Acta Diabetol (2015) 52:701–708 DOI 10.1007/s00592-014-0711-y
ORIGINAL ARTICLE
Association between body mass index and diabetic retinopathy in Chinese patients with type 2 diabetes Jun Lu • Xuhong Hou • Lei Zhang • Fusong Jiang • Cheng Hu • Yuqian Bao Weiping Jia
•
Received: 12 October 2014 / Accepted: 30 December 2014 / Published online: 22 January 2015 Ó Springer-Verlag Italia 2015
Abstract Aims To explore the factors mediating the relationship between body mass index (BMI) and diabetic retinopathy (DR) in Chinese type 2 diabetes patients. Methods This is a cross-sectional study. Data of 2,533 patients with type 2 diabetes were studied from the Shanghai Diabetes Registry Database. DR was assessed using non-mydriatic fundus photography and graded as non-DR, mild–moderate (DR I–II), and sight-threatening (DR III–IV). BMI (kg/m2) was classified as normal weight (18.5 B BMI \ 25), overweight (25 B BMI \ 30), and obese (BMI C 30). b cell function was evaluated by fasting C-peptide (FCP). Results DR was present in 701 (27.7 %) patients. Patients with DR had lower BMI (24.3 vs. 24.9 kg/m2, P = 0.001) and fasting C-peptide (1.46 vs. 1.86 ng/ml, P \ 0.001) than those without DR. The association between BMI (2 kg/m2 interval) and DR was U-shaped; patients with BMI 28–29.9 kg/m2 had the lowest DR rate. Compared with normal weight, overweight was associated with reduced risk of any DR [odds ratio (OR) 0.73], DR I–II
Managed by Antonio Secchi. Jun Lu and Xuhong Hou have contributed equally to this article.
Electronic supplementary material The online version of this article (doi:10.1007/s00592-014-0711-y) contains supplementary material, which is available to authorized users. J. Lu X. Hou L. Zhang F. Jiang C. Hu Y. Bao W. Jia (&) Department of Endocrinology and Metabolism, Shanghai Diabetes Institute, Shanghai Clinical Center for Diabetes, Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, 600 Yishan Road, Shanghai 200233, China e-mail: [email protected]
(OR 0.76), and DR III–IV (OR 0.64) after adjustment for sex, age at diabetes diagnosis, and duration of diabetes. This negative association attenuated after adjustment for other confounders and became nonsignificant after further adjustment for FCP. Patients with different BMI categories had similar DR risk when stratified by FCP tertiles. Conclusion Overweight patients have lower DR prevalence than normal weight individuals, which may be attributable to better b cell function in overweight patients. Keywords Type 2 diabetes Diabetic retinopathy Body mass index Fasting C-peptide
Introduction Overweight and obesity have been established as risk factors for diabetes [1]. A national survey reported that [60 % of diabetic patients in China are overweight or obese [2]. Traditionally, higher body mass index (BMI) is a risk factor for nephropathy and cardiovascular disease morbidity and mortality among type 2 diabetes patients [3– 5]. However, recent studies have reported that being overweight may confer a protective effect against all-cause mortality and on the life expectancy of cardiovascular disease in type 2 diabetes patients [6, 7]; this has been termed the obesity paradox [7]. Diabetic retinopathy (DR) is characterized as a microvascular complication of diabetes and is always accompanied with other macro- and microvascular complications of diabetes [8–10]. Studies also demonstrated that DR was the leading cause of new cases of blindness among adults aged 20–74 years [11] and severely influenced the quality of life in diabetic patients [12]. Previous studies have indicated that DR is more prevalent in patients at an earlier diagnosis
123
702
age of diabetes, with longer duration of diabetes, poorer b cell function and glycemic control, dyslipidemia, and hypertension [11, 13–20]. However, the association between BMI and DR is inconsistent. Earlier studies suggested that higher BMI increases DR risk [17, 18, 21–23], and there was higher DR prevalence in obese patients compared with normal weight patients [24]; other studies reported no association [15, 19, 25]. Recently, lower DR prevalence was reported in patients with higher BMI [26– 28]. It is not known whether this obesity paradox exists between BMI and DR among Chinese type 2 diabetes patients. We aimed to investigate the association between BMI and DR and to explore the factors mediating the relationship between BMI and DR in Chinese type 2 diabetes patients.
Subjects and methods Trial design and study population This was a retrospective study using the Shanghai Diabetes Registry (SDR) database, which was established in December 2001 at the Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, the Shanghai Clinical Center for Diabetes [29]. This computerized database was set up with the aim of evaluating the outcome of Chinese patients with diabetes. Data of demographics (including lifestyle, disease history, family history of disease, smoking and drinking habits), anthropometry, diagnoses, laboratory measurements and drug prescriptions was collected from the medical records. Current smokers were defined as those who had smoked C1 cigarette/day for at least 1 year. Current drinkers were defined as those who had consumed C30 g of alcohol/week on average for at least 1 year. The duration of diabetes was calculated as the time of non-mydriatic fundus photography performed minus the time of diagnosis of diabetes. Inclusion criteria of this study were as follows: complete data of demographics, laboratory measurements, and fundus photography; islet autoantibody (antibodies to glutamic acid decarboxylase and protein tyrosine phosphatase IA-2)negative; and BMI C 18.5 kg/m2. Patients with underweight were excluded due to a small sample size (n = 75). And a total of 2,533 patients with adult-onset diabetes between March 2003 and December 2012 were enrolled in this study. Anthropometric and laboratory investigation Anthropometric and biochemical measurements were taken at admission. The anthropometric indices of height and
123
Acta Diabetol (2015) 52:701–708
weight were measured while the patients were barefoot and in light clothing. The BMI was calculated as the weight divided by height squared (kg/m2). Blood pressure was measured twice using a mercury sphygmomanometer when subjects were supine, and then averaged. Mean blood pressure (mmHg) was calculated by the formula: (SBP ? 2 9 DBP)/3. After a minimum 10-h overnight fast, venous blood samples were drawn at 0, 30, and 120 min following the ingestion of a diabetic diet prescribed according to the criteria proposed by the Chinese Diabetes Society [30]. Hemoglobin A1c (HbA1c), lipid profiles, and fasting plasma glucose and fasting C-peptide were measured using fasting blood sample. Blood samples collected at 30 and 120 min were used to test postprandial C-Peptide. HbA1c was measured using high-performance liquid chromatography (HLC-73G7; Tosoh, Tokyo, Japan); fasting plasma glucose was measured using the glucose oxidase method (Roche, Mannheim, Germany); lipid profiles were determined using an auto-analyzer (Hitachi 7600, Tokyo, Japan); C-peptide (CP) was measured using radioimmunoassay (Linco Research, St Charles, MO, USA). All laboratory measurements met the Shanghai center for Clinical Laboratory criteria. Definitions of hypertension and overweight/obesity Hypertension was defined as blood pressure (BP) C 140/ 90 mmHg or usage of antihypertensive medications. BMI (kg/m2) was categorized using World Health Organization standards [31]: normal weight, 18.5 B BMI \ 25; overweight, 25 B BMI \ 30; and obese, BMI C 30. Assessment of diabetic retinopathy Fundus photography was performed for each patient following a standardized protocol. The participants were seated in a darkened room, and the posterior pole of each eye was photographed with a 45° 6.3-megapixel digital non-mydriatic camera (Canon CR6-45NM, Lake Success, New York, USA). Each fundus photography was graded by two independent ophthalmologists. If there was a disagreement, a third experienced ophthalmologist would check the results. DR was graded according to the 2003 standards proposed by the American Academy of Ophthalmology (AAO) [32]: no apparent retinopathy (DR 0), no abnormalities; mild nonproliferative DR (DR I), microaneurysms only; moderate nonproliferative DR (DR II), more than just microaneurysms but less than severe nonproliferative DR; severe nonproliferative DR (DR III), any of the following: [20 intraretinal hemorrhages in each of the four quadrants, definite venous beading in 2? quadrants, prominent intraretinal microvascular abnormalities in 1? quadrant and no signs of proliferative retinopathy;
Acta Diabetol (2015) 52:701–708
703
proliferative DR (DR IV), one or more of the following: neovascularization and vitreous/preretinal hemorrhage. Patients were grouped according to DR severity: non-DR (DR 0), mild–moderate retinopathy (DR I–II), and sightthreatening retinopathy (DR III–IV). Statistical analysis The distribution of continuous variables was assessed using the Kolmogorov–Smirnov test. Non-normally distributed continuous variables are expressed as median (interquartile range, IQR) and categorical variables as frequency (%). Differences in medians were examined using the Mann–Whitney U test between two groups and Kruskal–Wallis test among three groups; differences in proportions were tested using the Chi-square test (Table 1). The association between BMI and any DR was
assessed using multivariable binary logistic regression, and the association of BMI with different degrees of DR (DR I–II and DR III–IV) was assessed using multinomial logistic regression; odds ratios (ORs) and P values are presented (Tables 2, 3; Figs. 1, 2). Adjustment variables were first tested by collinearity diagnosis according to the following criteria: variance inflation factor (VIF) [10 or tolerance near 0.1; condition index [30; and variance proportions [50 %. Then, the variables including sex, age at diabetes diagnosis, duration of diabetes, mean blood pressure, HbA1c, TG, HDL-C, smoking status, drinking status, and insulin therapy were finally selected as adjustment variables for the multiple logistic regression analyses. Tests for trend were performed using the BMI categories and FCP tertiles as ordinal variables in the corresponding logistic regression models. All statistical analyses were performed using SPSS 12.0 (SPSS Inc.,
Table 1 Comparison of clinical characteristics in patients with and without diabetic retinopathy Variables
Non-DR (n = 1,832)
DR I–II (n = 511)
Gender (male)
1,101 (60.1)
295 (57.7)
DR III–IV (n = 190)
P value
95 (50)
0.010
DR (n = 701) 390 (55.6)
P value DR versus non-DR 0.041
Age (years)
57 (49–64)
57 (51–65)
59.5 (53–66)
0.003
58 (51–65)
0.002
Age at diagnosis of diabetes (years)
51 (43–58)
48 (41–54)
49 (41–54)
\0.001
48 (41–54)
\0.001
10 (5–14)
11 (6–15)
\0.001
10 (5–14)
\0.001
Duration of diabetes (years) 2
BMI (kg/m ) WC (cm) Abdominal obesity
4 (0.6–10) 24.9 (22.7–27.6) 89 (83–96)
24.3 (22.3–27.0) 88 (82–96)
24.4 (22.0–26.5)
0.002
87 (82–93)
0.024
80 (45.5)
0.012
943 (54.7)
239 (51.3)
SBP (mmHg)
130 (120–140)
130 (120–140)
140 (125–150)
DBP (mmHg)
80 (70–86)
80 (75–90)
80 (75–90)
24.3 (22.2–26.8) 88 (82–95) 319 (49.7)
0.001 0.037 0.029
Blood pressure \0.001 0.047
130 (120–147) 80 (75–90)
\0.001 0.021
100 (93–110)
\0.001
987 (54.4)
301 (59.4)
131 (69.3)
\0.001
8.6 (7.1–10.6)
9.2 (7.5–10.7)
9.3 (7.6–11.0)
0.007
9.2 (7.5–10.8)
0.003
FPG (mmol/L)
7.91 (6.43–10.0)
7.90 (6.32–10.5)
8.68 (6.83–11.4)
0.020
8.05 (6.45–10.9)
0.089
FCP (ng/ml) 30-min CP (ng/ml)
1.86 (1.20–2.63) 2.63 (1.65–3.90)
1.52 (0.93–2.15) 2.04 (1.26–3.09)
1.24 (0.79–2.03) \0.001 1.74 (1.09–2.60) \0.001
Mean BP (mmHg) HTN (%) HbA1c (%)
97 (90–103)
97 (93–107)
97 (93–107) 432 (62.1)
\0.001 0.001
1.46 (0.89–2.12) \0.001 1.94 (1.22–2.98) \0.001
120-min CP (ng/ml)
4.33 (2.50–5.93)
3.20 (1.89–5.06)
2.45 (1.39–3.67) \0.001
2.89 (1.73–4.72) \0.001
TC (mmol/L)
4.65 (4.01–5.30)
4.60 (4.00–5.22)
4.83 (4.05–5.72)
0.043
4.61 (4.02–5.40)
TG (mmol/L)
1.49 (1.03–2.20)
1.42 (1.00–2.12)
1.45 (0.92–1.99)
0.134
1.43 (1.00–2.07)
0.075
HDL-C (mmol/L)
1.05 (0.90–1.24)
1.07 (0.93–1.29)
1.14 (0.96–1.32)
0.003
1.08 (0.93–1.29)
0.007
LDL-C (mmol/L)
3.09 (2.51–3.71)
3.04 (2.48–3.65)
3.26 (2.40–4.01)
0.221
3.07 (2.45–3.74)
0.961
Current smoker (%)
522 (28.5)
146 (28.6)
43 (22.6)
0.193
189 (27.0)
0.443
Current drinker (%)
231 (12.6)
63 (12.3)
22 (11.6)
0.685
85 (12.1)
0.742
1,101 (60.1)
374 (73.2)
163 (85.8)
\0.001
537 (76.6)
\0.001
Insulin use (%)
0.860
Data are presented as median (IQR) for continuous variables and number (proportion) for category variables. Differences in medians were examined using the Mann–Whitney U test between two groups and Kruskal–Wallis test among three groups; Differences in proportions were tested using the Chi-square test BMI Body mass index, WC waist circumference, FPG fasting plasma glucose, TC total cholesterol, HTN hypertension, TG triglyceride, HDLC high-density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, FCP fasting C-peptide, 30-min CP 30-minute postprandial C-peptide, 120-min CP 120-minute postprandial C-peptide, SBP systolic blood pressure, DBP diastolic blood pressure
123
704
Acta Diabetol (2015) 52:701–708
Table 2 Association between body mass index (BMI) and any diabetic retinopathy Variables
No. at risk
Prevalence of DR (%)
Model 1
Model 2
OR (95 % CI)
P value
OR (95 % CI)
Model 3 P value
OR (95 % CI)
P value
BMI (kg/m2) 18.5–24.9
1,335
30.4
1
–
1
–
1
–
25.0–29.9
942
23.5
0.73 (0.60–0.89)
0.002
0.76 (0.61–0.95)
0.018
0.84 (0.66–1.05)
0.129
C30
256
28.9
0.99 (0.72–1.36)
0.958
1.07 (0.76–1.51)
0.704
1.28 (0.9–1.83)
–
0.100
37.2 27.8
– –
– –
– –
– –
1 0.79 (0.61–1.02)
18.8
–
–
–
–
0.51 (0.38–0.68)
P trend
0.027
0.402
0.168 0.757
FCP (ng/ml)a B1.27 1.28–2.17
806 854
C2.18
873
\0.001
P trend
– 0.070 0.000 \0.001
Model 1: adjusted for sex, age at diabetes diagnosis, and duration of diabetes; Model 2: adjusted for sex, age at diabetes diagnosis, duration of diabetes, mean blood pressure, HbA1c, TG, HDL-C, smoking status, drinking status, and insulin therapy; Model 3: further adjusted for FCP tertiles in addition to confounders in Model 2 OR odds ratio, 95 % CI 95 % confidence interval, FCP fasting C-peptide a
FCP was categorized by its tertiles
Table 3 Association between body mass index (BMI) and different stages of diabetic retinopathy Variables
Prevalence of DR (%)
Model 1 OR (95 % CI)
Model 2 P value
OR (95 % CI)
Model 3 P value
OR (95 % CI)
P value
DR I–II BMI (kg/m2) 18.5–24.9
21.9
1
25.0–29.9
17.7
0.76 (0.61–0.95)
0.015
1 0.78 (0.61–1.00)
0.054
1 0.86 (0.67–1.11)
0.259
C30
20.3
0.97 (0.68–1.37)
0.847
1.01 (0.69–1.48)
0.956
1.21 (0.82–1.80)
0.338
DR III–IV BMI (kg/m2) 18.5–24.9
8.5
1
25.0–29.9
5.7
0.64 (0.45–0.90)
0.011
1 0.69 (0.46–1.02)
0.065
1 0.76 (0.51–1.13)
0.177
C30
8.6
1.06 (0.64–1.75)
0.828
1.23 (0.71–2.13)
0.452
1.47 (0.83–2.58)
0.183
Model 1: adjusted for sex, age at diabetes diagnosis, and duration of diabetes; Model 2: adjusted for sex, age at diabetes diagnosis, duration of diabetes, mean blood pressure, HbA1c, TG, HDL-C, smoking status, drinking status, and insulin therapy; Model 3: further adjusted for FCP tertiles in addition to confounders in Model 2 OR odds ratio, 95 % CI 95 % confidence interval
Chicago, IL, USA); two-sided P values \0.05 were considered statistically significant.
Results Clinical characteristics of patients with and without DR Table 1 lists the clinical characteristics. Overall, 701 (27.7 %) patients had DR. Patients with DR had lower fasting C-peptide (FCP: 1.46 vs. 1.86 ng/ml, P \ 0.001), 30-min postprandial C-peptide (30-min CP: 1.94 vs. 2.63 ng/ml, P \ 0.001), 120-min postprandial C-peptide
123
(120-min CP: 2.89 vs. 4.33 ng/ml, P \ 0.001), and BMI (24.3 vs. 24.9 kg/m2, P = 0.001) compared to patients without DR. Patients with DR were diagnosed with diabetes at an earlier age (48 vs. 51 years, P \ 0.001) and had longer duration of diabetes (10 vs. 4 years, P \ 0.001), poorer glycemic control (HbA1c: 9.2 vs. 8.6 %, P = 0.003), and higher prevalence of hypertension (62.1 vs. 54.4 %, P = 0.001) than non-DR patients. The prevalence of current smoking status and current drinking status was similar between the two groups. However, a higher proportion of patients with DR required insulin therapy compared to those without DR (76.6 vs. 60.1 %, P \ 0.001).
Acta Diabetol (2015) 52:701–708
Association between BMI and DR The association between BMI (2-kg/m2 interval) and DR was U-shaped; the risk of DR decreased with BMI initially and reached the lowest when BMI was 28–29.9 kg/m2, and then increased in obesity (Fig. 1). Furthermore, compared with normal weight, overweight was associated with reduced risk of any DR, DR I–II, and DR III–IV after adjustment for sex, age at diabetes diagnosis, and duration of diabetes; the respective ORs [95 % confidence interval (CI)] were 0.73 (0.60–0.89), 0.76 (0.61–0.95), and 0.64
705
(0.45–0.90) (Model 1 in Tables 2, 3). However, the strength of negative association attenuated after further adjustment for HbA1c, mean blood pressure, high-density lipoprotein cholesterol (HDL-C), triglyceride (TG), smoking status, drinking status, and insulin therapy (Model 2 in Tables 2, 3), and became nonsignificant after adjustment for FCP in addition to the above confounders (Model 3 in Tables 2, 3). Obese patients had similar risk of DR compared with the normal weight patients (28.9 vs. 30.4 %). The association between BMI and DR was also assessed in patients with newly diagnosed diabetes (duration B1 year). As shown in supplementary Table 1, the results indicated that the protective effect of overweight for DR was more predominant in patients with duration B1 year than in the overall population; the corresponding ORs were 0.47 (0.27–0.81) versus 0.73 (0.60–0.89) after adjustment for sex, age at diabetes diagnosis, and duration of diabetes (in Model 1). However, the association of BMI with DR was nonsignificant after further adjustment for other confounders such as mean blood pressure, HbA1c, TG, HDL-C, smoking status, drinking status, and insulin therapy (in Model 2). Clinical characteristics correlating with BMI
Fig. 1 Odds ratios of BMI for any diabetic retinopathy (DR). ORs were adjusted for sex, age at diabetes diagnosis, and duration of diabetes. BMI was grouped by a 2-kg/m2 interval. And the group with a BMI of 28–29.9 kg/m2 was regarded as reference. The respective ORs (95 % CIs) in different BMI subgroups were 1.67 (1.05–2.66), 1.65 (1.13–2.43), 1.49 (1.05–2.13), 1.44 (1.01–2.07), 1.11 (0.76–1.63), 1 (reference), and 1.55 (1.02–2.34)
Correlation analysis showed that after adjustment for sex, age, and duration of diabetes, BMI was more closely associated with FCP and postprandial CP than with other clinical characteristics; there was a moderate positive relationship between BMI and FCP (partial correlation coefficient = 0.415, P \ 0.001). Other correlates of BMI
Fig. 2 Prevalence of any diabetic retinopathy (DR) among patients with different BMI categories stratified by FCP tertiles. P values were derived from binary logistic regression analysis and adjusted for sex, age at diabetes diagnosis, and duration of diabetes. At each FCP tertile, the group normal weight was regarded as reference
123
706
Acta Diabetol (2015) 52:701–708
Table 4 Clinical characteristics correlating with body mass index (BMI, kg/m2) Variables
Correlation coefficient
P value
Partial correlation coefficienta
Age at diagnosis of diabetes (years)
-0.004
0.036
–
–
Duration (years)
–
P value
-0.079
0.000
–
SBP (mmHg)
0.138
0.000
0.172
0.000
DBP (mmHg)
0.174
0.000
0.171
0.000
FCP (ng/ml)
0.416
0.000
0.415
0.000
30-min CP (ng/ml)
0.389
0.000
0.388
0.000
120-min CP (ng/ml)
0.321
0.000
0.320
0.000
-0.094
0.000
-0.106
0.000
TC (mmol/l)
0.108
0.000
0.105
0.000
Triglyceride (mmol/ l)
0.223
0.000
0.216
0.000
HDL-C (mmol/l)
-0.244
0.000
-0.246
0.000
LDL-C (mmol/l)
0.076
0.000
0.074
0.000
HbA1c (%)
a
Adjusted for sex, age, and duration of diabetes
SBP systolic blood pressure, DBP diastolic blood pressure, FCP fasting C-peptide, 30-min CP 30-min postprandial C-peptide, 120-min CP 120-min postprandial C-peptide, HbA1c hemoglobin A1c, TC total cholesterol, HDL-C high-density lipoprotein cholesterol, LDLC low-density lipoprotein cholesterol
included systolic blood pressure, diastolic blood pressure, HbA1c, TG, total cholesterol, low-density lipoprotein cholesterol (LDL-C), and HDL-C (all P \ 0.001, Table 4). Association between BMI and DR stratified by FCP tertiles Multiple logistic regression analysis indicated that FCP, 30-min CP, and 120-min CP were negatively associated with the presence of DR after adjustment for sex, age at diabetes diagnosis, duration of diabetes, HbA1c, mean blood pressure, HDL-C, TG, BMI, smoking status, drinking status, and insulin therapy; the ORs were 0.84 (0.75–0.94), 0.83 (0.76–0.90), and 0.85 (0.81–0.90), respectively (data not shown in Tables). As shown in Table 2, DR risk decreased to 51 % in patients with the highest FCP tertile compared with patients with the lowest FCP tertile (Table 2). The association between BMI and DR was further assessed at each FCP tertile. Following FCP tertile stratification, overweight and obese patients had comparable DR risk to normal weight patients after adjustment for sex, age at diabetes diagnosis, and duration of diabetes (Fig. 2). Discussion DR is the most frequent cause of new cases of blindness among adults aged 20–74 years, and the global
123
prevalence of DR is 34.6 % in diabetes patients [11]. In this study, DR prevalence was 27.7 %; DR was more prevalent in patients diagnosed with diabetes at an earlier age, with longer duration of diabetes, poorer glycemic control, and hypertension, which agrees with previous studies [11, 13–19]. Reports on the relationship between BMI and DR in type 2 diabetes patients were controversial. Earlier studies indicated that higher BMI increased DR risk [17, 18, 21– 24]. These studies usually treated BMI as a continuous variable [17, 18, 21–23], and the influence of overweight on DR presence was seldom investigated. Recently, Rooney et al. [26] noted that overweight and obesity exerted a protective effect against the development of any DR compared with normal/underweight in Asian population. In India, Raman et al. [27] reported that obesity (BMI C 23 kg/m2) played a protective role against any DR. However, we identified a U-shaped association between BMI and DR presence for the first time among Chinese type 2 diabetes patients, which is similar to the relationship between BMI and all-cause mortality in patients with type 2 diabetes [6]. Moreover, our data indicated that the protective effect of overweight on the presence of DR was more obvious in patients with newly diagnosed diabetes than in the total population. Previous studies contain few data on the underlying factors linking BMI with DR. Elevated BMI may have dual effects on DR development. On one hand, higher BMI correlates with elevated BP and worse lipid profile [26, 33, 34], known DR risk factors [11, 13, 14, 18, 21]. On the other hand, elevated BMI may exert beneficial effects: Increased BMI was associated with greater pancreatic b cell mass and higher CP levels [19, 35, 36]. Moreover, higher FCP and postprandial CP reduce DR risk in both type 1 and type 2 diabetes [19, 20, 37–42]. In our study, BMI was correlated with BP, lipid profile, HbA1c, and CP levels. BMI was more closely associated with FCP than with other correlates, with a moderate partial correlation coefficient of 0.415. Our data also demonstrated that FCP, 30-min CP, and 120-min CP were negatively associated with the risk of DR. We presume that the beneficial aspect of being overweight, which perhaps involves a complex of advantageous and disadvantageous factors, may compensate for its detrimental effect on DR development. Moreover, the negative association between overweight and DR became nonsignificant after further adjustment for FCP tertiles in addition to other confounders in the multiple regression models, the corresponding OR (95 % CI) for any DR changed from 0.73 (0.60–0.89) in Model 1 to 0.84 (0.66–1.05) in Model 3. Additionally, stratified by FCP tertiles, overweight patients had comparable DR risk to the normal weight individuals after adjustment for sex, age at diabetes diagnosis, and duration of diabetes. These results
Acta Diabetol (2015) 52:701–708
hint that FCP may mediate the protective action of overweight against DR presence. CP exerts direct and indirect protective function against DR. CP reflects endogenous insulin secretion in diabetes patients [43, 44]. Elevated CP is associated with improved glycemic control and less glycemic variability [38–42], which is beneficial for decreasing DR risk. Furthermore, CP may act as an active peptide hormone with potentially important physiological actions [45]; clinical and experimental data have shown that CP treatment improves diabetic nephropathy, neuropathy, and possibly retinopathy [45]. There were some limitations in our study. Firstly, this was a retrospective study. Secondly, data were collected during a 10-year period. Furthermore, prospective studies are warranted to explore the association of BMI with the development diabetic retinopathy in Chinese patients with diabetes. In conclusion, overweight patients have lower DR risk compared with normal weight individuals. Elevated C-peptide levels may partially account for the lower DR risk in overweight type 2 diabetes patients. Conflict of interest of interest.
The authors declare that they have no conflict
Ethical standard This study was approved by the Institutional Ethics Committee of Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, China. Human and Animal Rights disclosure All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5). Informed consent disclosure Informed consent was obtained from all patients for being included in the study.
References 1. Garvey WT, Garber AJ, Mechanick JI et al (2014) American association of clinical endocrinologists and american college of endocrinology position statement on the 2014 advanced framework for a new diagnosis of obesity as a chronic disease. Endocr Pract 20:977–989 2. Hou X, Lu J, Weng J et al (2013) Impact of waist circumference and body mass index on risk of cardiometabolic disorder and cardiovascular disease in Chinese adults: a national diabetes and metabolic disorders survey. PLoS ONE 8:e57319 3. Khalangot M, Tronko M, Kravchenko V, Kulchinska J, Hu G (2009) Body mass index and the risk of total and cardiovascular mortality among patients with type 2 diabetes: a large prospective study in Ukraine. Heart 95:454–460 4. Eeg-Olofsson K, Cederholm J, Nilsson PM et al (2009) Risk of cardiovascular disease and mortality in overweight and obese patients with type 2 diabetes: an observational study in 13,087 patients. Diabetologia 52:65–73
707 5. Maric C, Hall JE (2011) Obesity, metabolic syndrome and diabetic nephropathy. Contrib Nephrol 170:28–35 6. Logue J, Walker JJ, Leese G et al (2013) Association between BMI measured within a year after diagnosis of type 2 diabetes and mortality. Diabetes Care 36:887–893 7. Gruberg L, Weissman NJ, Waksman R et al (2002) The impact of obesity on the short-term and long-term outcomes after percutaneous coronary intervention: the obesity paradox? J Am Coll Cardiol 39:578–584 8. Karlberg C, Falk C, Green A, Sjolie AK, Grauslund J (2012) Proliferative retinopathy predicts nephropathy: a 25-year followup study of type 1 diabetic patients. Acta Diabetol 49:263–268 9. Pradeepa R, Surendar J, Indulekha K, Chella S, Anjana RM, Mohan V (2014) Relationship of diabetic retinopathy with coronary artery disease in Asian Indians with type 2 diabetes: The Chennai Urban Rural Epidemiology Study (CURES) Eye Study3. Diabetes Technol Ther [Epub ahead of print] 10. Miyamoto M, Kotani K, Okada K et al (2012) The correlation of common carotid arterial diameter with atherosclerosis and diabetic retinopathy in patients with type 2 diabetes mellitus. Acta Diabetol 49:63–68 11. Yau JW, Rogers SL, Kawasaki R et al (2012) Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care 35:556–564 12. Trento M, Passera P, Trevisan M et al (2013) Quality of life, impaired vision and social role in people with diabetes: a multicenter observational study. Acta Diabetol 50:873–877 13. Ding J, Wong TY (2012) Current epidemiology of diabetic retinopathy and diabetic macular edema. Curr Diabetes Rep 12:346–354 14. Brown JB, Pedula KL, Summers KH (2003) Diabetic retinopathy: contemporary prevalence in a well-controlled population. Diabetes Care 26:2637–2642 15. Diabetes Prevention Program Research Group (2007) The prevalence of retinopathy in impaired glucose tolerance and recentonset diabetes in the diabetes prevention program. Diabet Med 24:137–144 16. Writing Team for the Diabetes C, Complications Trial, Epidemiology of Diabetes I, Complications Research G (2002) Effect of intensive therapy on the microvascular complications of type 1 diabetes mellitus. JAMA 287:2563–2569 17. Keen H, Lee ET, Russell D, Miki E, Bennett PH, Lu M (2001) The appearance of retinopathy and progression to proliferative retinopathy: the WHO multinational study of vascular disease in diabetes. Diabetologia 44(Suppl 2):S22–S30 18. van Leiden HA, Dekker JM, Moll AC et al (2002) Blood pressure, lipids, and obesity are associated with retinopathy: the hoorn study. Diabetes Care 25:1320–1325 19. Cai X, Han X, Zhang S, Luo Y, Chen Y, Ji L (2014) Age at diagnosis and C-peptide level are associated with diabetic retinopathy in Chinese. PLoS ONE 9:e91174 20. Azad N, Agrawal L, Emanuele NV, Klein R, Bahn GD, Reaven P (2014) Association of blood glucose control and pancreatic reserve with diabetic retinopathy in the veterans affairs diabetes trial (VADT). Diabetologia 57:1124–1131 21. Tolonen N, Hietala K, Forsblom C et al (2013) Associations and interactions between lipid profiles, retinopathy and nephropathy in patients with type 1 diabetes: the FinnDiane Study. J Intern Med 274:469–479 22. Henricsson M, Nystrom L, Blohme G et al (2003) The incidence of retinopathy 10 years after diagnosis in young adult people with diabetes: results from the nationwide population-based diabetes incidence study in Sweden (DISS). Diabetes Care 26:349–354 23. Zhang L, Krzentowski G, Albert A, Lefebvre PJ (2001) Risk of developing retinopathy in diabetes control and complications trial type 1 diabetic patients with good or poor metabolic control. Diabetes Care 24:1275–1279
123
708 24. Dirani M, Xie J, Fenwick E et al (2011) Are obesity and anthropometry risk factors for diabetic retinopathy? The diabetes management project. Invest Ophthalmol Vis Sci 52:4416–4421 25. van Leiden HA, Dekker JM, Moll AC et al (2003) Risk factors for incident retinopathy in a diabetic and nondiabetic population: the Hoorn study. Arch Ophthalmol 121:245–251 26. Rooney D, Lye WK, Tan G, et al (2014) Body mass index and retinopathy in Asian populations with diabetes mellitus. Acta Diabetol. doi:10.1007/s00592-014-0602-2 27. Raman R, Rani PK, Gnanamoorthy P, Sudhir RR, Kumaramanikavel G, Sharma T (2010) Association of obesity with diabetic retinopathy: Sankara Nethralaya diabetic retinopathy epidemiology and molecular genetics study (SN-DREAMS Report no. 8). Acta Diabetol 47:209–215 28. Looker HC, Nyangoma SO, Cromie D et al (2012) Diabetic retinopathy at diagnosis of type 2 diabetes in Scotland. Diabetologia 55:2335–2342 29. Gu Y, Wang C, Zheng Y et al (2013) Cancer incidence and mortality in patients with type 2 diabetes treated with human insulin: a cohort study in Shanghai. PLoS ONE 8:e53411 30. Chinese Diabetes Society (2012) Guidelines for the prevention and treatment of type 2 diabetes in China (2010). Chin J Diabetes 20:81–117 31. (2000) Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser 894:i–xii, 1–253 32. Wilkinson CP, Ferris FL 3rd, Klein RE et al (2003) Proposed international clinical diabetic retinopathy and diabetic macular edema disease severity scales. Ophthalmology 110:1677–1682 33. de Mutsert R, Sun Q, Willett WC, Hu FB, van Dam RM (2014) Overweight in early adulthood, adult weight change, and risk of type 2 diabetes, cardiovascular diseases, and certain cancers in men: a cohort study. Am J Epidemiol 179:1353–1365 34. Hou X, Jia W, Bao Y et al (2008) Risk factors for overweight and obesity, and changes in body mass index of Chinese adults in Shanghai. BMC Public Health 8:389
123
Acta Diabetol (2015) 52:701–708 35. Rahier J, Guiot Y, Goebbels RM, Sempoux C, Henquin JC (2008) Pancreatic beta-cell mass in European subjects with type 2 diabetes. Diabetes Obes Metab 10(Suppl 4):32–42 36. Ahren J, Ahren B, Wierup N (2010) Increased beta-cell volume in mice fed a high-fat diet: a dynamic study over 12 months. Islets 2:353–356 37. Lachin JM, McGee P, Palmer JP (2014) Impact of C-peptide preservation on metabolic and clinical outcomes in the diabetes control and complications trial. Diabetes 63:739–748 38. Steffes MW, Sibley S, Jackson M, Thomas W (2003) Beta-cell function and the development of diabetes-related complications in the diabetes control and complications trial. Diabetes Care 26:832–836 39. Panero F, Novelli G, Zucco C et al (2009) Fasting plasma C-peptide and micro- and macrovascular complications in a large clinic-based cohort of type 1 diabetic patients. Diabetes Care 32:301–305 40. Suzuki K, Watanabe K, Motegi T, Kajinuma H (1989) High prevalence of proliferative retinopathy in diabetic patients with low pancreatic B-cell capacity. Diabetes Res Clin Pract 6:45–52 41. Boz M, Scheen AJ, Gerard PL, Castillo MJ, Lefebvre PJ (1995) Retinopathy, but not neuropathy, is influenced by the level of residual endogenous insulin secretion in type 2 diabetes. Diabete Metab 21:353–359 42. Bo S, Cavallo-Perin P, Gentile L, Repetti E, Pagano G (2000) Relationship of residual beta-cell function, metabolic control and chronic complications in type 2 diabetes mellitus. Acta Diabetol 37:125–129 43. Genuth SM (1996) Clinical utility of C-peptide measurement in diabetes mellitus. Endocr Pract 2:266–268 44. Jones AG, Hattersley AT (2013) The clinical utility of C-peptide measurement in the care of patients with diabetes. Diabet Med 30:803–817 45. Marques RG, Fontaine MJ, Rogers J (2004) C-peptide: much more than a byproduct of insulin biosynthesis. Pancreas 29:231–238
ORIGINAL ARTICLE
Association between body mass index and diabetic retinopathy in Chinese patients with type 2 diabetes Jun Lu • Xuhong Hou • Lei Zhang • Fusong Jiang • Cheng Hu • Yuqian Bao Weiping Jia
•
Received: 12 October 2014 / Accepted: 30 December 2014 / Published online: 22 January 2015 Ó Springer-Verlag Italia 2015
Abstract Aims To explore the factors mediating the relationship between body mass index (BMI) and diabetic retinopathy (DR) in Chinese type 2 diabetes patients. Methods This is a cross-sectional study. Data of 2,533 patients with type 2 diabetes were studied from the Shanghai Diabetes Registry Database. DR was assessed using non-mydriatic fundus photography and graded as non-DR, mild–moderate (DR I–II), and sight-threatening (DR III–IV). BMI (kg/m2) was classified as normal weight (18.5 B BMI \ 25), overweight (25 B BMI \ 30), and obese (BMI C 30). b cell function was evaluated by fasting C-peptide (FCP). Results DR was present in 701 (27.7 %) patients. Patients with DR had lower BMI (24.3 vs. 24.9 kg/m2, P = 0.001) and fasting C-peptide (1.46 vs. 1.86 ng/ml, P \ 0.001) than those without DR. The association between BMI (2 kg/m2 interval) and DR was U-shaped; patients with BMI 28–29.9 kg/m2 had the lowest DR rate. Compared with normal weight, overweight was associated with reduced risk of any DR [odds ratio (OR) 0.73], DR I–II
Managed by Antonio Secchi. Jun Lu and Xuhong Hou have contributed equally to this article.
Electronic supplementary material The online version of this article (doi:10.1007/s00592-014-0711-y) contains supplementary material, which is available to authorized users. J. Lu X. Hou L. Zhang F. Jiang C. Hu Y. Bao W. Jia (&) Department of Endocrinology and Metabolism, Shanghai Diabetes Institute, Shanghai Clinical Center for Diabetes, Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, 600 Yishan Road, Shanghai 200233, China e-mail: [email protected]
(OR 0.76), and DR III–IV (OR 0.64) after adjustment for sex, age at diabetes diagnosis, and duration of diabetes. This negative association attenuated after adjustment for other confounders and became nonsignificant after further adjustment for FCP. Patients with different BMI categories had similar DR risk when stratified by FCP tertiles. Conclusion Overweight patients have lower DR prevalence than normal weight individuals, which may be attributable to better b cell function in overweight patients. Keywords Type 2 diabetes Diabetic retinopathy Body mass index Fasting C-peptide
Introduction Overweight and obesity have been established as risk factors for diabetes [1]. A national survey reported that [60 % of diabetic patients in China are overweight or obese [2]. Traditionally, higher body mass index (BMI) is a risk factor for nephropathy and cardiovascular disease morbidity and mortality among type 2 diabetes patients [3– 5]. However, recent studies have reported that being overweight may confer a protective effect against all-cause mortality and on the life expectancy of cardiovascular disease in type 2 diabetes patients [6, 7]; this has been termed the obesity paradox [7]. Diabetic retinopathy (DR) is characterized as a microvascular complication of diabetes and is always accompanied with other macro- and microvascular complications of diabetes [8–10]. Studies also demonstrated that DR was the leading cause of new cases of blindness among adults aged 20–74 years [11] and severely influenced the quality of life in diabetic patients [12]. Previous studies have indicated that DR is more prevalent in patients at an earlier diagnosis
123
702
age of diabetes, with longer duration of diabetes, poorer b cell function and glycemic control, dyslipidemia, and hypertension [11, 13–20]. However, the association between BMI and DR is inconsistent. Earlier studies suggested that higher BMI increases DR risk [17, 18, 21–23], and there was higher DR prevalence in obese patients compared with normal weight patients [24]; other studies reported no association [15, 19, 25]. Recently, lower DR prevalence was reported in patients with higher BMI [26– 28]. It is not known whether this obesity paradox exists between BMI and DR among Chinese type 2 diabetes patients. We aimed to investigate the association between BMI and DR and to explore the factors mediating the relationship between BMI and DR in Chinese type 2 diabetes patients.
Subjects and methods Trial design and study population This was a retrospective study using the Shanghai Diabetes Registry (SDR) database, which was established in December 2001 at the Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, the Shanghai Clinical Center for Diabetes [29]. This computerized database was set up with the aim of evaluating the outcome of Chinese patients with diabetes. Data of demographics (including lifestyle, disease history, family history of disease, smoking and drinking habits), anthropometry, diagnoses, laboratory measurements and drug prescriptions was collected from the medical records. Current smokers were defined as those who had smoked C1 cigarette/day for at least 1 year. Current drinkers were defined as those who had consumed C30 g of alcohol/week on average for at least 1 year. The duration of diabetes was calculated as the time of non-mydriatic fundus photography performed minus the time of diagnosis of diabetes. Inclusion criteria of this study were as follows: complete data of demographics, laboratory measurements, and fundus photography; islet autoantibody (antibodies to glutamic acid decarboxylase and protein tyrosine phosphatase IA-2)negative; and BMI C 18.5 kg/m2. Patients with underweight were excluded due to a small sample size (n = 75). And a total of 2,533 patients with adult-onset diabetes between March 2003 and December 2012 were enrolled in this study. Anthropometric and laboratory investigation Anthropometric and biochemical measurements were taken at admission. The anthropometric indices of height and
123
Acta Diabetol (2015) 52:701–708
weight were measured while the patients were barefoot and in light clothing. The BMI was calculated as the weight divided by height squared (kg/m2). Blood pressure was measured twice using a mercury sphygmomanometer when subjects were supine, and then averaged. Mean blood pressure (mmHg) was calculated by the formula: (SBP ? 2 9 DBP)/3. After a minimum 10-h overnight fast, venous blood samples were drawn at 0, 30, and 120 min following the ingestion of a diabetic diet prescribed according to the criteria proposed by the Chinese Diabetes Society [30]. Hemoglobin A1c (HbA1c), lipid profiles, and fasting plasma glucose and fasting C-peptide were measured using fasting blood sample. Blood samples collected at 30 and 120 min were used to test postprandial C-Peptide. HbA1c was measured using high-performance liquid chromatography (HLC-73G7; Tosoh, Tokyo, Japan); fasting plasma glucose was measured using the glucose oxidase method (Roche, Mannheim, Germany); lipid profiles were determined using an auto-analyzer (Hitachi 7600, Tokyo, Japan); C-peptide (CP) was measured using radioimmunoassay (Linco Research, St Charles, MO, USA). All laboratory measurements met the Shanghai center for Clinical Laboratory criteria. Definitions of hypertension and overweight/obesity Hypertension was defined as blood pressure (BP) C 140/ 90 mmHg or usage of antihypertensive medications. BMI (kg/m2) was categorized using World Health Organization standards [31]: normal weight, 18.5 B BMI \ 25; overweight, 25 B BMI \ 30; and obese, BMI C 30. Assessment of diabetic retinopathy Fundus photography was performed for each patient following a standardized protocol. The participants were seated in a darkened room, and the posterior pole of each eye was photographed with a 45° 6.3-megapixel digital non-mydriatic camera (Canon CR6-45NM, Lake Success, New York, USA). Each fundus photography was graded by two independent ophthalmologists. If there was a disagreement, a third experienced ophthalmologist would check the results. DR was graded according to the 2003 standards proposed by the American Academy of Ophthalmology (AAO) [32]: no apparent retinopathy (DR 0), no abnormalities; mild nonproliferative DR (DR I), microaneurysms only; moderate nonproliferative DR (DR II), more than just microaneurysms but less than severe nonproliferative DR; severe nonproliferative DR (DR III), any of the following: [20 intraretinal hemorrhages in each of the four quadrants, definite venous beading in 2? quadrants, prominent intraretinal microvascular abnormalities in 1? quadrant and no signs of proliferative retinopathy;
Acta Diabetol (2015) 52:701–708
703
proliferative DR (DR IV), one or more of the following: neovascularization and vitreous/preretinal hemorrhage. Patients were grouped according to DR severity: non-DR (DR 0), mild–moderate retinopathy (DR I–II), and sightthreatening retinopathy (DR III–IV). Statistical analysis The distribution of continuous variables was assessed using the Kolmogorov–Smirnov test. Non-normally distributed continuous variables are expressed as median (interquartile range, IQR) and categorical variables as frequency (%). Differences in medians were examined using the Mann–Whitney U test between two groups and Kruskal–Wallis test among three groups; differences in proportions were tested using the Chi-square test (Table 1). The association between BMI and any DR was
assessed using multivariable binary logistic regression, and the association of BMI with different degrees of DR (DR I–II and DR III–IV) was assessed using multinomial logistic regression; odds ratios (ORs) and P values are presented (Tables 2, 3; Figs. 1, 2). Adjustment variables were first tested by collinearity diagnosis according to the following criteria: variance inflation factor (VIF) [10 or tolerance near 0.1; condition index [30; and variance proportions [50 %. Then, the variables including sex, age at diabetes diagnosis, duration of diabetes, mean blood pressure, HbA1c, TG, HDL-C, smoking status, drinking status, and insulin therapy were finally selected as adjustment variables for the multiple logistic regression analyses. Tests for trend were performed using the BMI categories and FCP tertiles as ordinal variables in the corresponding logistic regression models. All statistical analyses were performed using SPSS 12.0 (SPSS Inc.,
Table 1 Comparison of clinical characteristics in patients with and without diabetic retinopathy Variables
Non-DR (n = 1,832)
DR I–II (n = 511)
Gender (male)
1,101 (60.1)
295 (57.7)
DR III–IV (n = 190)
P value
95 (50)
0.010
DR (n = 701) 390 (55.6)
P value DR versus non-DR 0.041
Age (years)
57 (49–64)
57 (51–65)
59.5 (53–66)
0.003
58 (51–65)
0.002
Age at diagnosis of diabetes (years)
51 (43–58)
48 (41–54)
49 (41–54)
\0.001
48 (41–54)
\0.001
10 (5–14)
11 (6–15)
\0.001
10 (5–14)
\0.001
Duration of diabetes (years) 2
BMI (kg/m ) WC (cm) Abdominal obesity
4 (0.6–10) 24.9 (22.7–27.6) 89 (83–96)
24.3 (22.3–27.0) 88 (82–96)
24.4 (22.0–26.5)
0.002
87 (82–93)
0.024
80 (45.5)
0.012
943 (54.7)
239 (51.3)
SBP (mmHg)
130 (120–140)
130 (120–140)
140 (125–150)
DBP (mmHg)
80 (70–86)
80 (75–90)
80 (75–90)
24.3 (22.2–26.8) 88 (82–95) 319 (49.7)
0.001 0.037 0.029
Blood pressure \0.001 0.047
130 (120–147) 80 (75–90)
\0.001 0.021
100 (93–110)
\0.001
987 (54.4)
301 (59.4)
131 (69.3)
\0.001
8.6 (7.1–10.6)
9.2 (7.5–10.7)
9.3 (7.6–11.0)
0.007
9.2 (7.5–10.8)
0.003
FPG (mmol/L)
7.91 (6.43–10.0)
7.90 (6.32–10.5)
8.68 (6.83–11.4)
0.020
8.05 (6.45–10.9)
0.089
FCP (ng/ml) 30-min CP (ng/ml)
1.86 (1.20–2.63) 2.63 (1.65–3.90)
1.52 (0.93–2.15) 2.04 (1.26–3.09)
1.24 (0.79–2.03) \0.001 1.74 (1.09–2.60) \0.001
Mean BP (mmHg) HTN (%) HbA1c (%)
97 (90–103)
97 (93–107)
97 (93–107) 432 (62.1)
\0.001 0.001
1.46 (0.89–2.12) \0.001 1.94 (1.22–2.98) \0.001
120-min CP (ng/ml)
4.33 (2.50–5.93)
3.20 (1.89–5.06)
2.45 (1.39–3.67) \0.001
2.89 (1.73–4.72) \0.001
TC (mmol/L)
4.65 (4.01–5.30)
4.60 (4.00–5.22)
4.83 (4.05–5.72)
0.043
4.61 (4.02–5.40)
TG (mmol/L)
1.49 (1.03–2.20)
1.42 (1.00–2.12)
1.45 (0.92–1.99)
0.134
1.43 (1.00–2.07)
0.075
HDL-C (mmol/L)
1.05 (0.90–1.24)
1.07 (0.93–1.29)
1.14 (0.96–1.32)
0.003
1.08 (0.93–1.29)
0.007
LDL-C (mmol/L)
3.09 (2.51–3.71)
3.04 (2.48–3.65)
3.26 (2.40–4.01)
0.221
3.07 (2.45–3.74)
0.961
Current smoker (%)
522 (28.5)
146 (28.6)
43 (22.6)
0.193
189 (27.0)
0.443
Current drinker (%)
231 (12.6)
63 (12.3)
22 (11.6)
0.685
85 (12.1)
0.742
1,101 (60.1)
374 (73.2)
163 (85.8)
\0.001
537 (76.6)
\0.001
Insulin use (%)
0.860
Data are presented as median (IQR) for continuous variables and number (proportion) for category variables. Differences in medians were examined using the Mann–Whitney U test between two groups and Kruskal–Wallis test among three groups; Differences in proportions were tested using the Chi-square test BMI Body mass index, WC waist circumference, FPG fasting plasma glucose, TC total cholesterol, HTN hypertension, TG triglyceride, HDLC high-density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, FCP fasting C-peptide, 30-min CP 30-minute postprandial C-peptide, 120-min CP 120-minute postprandial C-peptide, SBP systolic blood pressure, DBP diastolic blood pressure
123
704
Acta Diabetol (2015) 52:701–708
Table 2 Association between body mass index (BMI) and any diabetic retinopathy Variables
No. at risk
Prevalence of DR (%)
Model 1
Model 2
OR (95 % CI)
P value
OR (95 % CI)
Model 3 P value
OR (95 % CI)
P value
BMI (kg/m2) 18.5–24.9
1,335
30.4
1
–
1
–
1
–
25.0–29.9
942
23.5
0.73 (0.60–0.89)
0.002
0.76 (0.61–0.95)
0.018
0.84 (0.66–1.05)
0.129
C30
256
28.9
0.99 (0.72–1.36)
0.958
1.07 (0.76–1.51)
0.704
1.28 (0.9–1.83)
–
0.100
37.2 27.8
– –
– –
– –
– –
1 0.79 (0.61–1.02)
18.8
–
–
–
–
0.51 (0.38–0.68)
P trend
0.027
0.402
0.168 0.757
FCP (ng/ml)a B1.27 1.28–2.17
806 854
C2.18
873
\0.001
P trend
– 0.070 0.000 \0.001
Model 1: adjusted for sex, age at diabetes diagnosis, and duration of diabetes; Model 2: adjusted for sex, age at diabetes diagnosis, duration of diabetes, mean blood pressure, HbA1c, TG, HDL-C, smoking status, drinking status, and insulin therapy; Model 3: further adjusted for FCP tertiles in addition to confounders in Model 2 OR odds ratio, 95 % CI 95 % confidence interval, FCP fasting C-peptide a
FCP was categorized by its tertiles
Table 3 Association between body mass index (BMI) and different stages of diabetic retinopathy Variables
Prevalence of DR (%)
Model 1 OR (95 % CI)
Model 2 P value
OR (95 % CI)
Model 3 P value
OR (95 % CI)
P value
DR I–II BMI (kg/m2) 18.5–24.9
21.9
1
25.0–29.9
17.7
0.76 (0.61–0.95)
0.015
1 0.78 (0.61–1.00)
0.054
1 0.86 (0.67–1.11)
0.259
C30
20.3
0.97 (0.68–1.37)
0.847
1.01 (0.69–1.48)
0.956
1.21 (0.82–1.80)
0.338
DR III–IV BMI (kg/m2) 18.5–24.9
8.5
1
25.0–29.9
5.7
0.64 (0.45–0.90)
0.011
1 0.69 (0.46–1.02)
0.065
1 0.76 (0.51–1.13)
0.177
C30
8.6
1.06 (0.64–1.75)
0.828
1.23 (0.71–2.13)
0.452
1.47 (0.83–2.58)
0.183
Model 1: adjusted for sex, age at diabetes diagnosis, and duration of diabetes; Model 2: adjusted for sex, age at diabetes diagnosis, duration of diabetes, mean blood pressure, HbA1c, TG, HDL-C, smoking status, drinking status, and insulin therapy; Model 3: further adjusted for FCP tertiles in addition to confounders in Model 2 OR odds ratio, 95 % CI 95 % confidence interval
Chicago, IL, USA); two-sided P values \0.05 were considered statistically significant.
Results Clinical characteristics of patients with and without DR Table 1 lists the clinical characteristics. Overall, 701 (27.7 %) patients had DR. Patients with DR had lower fasting C-peptide (FCP: 1.46 vs. 1.86 ng/ml, P \ 0.001), 30-min postprandial C-peptide (30-min CP: 1.94 vs. 2.63 ng/ml, P \ 0.001), 120-min postprandial C-peptide
123
(120-min CP: 2.89 vs. 4.33 ng/ml, P \ 0.001), and BMI (24.3 vs. 24.9 kg/m2, P = 0.001) compared to patients without DR. Patients with DR were diagnosed with diabetes at an earlier age (48 vs. 51 years, P \ 0.001) and had longer duration of diabetes (10 vs. 4 years, P \ 0.001), poorer glycemic control (HbA1c: 9.2 vs. 8.6 %, P = 0.003), and higher prevalence of hypertension (62.1 vs. 54.4 %, P = 0.001) than non-DR patients. The prevalence of current smoking status and current drinking status was similar between the two groups. However, a higher proportion of patients with DR required insulin therapy compared to those without DR (76.6 vs. 60.1 %, P \ 0.001).
Acta Diabetol (2015) 52:701–708
Association between BMI and DR The association between BMI (2-kg/m2 interval) and DR was U-shaped; the risk of DR decreased with BMI initially and reached the lowest when BMI was 28–29.9 kg/m2, and then increased in obesity (Fig. 1). Furthermore, compared with normal weight, overweight was associated with reduced risk of any DR, DR I–II, and DR III–IV after adjustment for sex, age at diabetes diagnosis, and duration of diabetes; the respective ORs [95 % confidence interval (CI)] were 0.73 (0.60–0.89), 0.76 (0.61–0.95), and 0.64
705
(0.45–0.90) (Model 1 in Tables 2, 3). However, the strength of negative association attenuated after further adjustment for HbA1c, mean blood pressure, high-density lipoprotein cholesterol (HDL-C), triglyceride (TG), smoking status, drinking status, and insulin therapy (Model 2 in Tables 2, 3), and became nonsignificant after adjustment for FCP in addition to the above confounders (Model 3 in Tables 2, 3). Obese patients had similar risk of DR compared with the normal weight patients (28.9 vs. 30.4 %). The association between BMI and DR was also assessed in patients with newly diagnosed diabetes (duration B1 year). As shown in supplementary Table 1, the results indicated that the protective effect of overweight for DR was more predominant in patients with duration B1 year than in the overall population; the corresponding ORs were 0.47 (0.27–0.81) versus 0.73 (0.60–0.89) after adjustment for sex, age at diabetes diagnosis, and duration of diabetes (in Model 1). However, the association of BMI with DR was nonsignificant after further adjustment for other confounders such as mean blood pressure, HbA1c, TG, HDL-C, smoking status, drinking status, and insulin therapy (in Model 2). Clinical characteristics correlating with BMI
Fig. 1 Odds ratios of BMI for any diabetic retinopathy (DR). ORs were adjusted for sex, age at diabetes diagnosis, and duration of diabetes. BMI was grouped by a 2-kg/m2 interval. And the group with a BMI of 28–29.9 kg/m2 was regarded as reference. The respective ORs (95 % CIs) in different BMI subgroups were 1.67 (1.05–2.66), 1.65 (1.13–2.43), 1.49 (1.05–2.13), 1.44 (1.01–2.07), 1.11 (0.76–1.63), 1 (reference), and 1.55 (1.02–2.34)
Correlation analysis showed that after adjustment for sex, age, and duration of diabetes, BMI was more closely associated with FCP and postprandial CP than with other clinical characteristics; there was a moderate positive relationship between BMI and FCP (partial correlation coefficient = 0.415, P \ 0.001). Other correlates of BMI
Fig. 2 Prevalence of any diabetic retinopathy (DR) among patients with different BMI categories stratified by FCP tertiles. P values were derived from binary logistic regression analysis and adjusted for sex, age at diabetes diagnosis, and duration of diabetes. At each FCP tertile, the group normal weight was regarded as reference
123
706
Acta Diabetol (2015) 52:701–708
Table 4 Clinical characteristics correlating with body mass index (BMI, kg/m2) Variables
Correlation coefficient
P value
Partial correlation coefficienta
Age at diagnosis of diabetes (years)
-0.004
0.036
–
–
Duration (years)
–
P value
-0.079
0.000
–
SBP (mmHg)
0.138
0.000
0.172
0.000
DBP (mmHg)
0.174
0.000
0.171
0.000
FCP (ng/ml)
0.416
0.000
0.415
0.000
30-min CP (ng/ml)
0.389
0.000
0.388
0.000
120-min CP (ng/ml)
0.321
0.000
0.320
0.000
-0.094
0.000
-0.106
0.000
TC (mmol/l)
0.108
0.000
0.105
0.000
Triglyceride (mmol/ l)
0.223
0.000
0.216
0.000
HDL-C (mmol/l)
-0.244
0.000
-0.246
0.000
LDL-C (mmol/l)
0.076
0.000
0.074
0.000
HbA1c (%)
a
Adjusted for sex, age, and duration of diabetes
SBP systolic blood pressure, DBP diastolic blood pressure, FCP fasting C-peptide, 30-min CP 30-min postprandial C-peptide, 120-min CP 120-min postprandial C-peptide, HbA1c hemoglobin A1c, TC total cholesterol, HDL-C high-density lipoprotein cholesterol, LDLC low-density lipoprotein cholesterol
included systolic blood pressure, diastolic blood pressure, HbA1c, TG, total cholesterol, low-density lipoprotein cholesterol (LDL-C), and HDL-C (all P \ 0.001, Table 4). Association between BMI and DR stratified by FCP tertiles Multiple logistic regression analysis indicated that FCP, 30-min CP, and 120-min CP were negatively associated with the presence of DR after adjustment for sex, age at diabetes diagnosis, duration of diabetes, HbA1c, mean blood pressure, HDL-C, TG, BMI, smoking status, drinking status, and insulin therapy; the ORs were 0.84 (0.75–0.94), 0.83 (0.76–0.90), and 0.85 (0.81–0.90), respectively (data not shown in Tables). As shown in Table 2, DR risk decreased to 51 % in patients with the highest FCP tertile compared with patients with the lowest FCP tertile (Table 2). The association between BMI and DR was further assessed at each FCP tertile. Following FCP tertile stratification, overweight and obese patients had comparable DR risk to normal weight patients after adjustment for sex, age at diabetes diagnosis, and duration of diabetes (Fig. 2). Discussion DR is the most frequent cause of new cases of blindness among adults aged 20–74 years, and the global
123
prevalence of DR is 34.6 % in diabetes patients [11]. In this study, DR prevalence was 27.7 %; DR was more prevalent in patients diagnosed with diabetes at an earlier age, with longer duration of diabetes, poorer glycemic control, and hypertension, which agrees with previous studies [11, 13–19]. Reports on the relationship between BMI and DR in type 2 diabetes patients were controversial. Earlier studies indicated that higher BMI increased DR risk [17, 18, 21– 24]. These studies usually treated BMI as a continuous variable [17, 18, 21–23], and the influence of overweight on DR presence was seldom investigated. Recently, Rooney et al. [26] noted that overweight and obesity exerted a protective effect against the development of any DR compared with normal/underweight in Asian population. In India, Raman et al. [27] reported that obesity (BMI C 23 kg/m2) played a protective role against any DR. However, we identified a U-shaped association between BMI and DR presence for the first time among Chinese type 2 diabetes patients, which is similar to the relationship between BMI and all-cause mortality in patients with type 2 diabetes [6]. Moreover, our data indicated that the protective effect of overweight on the presence of DR was more obvious in patients with newly diagnosed diabetes than in the total population. Previous studies contain few data on the underlying factors linking BMI with DR. Elevated BMI may have dual effects on DR development. On one hand, higher BMI correlates with elevated BP and worse lipid profile [26, 33, 34], known DR risk factors [11, 13, 14, 18, 21]. On the other hand, elevated BMI may exert beneficial effects: Increased BMI was associated with greater pancreatic b cell mass and higher CP levels [19, 35, 36]. Moreover, higher FCP and postprandial CP reduce DR risk in both type 1 and type 2 diabetes [19, 20, 37–42]. In our study, BMI was correlated with BP, lipid profile, HbA1c, and CP levels. BMI was more closely associated with FCP than with other correlates, with a moderate partial correlation coefficient of 0.415. Our data also demonstrated that FCP, 30-min CP, and 120-min CP were negatively associated with the risk of DR. We presume that the beneficial aspect of being overweight, which perhaps involves a complex of advantageous and disadvantageous factors, may compensate for its detrimental effect on DR development. Moreover, the negative association between overweight and DR became nonsignificant after further adjustment for FCP tertiles in addition to other confounders in the multiple regression models, the corresponding OR (95 % CI) for any DR changed from 0.73 (0.60–0.89) in Model 1 to 0.84 (0.66–1.05) in Model 3. Additionally, stratified by FCP tertiles, overweight patients had comparable DR risk to the normal weight individuals after adjustment for sex, age at diabetes diagnosis, and duration of diabetes. These results
Acta Diabetol (2015) 52:701–708
hint that FCP may mediate the protective action of overweight against DR presence. CP exerts direct and indirect protective function against DR. CP reflects endogenous insulin secretion in diabetes patients [43, 44]. Elevated CP is associated with improved glycemic control and less glycemic variability [38–42], which is beneficial for decreasing DR risk. Furthermore, CP may act as an active peptide hormone with potentially important physiological actions [45]; clinical and experimental data have shown that CP treatment improves diabetic nephropathy, neuropathy, and possibly retinopathy [45]. There were some limitations in our study. Firstly, this was a retrospective study. Secondly, data were collected during a 10-year period. Furthermore, prospective studies are warranted to explore the association of BMI with the development diabetic retinopathy in Chinese patients with diabetes. In conclusion, overweight patients have lower DR risk compared with normal weight individuals. Elevated C-peptide levels may partially account for the lower DR risk in overweight type 2 diabetes patients. Conflict of interest of interest.
The authors declare that they have no conflict
Ethical standard This study was approved by the Institutional Ethics Committee of Shanghai Jiao Tong University Affiliated Sixth People’s Hospital, China. Human and Animal Rights disclosure All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5). Informed consent disclosure Informed consent was obtained from all patients for being included in the study.
References 1. Garvey WT, Garber AJ, Mechanick JI et al (2014) American association of clinical endocrinologists and american college of endocrinology position statement on the 2014 advanced framework for a new diagnosis of obesity as a chronic disease. Endocr Pract 20:977–989 2. Hou X, Lu J, Weng J et al (2013) Impact of waist circumference and body mass index on risk of cardiometabolic disorder and cardiovascular disease in Chinese adults: a national diabetes and metabolic disorders survey. PLoS ONE 8:e57319 3. Khalangot M, Tronko M, Kravchenko V, Kulchinska J, Hu G (2009) Body mass index and the risk of total and cardiovascular mortality among patients with type 2 diabetes: a large prospective study in Ukraine. Heart 95:454–460 4. Eeg-Olofsson K, Cederholm J, Nilsson PM et al (2009) Risk of cardiovascular disease and mortality in overweight and obese patients with type 2 diabetes: an observational study in 13,087 patients. Diabetologia 52:65–73
707 5. Maric C, Hall JE (2011) Obesity, metabolic syndrome and diabetic nephropathy. Contrib Nephrol 170:28–35 6. Logue J, Walker JJ, Leese G et al (2013) Association between BMI measured within a year after diagnosis of type 2 diabetes and mortality. Diabetes Care 36:887–893 7. Gruberg L, Weissman NJ, Waksman R et al (2002) The impact of obesity on the short-term and long-term outcomes after percutaneous coronary intervention: the obesity paradox? J Am Coll Cardiol 39:578–584 8. Karlberg C, Falk C, Green A, Sjolie AK, Grauslund J (2012) Proliferative retinopathy predicts nephropathy: a 25-year followup study of type 1 diabetic patients. Acta Diabetol 49:263–268 9. Pradeepa R, Surendar J, Indulekha K, Chella S, Anjana RM, Mohan V (2014) Relationship of diabetic retinopathy with coronary artery disease in Asian Indians with type 2 diabetes: The Chennai Urban Rural Epidemiology Study (CURES) Eye Study3. Diabetes Technol Ther [Epub ahead of print] 10. Miyamoto M, Kotani K, Okada K et al (2012) The correlation of common carotid arterial diameter with atherosclerosis and diabetic retinopathy in patients with type 2 diabetes mellitus. Acta Diabetol 49:63–68 11. Yau JW, Rogers SL, Kawasaki R et al (2012) Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care 35:556–564 12. Trento M, Passera P, Trevisan M et al (2013) Quality of life, impaired vision and social role in people with diabetes: a multicenter observational study. Acta Diabetol 50:873–877 13. Ding J, Wong TY (2012) Current epidemiology of diabetic retinopathy and diabetic macular edema. Curr Diabetes Rep 12:346–354 14. Brown JB, Pedula KL, Summers KH (2003) Diabetic retinopathy: contemporary prevalence in a well-controlled population. Diabetes Care 26:2637–2642 15. Diabetes Prevention Program Research Group (2007) The prevalence of retinopathy in impaired glucose tolerance and recentonset diabetes in the diabetes prevention program. Diabet Med 24:137–144 16. Writing Team for the Diabetes C, Complications Trial, Epidemiology of Diabetes I, Complications Research G (2002) Effect of intensive therapy on the microvascular complications of type 1 diabetes mellitus. JAMA 287:2563–2569 17. Keen H, Lee ET, Russell D, Miki E, Bennett PH, Lu M (2001) The appearance of retinopathy and progression to proliferative retinopathy: the WHO multinational study of vascular disease in diabetes. Diabetologia 44(Suppl 2):S22–S30 18. van Leiden HA, Dekker JM, Moll AC et al (2002) Blood pressure, lipids, and obesity are associated with retinopathy: the hoorn study. Diabetes Care 25:1320–1325 19. Cai X, Han X, Zhang S, Luo Y, Chen Y, Ji L (2014) Age at diagnosis and C-peptide level are associated with diabetic retinopathy in Chinese. PLoS ONE 9:e91174 20. Azad N, Agrawal L, Emanuele NV, Klein R, Bahn GD, Reaven P (2014) Association of blood glucose control and pancreatic reserve with diabetic retinopathy in the veterans affairs diabetes trial (VADT). Diabetologia 57:1124–1131 21. Tolonen N, Hietala K, Forsblom C et al (2013) Associations and interactions between lipid profiles, retinopathy and nephropathy in patients with type 1 diabetes: the FinnDiane Study. J Intern Med 274:469–479 22. Henricsson M, Nystrom L, Blohme G et al (2003) The incidence of retinopathy 10 years after diagnosis in young adult people with diabetes: results from the nationwide population-based diabetes incidence study in Sweden (DISS). Diabetes Care 26:349–354 23. Zhang L, Krzentowski G, Albert A, Lefebvre PJ (2001) Risk of developing retinopathy in diabetes control and complications trial type 1 diabetic patients with good or poor metabolic control. Diabetes Care 24:1275–1279
123
708 24. Dirani M, Xie J, Fenwick E et al (2011) Are obesity and anthropometry risk factors for diabetic retinopathy? The diabetes management project. Invest Ophthalmol Vis Sci 52:4416–4421 25. van Leiden HA, Dekker JM, Moll AC et al (2003) Risk factors for incident retinopathy in a diabetic and nondiabetic population: the Hoorn study. Arch Ophthalmol 121:245–251 26. Rooney D, Lye WK, Tan G, et al (2014) Body mass index and retinopathy in Asian populations with diabetes mellitus. Acta Diabetol. doi:10.1007/s00592-014-0602-2 27. Raman R, Rani PK, Gnanamoorthy P, Sudhir RR, Kumaramanikavel G, Sharma T (2010) Association of obesity with diabetic retinopathy: Sankara Nethralaya diabetic retinopathy epidemiology and molecular genetics study (SN-DREAMS Report no. 8). Acta Diabetol 47:209–215 28. Looker HC, Nyangoma SO, Cromie D et al (2012) Diabetic retinopathy at diagnosis of type 2 diabetes in Scotland. Diabetologia 55:2335–2342 29. Gu Y, Wang C, Zheng Y et al (2013) Cancer incidence and mortality in patients with type 2 diabetes treated with human insulin: a cohort study in Shanghai. PLoS ONE 8:e53411 30. Chinese Diabetes Society (2012) Guidelines for the prevention and treatment of type 2 diabetes in China (2010). Chin J Diabetes 20:81–117 31. (2000) Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser 894:i–xii, 1–253 32. Wilkinson CP, Ferris FL 3rd, Klein RE et al (2003) Proposed international clinical diabetic retinopathy and diabetic macular edema disease severity scales. Ophthalmology 110:1677–1682 33. de Mutsert R, Sun Q, Willett WC, Hu FB, van Dam RM (2014) Overweight in early adulthood, adult weight change, and risk of type 2 diabetes, cardiovascular diseases, and certain cancers in men: a cohort study. Am J Epidemiol 179:1353–1365 34. Hou X, Jia W, Bao Y et al (2008) Risk factors for overweight and obesity, and changes in body mass index of Chinese adults in Shanghai. BMC Public Health 8:389
123
Acta Diabetol (2015) 52:701–708 35. Rahier J, Guiot Y, Goebbels RM, Sempoux C, Henquin JC (2008) Pancreatic beta-cell mass in European subjects with type 2 diabetes. Diabetes Obes Metab 10(Suppl 4):32–42 36. Ahren J, Ahren B, Wierup N (2010) Increased beta-cell volume in mice fed a high-fat diet: a dynamic study over 12 months. Islets 2:353–356 37. Lachin JM, McGee P, Palmer JP (2014) Impact of C-peptide preservation on metabolic and clinical outcomes in the diabetes control and complications trial. Diabetes 63:739–748 38. Steffes MW, Sibley S, Jackson M, Thomas W (2003) Beta-cell function and the development of diabetes-related complications in the diabetes control and complications trial. Diabetes Care 26:832–836 39. Panero F, Novelli G, Zucco C et al (2009) Fasting plasma C-peptide and micro- and macrovascular complications in a large clinic-based cohort of type 1 diabetic patients. Diabetes Care 32:301–305 40. Suzuki K, Watanabe K, Motegi T, Kajinuma H (1989) High prevalence of proliferative retinopathy in diabetic patients with low pancreatic B-cell capacity. Diabetes Res Clin Pract 6:45–52 41. Boz M, Scheen AJ, Gerard PL, Castillo MJ, Lefebvre PJ (1995) Retinopathy, but not neuropathy, is influenced by the level of residual endogenous insulin secretion in type 2 diabetes. Diabete Metab 21:353–359 42. Bo S, Cavallo-Perin P, Gentile L, Repetti E, Pagano G (2000) Relationship of residual beta-cell function, metabolic control and chronic complications in type 2 diabetes mellitus. Acta Diabetol 37:125–129 43. Genuth SM (1996) Clinical utility of C-peptide measurement in diabetes mellitus. Endocr Pract 2:266–268 44. Jones AG, Hattersley AT (2013) The clinical utility of C-peptide measurement in the care of patients with diabetes. Diabet Med 30:803–817 45. Marques RG, Fontaine MJ, Rogers J (2004) C-peptide: much more than a byproduct of insulin biosynthesis. Pancreas 29:231–238
