairment with/without any other impairment” than those without diabetes (9.3% vs. 2.7%). The higher probability was still significant after adjustment for age and education (second model), for medical conditions generally related to diabetes (third model), and after full adjustment (fourth model). However, there were no significant differences between men with and without diabetes in other functional impairment statuses (Table 3).

Women with diabetes were significantly more likely to report functional impairment related to “higher functioning impairment with/without physical function impairment” (52.0% vs. 30.7%) and to “self-care impairment with/without any other impairment” (12.0% vs. 5.3%) than those without diabetes. The probability of reporting impairment in at least one task was significantly higher in women with diabetes than in those without (86.7% vs. 74.6%). After

adjusting for age and education (second model), the significant association of diabetes with “higher functioning impairment” and with “self-care impairment” remained, while the association with “impairment in any task” was attenuated (p = 0.078). With further adjustment for the four medical conditions that generally related to diabetes (third model), the significant association of diabetes with “higher functioning impairment” still remained, but the association with “self-care impairment” was attenuated (p = 0.055), and the association with “impairment in any task” became insignificant. After full adjustment (fourth model), the significant association of diabetes with “higher functioning impairment” was attenuated (p = 0.086), and the association with self-care impairment became insignificant.

4. Discussion

The present research shows that sex and age differences should be considered when understanding the relationships between diabetes and functional impairments in older adults. Women in the younger old group with diabetes were at higher risk for functional impairment than their counterparts without diabetes in three of nine physical function tasks and in three of five higher functioning tasks. However, women in the older old group with diabetes were only at higher risk in two of the nine physical function tasks. The reason for this might be the increasing rate of functional impairments in those older old women without diabetes. For instance, functional impairment was reported in 35.4% of the younger old women without diabetes, but increased to 63.2% in the older old women group for the heavy housework task. A similar pattern was also seen in men; younger old men with diabetes were at higher risk in three tasks, while no significant differences were found in any tasks in the older old men group. These results indicate that having diabetes might be a stronger risk factor for functional impairment around 65 to 74 years, especially in women.

Sex differences were found after the analysis of functional impairment status. The strongest association between diabetes and functional impairment was found in the “self-care impairment with/without any other impairment” status in men, and in the “higher functioning impairment with/without physical function impairment” status in women. These associations were not

substantially reduced by the most common vascular complications of diabetes (heart disease, stroke, and cataracts) and one highly co-occurring chronic condition (high blood pressure).

Compared to findings of Maty et al. (2004) in older women, we found no significant difference in the “physical function only” status, and the significant difference in the “self-care impairment with/without any other impairment” was attenuated after adjusted for diabetes-related medical conditions in Taiwanese old women. However, our results in old women were similar to the findings in Hong Kong old adults (Chou and Chi, 2005), although their study did not classify patients by sex. We do not know if these differences among Eastern and Western studies are caused by cultural differences. In a review paper (Chia et al., 1997) concerning attitudes toward women in Taiwan, it was pointed out that Taiwanese men and women were less liberal and egalitarian than Americans. This might imply that the traditional gender roles, such as the most important “job” for women is to serve their husband and raise their children, was more acceptable in Taiwan than in the USA. Thus, it may be reasonable to assume that older Taiwanese women reporting difficulties with higher functioning tasks would be culturally more acceptable than reporting problems with self-care related tasks. We do not know if these hypotheses are true in older American populations. A longitudinal study (Murtagh and Hubert, 2004) that took place in the United States also found that older women tended to report “a greater degree of disability, particular among IADL categories”. Further studies are needed to clarify if there really are cultural or ethnic differences.

Besides providing data relating to Asian populations, the other major strengths of the current study are that it: (1) used a nationally representative sample of Taiwanese older people; (2) had blood glucose data to identify people with or without diabetes more objectively; (3) focused on sex and age differences to further explore the relationships between diabetes and functional impairments. However, like all cross-sectional studies, our study cannot provide enough information to make any inferences over time. Since the TLSA is an ongoing survey, further explorations of related issues may be possible in the near future.

5. Conclusion

Based on data from an Eastern population, Taiwan, the current study found gender and age differences in the relationship between of diabetes and functional impairment. Even after adjustment for age, education, and co-morbidities, men with diabetes were about four times more likely to have difficulties related to self-care impairment; while women with diabetes were about two to three times as likely to have difficulties related to higher functioning impairment, than their non-diabetic counterparts.

Thus, it is suggested that gender differences should be considered in understanding the relationships between diabetes and functional impairment in older adults.

Conflict of interest statement

None.

Acknowledgements

The descriptions or conclusions herein do not represent the viewpoint of the Bureau of Health Promotion, Department of Health, Taiwan.

References

Bureau of Health Promotion, Department of Health, Taiwan, 2000. Results from the social environment and biomarkers of aging study (SEBAS) survey report 2000. Chen, K-T., Chen, C-J., Gregg, E.W., Engelgau, M.M., Narayan, K.M.V., 2001. Prevalence

of type 2 diabetes mellitus in Taiwan, ethnic variation and risk factors. Diabetes Res. Clin. Pract. 51, 59–66.

Chia, R-C., Allred, L-J., Jerzak, P-A., 1997. Attitudes toward women in Taiwan and China: current status, problems, and suggestions for future research. Psychol. Women Q. 21, 137–150.

Chou, K-L., Chi, I., 2005. Functional disability related to diabetes mellitus in older Hong Kong Chinese adults. Gerontology 51, 334–339.

Department of Health, Taiwan, 2006. Health statistics in Taiwan 2006. Part II, Statistic on cause of death. Online at: http://www.doh.gov.tw/ufile/doc/ chapter2.pdf. (Last accessed on July, 30, 2009).

Gregg, E.W., Beckles, G.L., Williamson, D.F., Leveille, S.G., Langlois, J.A., Engel-gau, M.M., Narayan, K.M., 2000. Diabetes and physical disability among older U.S. adults. Diabetes Care 23, 1272–1277.

Liang, J., Bennett, J.M., Shaw, B.A., Qui ˜nones, A.R., Ye, W., Xu, X., Ofstedal, M.B., 2008. Gender differences in functional status in middle and older age: are there any age variations? J. Gerontol. B: Psychol. Sci. Soc. Sci. 63, S282–S289.

Maty, S.C., Fried, L.P., Volpato, S., Williamson, J., Brancati, F.L., Blaum, C.S., 2004. Patterns of disability related to diabetes mellitus in older women. J. Gerontol. A: Biol. Sci. Med. Sci. 59, M148–M153.

Murtagh, K.N., Hubert, H.B., 2004. Gender differences in physical disability among an elderly cohort. Am. J. Public Health 64, 1406–1411.

Rahman, M.O., Liu, J-H., 2000. Gender differences in functioning for older adults in rural Bangladesh. The impact of differential reporting? J. Gerontol. A: Biol. Sci. Med. Sci. 55, M28–M33.

Rekeneire, N., Resnick, H.E., Schwartz, A.V., 2003. Diabetes is associated with subclinical functional limitation in nondisabled older individuals. Diabetes Care 26, 3257–3263.

Sinclair, A.J., Conroy, S.P., Bayer, A.J., 2008. Impact of diabetes in physical function in older people. Diabetes Care 31, 233–235.

Tseng, C.H., 2004. Mortality and causes of death in a national sample of diabetic patients in Taiwan. Diabetes Care 27, 1605–1609.

Wray, L.A., Ofstedal, M.B., Langa, K.M., Blaum, C.S., 2005. The effect of diabetes on disability in middle-aged and older adults. J. Gerontol. A: Biol. Sci. Med. Sci. 60, 1206–1211.

Wu, S.C., Chang, M.C., 1997. Health care for the elderly in Taiwan: A fact book. Taipei, Taiwan: Taiwan Institute of Family Planning.