Abstract

Background. Diabetic nephropathy is the most common disease leading to end-stage renal disease (ESRD) in many countries including Germany. Some previous studies, mainly from the US, suggest that low socioeconomic status (SES) may increase the risk of ESRD. No data are available whether the SES influences the development of diabetic nephropathy in patients with diabetes mellitus in Germany.

Methods. This cross-sectional study was performed on patients treated at a large university outpatient department for endocrinology and metabolic diseases. A total of 174 patients with type 1 and 651 patients with type 2 diabetes and chronic preterminal diabetic nephropathy were studied [mainly chronic kidney disease (CKD) Stages 2 and 3]. Only very few CKD Stage 5 patients were included. Patients with acute kidney injury or abnormal urinary sediment were excluded. Patients were asked about their social status using a questionnaire. Social status was determined by three components: education, highest professional position achieved and household net income. Each component was assessed by a score with 1 to 7 points to generate a total score with a minimum of 3 up to maximum of 21 points. Smoking habits were also assessed by questionnaire. HbA1c, systolic and diastolic blood pressure and body mass index from the last observation were recorded. Estimated glomerular filtration rate (eGFR) was calculated according to the modified equation 7 MDRD formula. Patients were grouped into the CKD stages according to eGFR and presence of albuminuria. Multivariate analysis was used for data analysis.

Results. Patients were grouped in tertiles according to their social status (Tertile 1: 307, Tertile 2: 269, Tertile 3: 269 patients). The majority of type 1 (50.9%) and type 2 (64.9%) patients were in CKD Stages 2 and 3. Multivariate analysis revealed that SES is an independent predictor of renal function in all patients as well as in type 2 diabetic patients with diabetic nephropathy. This relationship was independent of smoking behaviour, duration of diabetes and HbA1c values. There was no association between renal function and SES in type 1 diabetic patients, but a type 2 error caused by low patient number cannot be excluded. Furthermore, no significant association between SES and albuminuria (defined ≥20 mg/L) existed. There was no significant difference in the number of visits to the clinic in regard to SES excluding referral bias.

Conclusions. A lower SES was associated with the presence of diabetic nephropathy in patients with type 2 diabetes in a German population. The causes for this association are likely multiple.

Introduction

Diabetic nephropathy due to type 2 diabetes is in many countries including Germany a leading cause of end-stage renal disease (ESRD) [ 1 , 2 ].

Analysis in several countries (Sweden, UK, Asia and the USA) suggest a higher incidence of renal disease in patients with low socioeconomic status (SES), defined differently in various studies [ 3–10 ]. However, part of this association may be due to ethnic minorities with lower SES, but other genetic risk factors, making the socioeconomic factor an important confounding factor [ 11 , 12 ]. Therefore, the present study was undertaken to analyse a potential relationship between socioeconomic factor and diabetic nephropathy in a relatively homogeneous population of Caucasian people cared for at a tertiary university diabetes centre. Our cross-sectional study revealed a close correlation between low SES and diabetic nephropathy in patients with type 2 diabetes independent of other factors. To the best of our knowledge, this is the first study showing such a relationship in a well-defined collective of German patients.

Patients and methods

This retrospective cohort study was performed to identify a potential association between SES and renal function in patients with types 1 and 2 diabetes from a tertiary university hospital with a large diabetes clinic. Patient data are stored in an electronic database EMIL™ ( http://www.itc-ms.de ). Data from the years 1989 to 2007 were analysed. A total of 174 patients with type 1 and 651 patients with type 2 diabetes were investigated who had data for at least two standardized creatinine and albuminuria measurements. Only Caucasian patients were studied (determined by asking the ethnic origin of both parents). Electronic medical records were examined [ 13 , 14 ]. Patients attended the clinic normally every 3 months or according to clinical needs. Access to health care at the University Outpatient Clinic was similar for all. The University Hospital is the only hospital in town and also functions as a municipal hospital. Besides our clinic, there are only two additional primary physicians in Jena specialized in the treatment of diabetic patients. We care for ∼75% of all diabetic patients in the region. Therefore, we strongly believe that the studied collective reflects very well the general population of diabetic patients in our town. The overall number of visits to the ambulatory care was independent of the SES indicating equal access for all patients (patients with the lowest SES 36.6 ± 27.3; patients in the second quartile: 34.6 ± 23.1; patients in the third quartile: 36.4 ± 24.8, patients in the with the highest SES: 32.9 ± 24.2 total visits; not significant). In addition, we offer no specific care for private-paying patients. Therefore, we do not think that referral bias or unequal access to the clinic may have influenced our analysis. Only very few dialysis patients were part of this ambulatory population [ 14 ] and all included dialysis patients were employed. Only stable outpatient diabetics were included in this study. Patients with acute kidney injury [patients admitted to the hospital with evidence and/or risk factors for acute kidney injury—e.g. increase in serum creatinine 2× of baseline, patients after invasive procedures such as surgery or contrast media application, acute infection (diabetic foot), abnormal acute urinary sediment findings or pathological sonography] were excluded. Albuminuria was determined by nephelometry the first morning through urine samples and is expressed in mg/L according to the recommendations of the German Diabetes Society. Estimated glomerular filtration rate (eGFR) was calculated according to the modified equation 7 MDRD formula [ 15 ]. Patients were grouped into the chronic kidney disease (CKD) stages according to the eGFR and presence of albuminuria [ 16 ]. Patients were considered as chronic if CKD stages were confirmed at two separate visits during 1 year. The majority of patients had no urinary sediment findings suggesting glomerulonephritis (e.g. a haematuria, acanthocytes), but not all patients had a recent urinary sediment analysis. SES was determined by the three components education, highest professional position achieved and household net income. Each component was assessed by a score of 1–7 points to generate a total score with a minimum of 3 and maximum of 21 points [ 17 ]. This score has been previously used in SES studies in Germany and is validated [ 17 ]. Patients were grouped in tertiles according to their SES. HbA1c levels were analysed by high-performance liquid chromatography. HbA1c values were adjusted to Diabetes Control and Complications Trial standards with an evaluated standardization procedure [ 18 ]. HbA1c is registered as absolute and relative value by dividing the absolute value by the mean of local healthy subjects. For each patient, the mean HbA1c of all visits was used for calculation. Analysis of HbA1c values was weighted because not every patient had the same number of HbA1c measurements [ 18 ].

Statistical analyses

For descriptive statistical analysis, means, SD, absolute and relative frequencies were calculated. Statistical analysis was performed with the t -test. For estimating possible relationships of eGFR ≤60 mL/min with SES, binary logistic regression models were constructed for all patients according to the tertiles of SES as the dependent variable. In another model, albuminuria (>20 mg/L) was considered as dependent variable. The multivariate models were stepwise backwards and forwards adjusted for hypertension (blood pressure > 140/70 mmHg), gender, duration of diabetes, mean adjusted HbA1c, body mass index (BMI), smoking and age [ 14 ]. A P <0.05 was considered significant. Statistical analysis was performed with SPSS 16.0 (SPSS, Chicago, IL) and SAS 9.1 (SAS Institute, Inc., Cary, NC).

Results

About 58.6% of all patients filled out the questionnaire completely (174 type 1 diabetes, 651 type 2 diabetes), 99.5% of all patients were members of the German ‘public’ mandatory health insurance and there was no difference between the SES tertiles making it extremely unlikely that patients with higher SES may have received better or other care outside the University Hospital, (for example, by private physicians). The most common reason for incomplete questionnaire was the refusal to give information concerning household net income or allowance to save these data in the electronic patient record. At the time of the last observation, the study population for types 1 and 2 diabetic patients had the characteristics shown in Table 1 . The studied group of 58.6% patients was representative for the whole population treated in our outpatient clinic ( n = 2452 patients, minimal changes were found in age with the whole collective’s mean age being 2 years younger than the study group, and duration of disease where the mean duration was 1 year less in the whole population compared with the study group, and no differences were found for blood pressure, HbA1c values and BMI). Data for grouped patients with types 1 and 2 diabetes and the tertiles SES are shown in Tables 2–4 . No significant difference in the number of visits to the outpatient clinic was found among the different SES tertiles indicating that the access to health care was similar for all patients. Because the number of type 1 diabetes patients was rather low, analysis of all (types 1 + 2 patients) as well as only type 2 patients was performed. Figure 1 shows the relationship between the CKD Stage (1–5) and the tertiles of SES for all (diabetes types 1 and 2) patients. It is clearly visible that the number of patients in CKD Stages 2–5 is increasing with decreasing SES. A similar relationship was found when only type 2 diabetes patients were analysed ( Figure 2 ). In contrast, no clear association between tertiles of the SES and CKD stage was found in patients with type 1 diabetes ( Figure 3 ). Multivariate regression analysis for all (types 1 + 2 diabetes) patients revealed that the SES as independent variable was highly significantly correlated with an estimated GFR <60 mL/min (see Table 5 ), whereas such a relationship was not found in type 1 diabetes patients ( Table 1 , Figure 3 ). However, in type 1 diabetic patients, a lower SES was significantly correlated with the degree of albuminuria ( Table 3 ).

Table 1.

Total group of 825 patients with type 1 ( n = 174) and 2 ( n = 651) diabetes according to tertiles in SES

Tertiles of socioeconomic score ≤9 10–12 ≥13  P-value a 
n 307 269 249  
Socioeconomic score 7.9 ± 1.4 (3–9) 10.8 ± 0.8 (10–12) 15.4 ± 2.2 (13–21)  
Gender (female/male) 156 (51)/151 (49) 129 (48)/140 (52) 92 (37)/157 (63) 0.003 
Age (years) 64.5 ± 11.8 (18–88) 63.1 ± 13.5 (20–92) 60.6 ± 13.8 (24–91) 0.002 
Duration of diabetes (years) 15.9 ± 9.5 (0-57) 15.9 ± 11.8 (0–67) 14.5 ± 10.2 (0–61) 0.131 
Adjusted mean HbA1c (%) 7.8 ± 1.1 (5.2–13.9) 7.6 ± 0.9 (5.7–11.0) 7.5 ± 0.9 (5.4–11.7) <0.001 
BMI (kg/m 2 )  32.0 ± 6.2 (20–62) 31.2 ± 6.2 (19–67) 28.7 ± 5.0 (18–45) <0.001 
Mean systolic blood pressure (mmHg) 145 ± 13 (107–190) 144 ± 13 (111–195) 141 ± 14 (105–182) 0.015 
Mean diastolic blood pressure (mmHg) 81 ± 8 (58–110) 82 ± 9 (62–109) 83 ± 9 (61–119) 0.781 
Hypertension (yes/no) 201 (66.5)/106 (34.5) 171 (64)/98 (36) 143 (57)/106 (43) 0.138 
Smoking (yes/no) 119 (39)/188 (61) 125 (46.5)/144 (53.5) 101 (41)/148 (59) 0.157 
Mean serum creatinine (μmol/L) 96 ± 47 (37–629) 92 ± 28 (50–273) 89 ± 29 (54–346) 0.190 
Mean eGFR (mL/min) 84 ± 28 (13–268) 88 ± 27 (27–177) 91 ± 22 (23–137) 0.001 
Mean urinary albumin (mg/L) 99 ± 267 (0–2380) 113 ± 354 (0–3638) 60 ± 261 (0–3545) 0.006 
Tertiles of socioeconomic score ≤9 10–12 ≥13  P-value a 
n 307 269 249  
Socioeconomic score 7.9 ± 1.4 (3–9) 10.8 ± 0.8 (10–12) 15.4 ± 2.2 (13–21)  
Gender (female/male) 156 (51)/151 (49) 129 (48)/140 (52) 92 (37)/157 (63) 0.003 
Age (years) 64.5 ± 11.8 (18–88) 63.1 ± 13.5 (20–92) 60.6 ± 13.8 (24–91) 0.002 
Duration of diabetes (years) 15.9 ± 9.5 (0-57) 15.9 ± 11.8 (0–67) 14.5 ± 10.2 (0–61) 0.131 
Adjusted mean HbA1c (%) 7.8 ± 1.1 (5.2–13.9) 7.6 ± 0.9 (5.7–11.0) 7.5 ± 0.9 (5.4–11.7) <0.001 
BMI (kg/m 2 )  32.0 ± 6.2 (20–62) 31.2 ± 6.2 (19–67) 28.7 ± 5.0 (18–45) <0.001 
Mean systolic blood pressure (mmHg) 145 ± 13 (107–190) 144 ± 13 (111–195) 141 ± 14 (105–182) 0.015 
Mean diastolic blood pressure (mmHg) 81 ± 8 (58–110) 82 ± 9 (62–109) 83 ± 9 (61–119) 0.781 
Hypertension (yes/no) 201 (66.5)/106 (34.5) 171 (64)/98 (36) 143 (57)/106 (43) 0.138 
Smoking (yes/no) 119 (39)/188 (61) 125 (46.5)/144 (53.5) 101 (41)/148 (59) 0.157 
Mean serum creatinine (μmol/L) 96 ± 47 (37–629) 92 ± 28 (50–273) 89 ± 29 (54–346) 0.190 
Mean eGFR (mL/min) 84 ± 28 (13–268) 88 ± 27 (27–177) 91 ± 22 (23–137) 0.001 
Mean urinary albumin (mg/L) 99 ± 267 (0–2380) 113 ± 354 (0–3638) 60 ± 261 (0–3545) 0.006 
a

Means (minimum−maximum) or number (%), as appropriate 1 according to Kruskal–Wallis test or X 2 test (Fisher exact test).

Table 2.

Data of 651 patients with diabetes mellitus type 2 according to tertiles in SES

Tertiles of socioeconomic score ≤9 10–12 ≥13  P-value a 
n 256 214 181  
Socioeconomic score 8.0 ± 1.3 (3–9) 10.9 ± 0.8 (10–12) 15.5 ± 2.2 (13–21)  
Gender (female/male) 133 (52)/123 (48) 101 (47)/113 (53) 66 (36.5)/115 (63.5) 0.006 
Age (years) 65.9 ± 10.5 (27–88) 66.2 ± 10.8 (33–92) 65.3 ± 10.5 (32–91) 0.442 
Duration of diabetes (years) 15.3 ± 8.6 (0–43) 13.6 ± 9.5 (0–40) 13.0 ± 9.0 (0–38) 0.004 
Adjusted mean HbA1c (%) 7.7 ± 1.0 (5.2–12.8) 7.6 ± 0.9 (5.8–11.0) 7.4 ± 0.9 (5.4–10.4) 0.001 
BMI (kg/m 2 )  33.0 ± 6.0 (21–62) 32.2 ± 6.2 (21–67) 30.1 ± 4.9 (18–45) <0.001 
Mean systolic blood pressure (mmHg) 146 ± 13 (107–190) 145 ± 13 (117–195) 145 ± 13 (108–182) 0.590 
Mean diastolic blood pressure (mmHg) 82 ± 8 (58–110) 83 ± 9 (62–109) 84 ± 9 (65–119) 0.203 
Hypertension (yes/no) 178 (69.5)/78 (30.5) 149 (70)/65 (30) 125 (69)/56 (31) 0.988 
Smoking (yes/no) 94 (37)/162 (63) 99 (46)/115 (54) 79 (44)/102 (56) 0.093 
Mean serum creatinine (μmol/L) 97 ± 50 (37–629) 94 ± 31 (50–273) 91 ± 29 (56–346) 0.625 
Mean eGFR (mL/min) 84 ± 29 (13–268) 88 ± 28 (27–177) 88 ± 22 (23–137) 0.026 
Mean urinary albumin (mg/L) 96 ± 241 (0–2380) 128 ± 387 (0–3638) 59 ± 159 (0–1901) 0.676 
Tertiles of socioeconomic score ≤9 10–12 ≥13  P-value a 
n 256 214 181  
Socioeconomic score 8.0 ± 1.3 (3–9) 10.9 ± 0.8 (10–12) 15.5 ± 2.2 (13–21)  
Gender (female/male) 133 (52)/123 (48) 101 (47)/113 (53) 66 (36.5)/115 (63.5) 0.006 
Age (years) 65.9 ± 10.5 (27–88) 66.2 ± 10.8 (33–92) 65.3 ± 10.5 (32–91) 0.442 
Duration of diabetes (years) 15.3 ± 8.6 (0–43) 13.6 ± 9.5 (0–40) 13.0 ± 9.0 (0–38) 0.004 
Adjusted mean HbA1c (%) 7.7 ± 1.0 (5.2–12.8) 7.6 ± 0.9 (5.8–11.0) 7.4 ± 0.9 (5.4–10.4) 0.001 
BMI (kg/m 2 )  33.0 ± 6.0 (21–62) 32.2 ± 6.2 (21–67) 30.1 ± 4.9 (18–45) <0.001 
Mean systolic blood pressure (mmHg) 146 ± 13 (107–190) 145 ± 13 (117–195) 145 ± 13 (108–182) 0.590 
Mean diastolic blood pressure (mmHg) 82 ± 8 (58–110) 83 ± 9 (62–109) 84 ± 9 (65–119) 0.203 
Hypertension (yes/no) 178 (69.5)/78 (30.5) 149 (70)/65 (30) 125 (69)/56 (31) 0.988 
Smoking (yes/no) 94 (37)/162 (63) 99 (46)/115 (54) 79 (44)/102 (56) 0.093 
Mean serum creatinine (μmol/L) 97 ± 50 (37–629) 94 ± 31 (50–273) 91 ± 29 (56–346) 0.625 
Mean eGFR (mL/min) 84 ± 29 (13–268) 88 ± 28 (27–177) 88 ± 22 (23–137) 0.026 
Mean urinary albumin (mg/L) 96 ± 241 (0–2380) 128 ± 387 (0–3638) 59 ± 159 (0–1901) 0.676 
a

Means (minimum−maximum) or number (%), as appropriate 1 according to Kruskal–Wallis test or X 2 test (Fisher exact test).

Table 3.

Data of 174 patients with diabetes mellitus type 1 according to tertiles in SES

Tertiles of socioeconomic score ≤9 10–12 ≥13  P-value a 
n 51 55 68  
Socioeconomic score 7.8 ± 1.6 (3–9) 10.8 ± 0.8 (10–12) 15.4 ± 2.2 (13–20)  
Gender (female/male) 23 (45)/28 (55) 28 (51)/27 (49) 26 (38)/42 (62) 0.373 
Age (years) 57.5 ± 15.3 (18–82) 51.0 ± 16.0 (20–77) 48.2 ± 14.0 (24–76) 0.002 
Duration of diabetes (years) 18.6 ± 12.8 (0–57) 25.3 ± 15.1 (1–67) 18.6 ± 11.9 (1–61) 0.016 
Adjusted mean HbA1c (%) 8.2 ± 1.3 (5.8–13.9) 7.9 ± 0.9 (5.7–10.2) 7.6 ± 1.0 (5.8–11.7) 0.016 
BMI (kg/m 2 )  26.5 ± 4.0 (20–40) 27.1 ± 4.4 (19–40) 24.5 ± 2.9 (20–32) 0.009 
Mean systolic blood pressure (mmHg) 138 ± 11 (111–157) 137 ± 14 (111–176) 132 ± 12 (105–171) 0.021 
Mean diastolic blood pressure (mmHg) 80 ± 6 (66–94) 79 ± 6 (68–92) 79 ± 7 (61–95) 0.439 
Hypertension (yes/no) 23 (45)/28 (55) 22 (40)/33 (60) 18 (26.5)/50 (73.5) 0.086 
Smoking (yes/no) 25 (49)/26 (51) 26 (47)/29 (52) 22 (32)/46 (68) 0.118 
Mean serum creatinine (μmol/L) 92 ± 31 (50–212) 83 ± 25 (53–222) 84 ± 28 (54–268) 0.241 
Mean eGFR (mL/min) 89 ± 27 (26–141) 95 ± 23 (32–144) 97 ± 21 (34–135) 0.206 
Mean urinary albumin (mg/L) 115 ± 377 (1–2283) 54 ± 170 (3–1211) 64 ± 429 (0–3545) <0.001 
Tertiles of socioeconomic score ≤9 10–12 ≥13  P-value a 
n 51 55 68  
Socioeconomic score 7.8 ± 1.6 (3–9) 10.8 ± 0.8 (10–12) 15.4 ± 2.2 (13–20)  
Gender (female/male) 23 (45)/28 (55) 28 (51)/27 (49) 26 (38)/42 (62) 0.373 
Age (years) 57.5 ± 15.3 (18–82) 51.0 ± 16.0 (20–77) 48.2 ± 14.0 (24–76) 0.002 
Duration of diabetes (years) 18.6 ± 12.8 (0–57) 25.3 ± 15.1 (1–67) 18.6 ± 11.9 (1–61) 0.016 
Adjusted mean HbA1c (%) 8.2 ± 1.3 (5.8–13.9) 7.9 ± 0.9 (5.7–10.2) 7.6 ± 1.0 (5.8–11.7) 0.016 
BMI (kg/m 2 )  26.5 ± 4.0 (20–40) 27.1 ± 4.4 (19–40) 24.5 ± 2.9 (20–32) 0.009 
Mean systolic blood pressure (mmHg) 138 ± 11 (111–157) 137 ± 14 (111–176) 132 ± 12 (105–171) 0.021 
Mean diastolic blood pressure (mmHg) 80 ± 6 (66–94) 79 ± 6 (68–92) 79 ± 7 (61–95) 0.439 
Hypertension (yes/no) 23 (45)/28 (55) 22 (40)/33 (60) 18 (26.5)/50 (73.5) 0.086 
Smoking (yes/no) 25 (49)/26 (51) 26 (47)/29 (52) 22 (32)/46 (68) 0.118 
Mean serum creatinine (μmol/L) 92 ± 31 (50–212) 83 ± 25 (53–222) 84 ± 28 (54–268) 0.241 
Mean eGFR (mL/min) 89 ± 27 (26–141) 95 ± 23 (32–144) 97 ± 21 (34–135) 0.206 
Mean urinary albumin (mg/L) 115 ± 377 (1–2283) 54 ± 170 (3–1211) 64 ± 429 (0–3545) <0.001 
a

Means (minimum−maximum) or number (%), as appropriate1 according to Kruskal–Wallis test or X 2 test (Fisher exact test).

Table 4.

Multivariate logistic regression analysis in the total group of 651 type 2 and 174 type 1 diabetic patients, adjusted for the covariates given, respectively: eGFR <60 mL/min/1.73m 2 as dependent variable

Indicators Regression coefficient Standard error Odds ratio 95% confidence interval P-value 
Socioeconomic score a versus highest tertile 3 (≥13), respectively       
    1 (lowest tertile, ≤9) 1.086 0.344 2.96 1.51–5.82 0.002 
    2 (10–12) 0.656 0.365 1.93 0.94–3.94 0.072 
Age a (per year increase)  0.146 0.017 1.16 1.12–1.20 <0.001 
Urinary albumin a (mg/L, per doubling)  0.386 0.063 1.47 1.30–1.66 <0.001 
Smoking a (yes versus no)  −0.599 0.268 0.55 0.33–0.93 0.026 
BMI b (per increase of 1 kg/m 2 )  0.053 0.025 1.05 1.01–1.11 0.033 
Type of diabetes b (type 2 versus type 1 as reference)  −0.819 0.444 0.44 0.19–1.05 0.065 
Adjusted mean HbA1c b (per increase of 1%)  −0.240 0.144 0.79 0.59–1.04 0.096 
Duration of diabetes b (per year increase)  −0.003 0.013 1.00 0.97–1.02 0.792 
Hypertension c (yes versus no) (blood pressure ≥ 140/90 mmHg)  −0.031 0.289 0.97 0.55–1.71 0.914 
Indicators Regression coefficient Standard error Odds ratio 95% confidence interval P-value 
Socioeconomic score a versus highest tertile 3 (≥13), respectively       
    1 (lowest tertile, ≤9) 1.086 0.344 2.96 1.51–5.82 0.002 
    2 (10–12) 0.656 0.365 1.93 0.94–3.94 0.072 
Age a (per year increase)  0.146 0.017 1.16 1.12–1.20 <0.001 
Urinary albumin a (mg/L, per doubling)  0.386 0.063 1.47 1.30–1.66 <0.001 
Smoking a (yes versus no)  −0.599 0.268 0.55 0.33–0.93 0.026 
BMI b (per increase of 1 kg/m 2 )  0.053 0.025 1.05 1.01–1.11 0.033 
Type of diabetes b (type 2 versus type 1 as reference)  −0.819 0.444 0.44 0.19–1.05 0.065 
Adjusted mean HbA1c b (per increase of 1%)  −0.240 0.144 0.79 0.59–1.04 0.096 
Duration of diabetes b (per year increase)  −0.003 0.013 1.00 0.97–1.02 0.792 
Hypertension c (yes versus no) (blood pressure ≥ 140/90 mmHg)  −0.031 0.289 0.97 0.55–1.71 0.914 
a

Resulting from the final forward model, all covariates of this model showed comparable results in the final backward model, too.

b

Resulting from the final backward model.

c

Resulting from the initial backward model.

Fig. 1.

Graphic demonstration of the number of patients in the various CKD stages according to the SES tertiles for the whole population (types 1:174 and 2:651 diabetic patients). It is clearly visible that the number of patients in CKD Stages 2–5 is increasing with decreasing SES.

Fig. 1.

Graphic demonstration of the number of patients in the various CKD stages according to the SES tertiles for the whole population (types 1:174 and 2:651 diabetic patients). It is clearly visible that the number of patients in CKD Stages 2–5 is increasing with decreasing SES.

Fig. 2.

Graphic demonstration of the number of patients in the various CKD stages according to the SES tertiles for only type 2 diabetic patients. Similar to the whole population shown in Figure 1 , a clear relationship between the SES and CKD stages is visible ( n = 174 patients).

Fig. 2.

Graphic demonstration of the number of patients in the various CKD stages according to the SES tertiles for only type 2 diabetic patients. Similar to the whole population shown in Figure 1 , a clear relationship between the SES and CKD stages is visible ( n = 174 patients).

Fig. 3.

Relationship between the number of type 1 diabetic patients, the SES and CKD stages. In contrast to all patient (types 1 and 2) as well as only type 2 patients, no clear relationship exits between SES and renal function. However, albuminuria was significantly higher in type 1 diabetic patients with low SES (see Table 3 ).

Fig. 3.

Relationship between the number of type 1 diabetic patients, the SES and CKD stages. In contrast to all patient (types 1 and 2) as well as only type 2 patients, no clear relationship exits between SES and renal function. However, albuminuria was significantly higher in type 1 diabetic patients with low SES (see Table 3 ).

Moreover, no association was found between albuminuria ≥ 20 mg/L as dependent variable and SES ( Table 6 ), although there was a close significant correlation with ‘classical’ risk factors for albuminuria (hypertension, smoking, adjusted mean HbA1c, duration of diabetes).

Table 5.

Multivariate logistic regression analysis in 651 type 2 diabetic patients, adjusted for the covariates given, respectively: eGFR <60 mL/min/1.73m 2 as dependent variable

Indicators Regression coefficient Standard error Odds ratio 95% confidence interval P-value 
SES a versus highest tertiles 3 (≥13), respectively       
    1 (lowest tertiles, ≤9) 0.993 0.376 2.70 1.29–5.64 0.008 
    2 (10–12) 0.764 0.388 2.15 1.00–4.59 0.049 
Age a (per year increase)  0.158 0.020 1.17 1.13–1.22 <0.001 
Urinary albumin a (mg/L, per doubling)  0.348 0.066 1.42 1.24–1.61 <0.001 
Smoking a (yes versus no)  −0.610 0.284 0.54 0.31–0.95 0.032 
BMI a (per increase of 1 kg/m 2 )  0.053 0.025 1.06 1.00–1.11 0.037 
Adjusted mean HbA1c b (per increase of 1%)  −0.200 0.148 0.82 0.61–1.10 0.178 
Duration of diabetes b (per year increase)  0.010 0.015 1.01 0.98–1.04 0.506 
Hypertension b (yes versus no) (blood pressure ≥ 140/90 mmHg)  −0.122 0.302 0.89 0.49–1.60 0.685 
Indicators Regression coefficient Standard error Odds ratio 95% confidence interval P-value 
SES a versus highest tertiles 3 (≥13), respectively       
    1 (lowest tertiles, ≤9) 0.993 0.376 2.70 1.29–5.64 0.008 
    2 (10–12) 0.764 0.388 2.15 1.00–4.59 0.049 
Age a (per year increase)  0.158 0.020 1.17 1.13–1.22 <0.001 
Urinary albumin a (mg/L, per doubling)  0.348 0.066 1.42 1.24–1.61 <0.001 
Smoking a (yes versus no)  −0.610 0.284 0.54 0.31–0.95 0.032 
BMI a (per increase of 1 kg/m 2 )  0.053 0.025 1.06 1.00–1.11 0.037 
Adjusted mean HbA1c b (per increase of 1%)  −0.200 0.148 0.82 0.61–1.10 0.178 
Duration of diabetes b (per year increase)  0.010 0.015 1.01 0.98–1.04 0.506 
Hypertension b (yes versus no) (blood pressure ≥ 140/90 mmHg)  −0.122 0.302 0.89 0.49–1.60 0.685 
a

Resulting from the final forward model, all covariates of this model showed comparable results in the final backward model, too.

b

Resulting from the last step of the backward model containing this covariate, respectively.

Table 6.

Multivariate logistic regression analysis in the total group of 651 type 2 and 174 type 1 diabetic patients, adjusted for the covariates given, respectively: mean urinary albumin ≥ 20 mg/L as dependent variable

Indicators Regression coefficient Standard error Odds ratio 95% confidence interval P-value 
Socioeconomic score a versus highest tertiles 3 (≥13), respectively       
    1 (lowest tertiles, ≤9) 0.072 0.194 1.08 0.74–1.57 0.709 
    2 (10–12) 0.128 0.195 1.14 0.78–1.67 0.514 
Age a (per year increase)  0.004 0.009 1.004 0.987–1.022 0.632 
eGFR b (per increase of 1 mL/min)  −0.010 0.003 0.99 0.984–0.996 0.001 
Smoking b (yes versus no)  0.556 0.158 1.74 1.281–2.375 <0.001 
BMI b (per increase of 1 kg/m 2 )  0.032 0.014 1.03 1.005–1.061 0.022 
Type of diabetes b (type 2 versus type 1 as reference)  1.133 0.229 3.10 1.982–4.862 <0.001 
Adjusted mean HbA1c b (per increase of 1%)  0.441 0.083 1.55 1.320–1.831 <0.001 
Duration of diabetes a (per year increase)  0.013 0.008 1.01 0.998–1.029 0.081 
Hypertension b (yes versus no) (blood pressure ≥ 140/90 mmHg)  0.553 0.163 1.73 1.263–2.392 0.001 
Indicators Regression coefficient Standard error Odds ratio 95% confidence interval P-value 
Socioeconomic score a versus highest tertiles 3 (≥13), respectively       
    1 (lowest tertiles, ≤9) 0.072 0.194 1.08 0.74–1.57 0.709 
    2 (10–12) 0.128 0.195 1.14 0.78–1.67 0.514 
Age a (per year increase)  0.004 0.009 1.004 0.987–1.022 0.632 
eGFR b (per increase of 1 mL/min)  −0.010 0.003 0.99 0.984–0.996 0.001 
Smoking b (yes versus no)  0.556 0.158 1.74 1.281–2.375 <0.001 
BMI b (per increase of 1 kg/m 2 )  0.032 0.014 1.03 1.005–1.061 0.022 
Type of diabetes b (type 2 versus type 1 as reference)  1.133 0.229 3.10 1.982–4.862 <0.001 
Adjusted mean HbA1c b (per increase of 1%)  0.441 0.083 1.55 1.320–1.831 <0.001 
Duration of diabetes a (per year increase)  0.013 0.008 1.01 0.998–1.029 0.081 
Hypertension b (yes versus no) (blood pressure ≥ 140/90 mmHg)  0.553 0.163 1.73 1.263–2.392 0.001 
a

Resulting from the last step of the backward model containing this covariate, respectively.

b

Resulting from the final forward model, all covariates of this model showed comparable results in the final backward model, too.

Discussion

Our study suggests that SES is an independent factor for low renal function in a population of type 2 diabetic patients, but not for type 1 diabetes patients. Since a similar relationship was found in all patients (combining types 1 and 2 diabetic patients), we cannot exclude that the relatively low number of type 1 diabetes patients induced a type 2 error in our analysis.

A relationship between SES and progression of renal disease and/or degree of albuminuria has been found in several studies in various countries [ 3–10 ]. Such relationships exist for diabetic as well as nondiabetic nephropathies [ 19 , 20 ]. In addition, the risk of diabetes manifestations also increases with a low SES [ 21 ]. However, minority populations with lower SES in other countries such as African-Americans and Native American Indians may have certain genetic risk factors making it difficult to separate between genetic and other factors [ 4 , 22 , 23 ]. Moreover, patients with lower SES are often uninsured and have a restricted access to health facilities in certain countries (e.g. USA) [ 4 , 8 ]. Therefore, a major advantage of our study is a relatively homologous population of Caucasian patients. Moreover, in Germany, mandatory health insurance exists with only a relatively low number of uninsured people [ 24 ]. In addition, the number of patient visits to the outpatient clinic was independent from the SES in our study. Since the majority of our patients were members of the public mandatory health insurance and were not members of ‘private’ insurance companies, it is very unlikely that patients with higher SES were treated outside the university by private physicians and may have received better care (referral bias). Therefore, our study excludes important confounding factors of previous studies which show a relationship between SES and renal disease.

Nevertheless, one may critically ask what is the advantage of our relatively small single-centre study compared with the analysis from Sweden with almost 2000 patients [ 6 ]. We believe that our single-centre study has clear advantages. First, we focused only on diabetic patients who are treated with a standardized regime in a single centre and we believe that even minor changes in therapy (e.g. use of angiotensin I conversion enzyme inhibitor versus other antihypertensive drugs) can influence the outcome in multicentre studies, second, calculation of eGFR and classification of CKD depend on the correct measurement of serum creatinine and current recommendations have been made to improve this critical point [ 25 , 26 ]. A single-centre study such as ours offers the advantage of measurement of laboratory parameters in a central laboratory under standardized conditions that may not be achievable if data from multiple practitioners are used who send their specimens in from various laboratories.

There is clearly a widening of the social gradient in many European countries including Germany [ 21 , 27 , 28 ]. The present study was performed in Jena, a university-oriented small town with 105 000 inhabitants of which ∼20 000 are students localized in the former East Germany. The average gross domestic income is 33 880 Euro (average in Germany: 28 534 Euro). The unemployment rate is 7.4% for Jena (September 2010) and for Germany 6.9%. Thus, our study was not performed in a very ‘poor’ or problematic city with many structural problems. Nevertheless, our analysis revealed an obvious relationship between low SES and renal function in patients with mainly type 2 diabetes. The reasons for this relationship are presumably multifactorial, but a few suggestions can be made. All patients with diabetes being in the lowest tertile of SES were significantly older and exhibited a higher BMI compared with a higher SES. In addition, by looking only at patients with type 2 diabetes, those with a lower SES had a significantly worse HbA1C and were significantly often of female gender.

Women may exhibit a lower SES than men because of likely lower income and more unemployment making an independent analysis of gender difficult.

Obesity is a well-known factor for progression of renal disease by various mechanisms (e.g. hyperfiltration, direct effects of leptin; for review see [ 29 ]). On the other hand, patients with low SES have a higher BMI [ 30 , 31 ]. Interestingly, smoking, a well-known factor for progression of renal disease, was not significantly associated with a low SES in our study [ 32 , 33 ]. Similarly, blood pressure and the presence of hypertension, a well-known factor contributing to progression of renal disease in diabetes [ 34 ], were not significantly different in the SES tertiles. However, the significant difference between low SES and higher SES and the HbA1c in the group of type 2 diabetic patients may suggest compliance problems.

Our study has obvious limitations. The study population of a university outpatient department may not be representative of other diabetic patients treated in primary care. In comparison to the German Disease Management Programme North Rhine 2007 [ 35 ] with 15 540 patients with type 1 and 330 000 with type 2 diabetes, our patients with type 1 diabetes were older (52 versus 43 years), had a longer duration of diabetes (20 versus 16 years) and a lower HbA1c (7.5 versus 7.8%). Age in people with diabetes type 2 in our study was comparable to the 330 000 unselected patients in the disease management program (66 versus 67 years) but diabetes duration in our study was longer (14 versus 8 years). A further limitation of the present study is that the analysis is based on current and retrospective data. Since only very few CKD Stage 5 patients were included in our study, we believe that a major influence of SES change due to loss of work and/or inability to work after initiating dialysis is unlikely. Nevertheless, changes in SES during lifetime could occur and were not assessed in our study. Therefore, a selection bias cannot be totally excluded, but seems unlikely. Nevertheless, our result should be interpreted cautiously. Although we used a previously validated instrument to measure SES [ 17 ], problems may occur with this approach. For example, although elderly people in Germany often experience a decline in their SES, and indeed, patients with the highest SES were significantly younger in our study, that is not necessarily true for everybody and linking the SES with education and income may be problematic in patients ≥65 years. Finally, selective loss of patients with rapidly progressive diseases resulting in early death may occur.

Finally, one may asked what is the value of our study if proteinuria does not correlate with SES. All included patients had albuminuria and classification done according to the CKD stages. Albuminuria is, particularly in diabetic patients an unspecific finding, may change due to good therapy, and is after all a surrogate parameter. Moreover, it has been recently appreciated that there is a wide variation between albuminuria and not all CKD in type 2 diabetes is proteinuric [ 36–39 ]. Recent studies provide strong arguments that small vessel disease may cause progressive loss of renal function in these patients [ 39 ], but we have not studied this hypothesis further in our collective. Therefore, we believe our study clearly demonstrated an association between renal function and SES but not albuminuria and SES makes our conclusion even stronger.

Part of the study was supported by a grant from the Deutsche Forschungsgemeinschaft to G.W.

Conflict of interest statement . None declared.

References

1.
Wolf
G
Müller
N
Tschauner
T
, et al.  . 
Prevalence of renal insufficiency in the diabetes TÜV of the Deutsche BKK
Med Klin (Munich)
 , 
2006
, vol. 
101
 (pg. 
441
-
447
)
2.
Ritz
E
Orth
SR
Nephropathy in patients with type 2 diabetes mellitus
N Engl J Med
 , 
1999
, vol. 
341
 (pg. 
1127
-
1133
)
3.
Bello
AK
Peters
J
Rigby
J
, et al.  . 
Socioeconomic status and chronic kidney disease at presentation to a renal service in the United Kingdom
Clin J Am Soc Nephrol
 , 
2008
, vol. 
3
 (pg. 
1316
-
1323
)
4.
Norris
K
Nissenson
AR
Race, Gender, and socioeconomic disparities in CKD in the United States
J Am Soc Nephrol
 , 
2008
, vol. 
19
 (pg. 
1261
-
1270
)
5.
Sabanayagam
C
Shankar
A
Saw
SM
, et al.  . 
Socioeconomic status and microalbuminuria in an Asian population
Nephrol Dial Transplant
 , 
2009
, vol. 
24
 (pg. 
123
-
129
)
6.
Fored
CM
Ejerblad
E
Fryzek
JP
, et al.  . 
Socio-economic status and chronic renal failure a population-based case-control study in Sweden
Nephrol Dial Transplant
 , 
2003
, vol. 
18
 (pg. 
82
-
88
)
7.
Ward
MM
Socioeconomic status and the incidence of ESRD
Am J Kidney Dis
 , 
2008
, vol. 
51
 (pg. 
563
-
572
)
8.
Lopes
AA
End-stage renal disease due to diabetes in racial/ethnic minorities and disadvantaged populations
Ethn Dis
 , 
2009
, vol. 
19
 
Suppl 1
(pg. 
S47
-
S51
)
9.
Martin
D
Tarreen
N
Zadshir
A
, et al.  . 
The association of poverty with the prevalence of albuminuria: data from the Third National Health and Nutrition Examination Survey (NHANES III)
Am J Kidney Dis
 , 
2006
, vol. 
47
 (pg. 
965
-
971
)
10.
Udayaraj
UP
Ben-Shlomo
Y
Roderick
P
, et al.  . 
Ethnicity, socioeconomic status, and attainment of clinical practice guideline standards in dialysis patient in the United kingdom
Clin J Am Soc Nephrol
 , 
2009
, vol. 
4
 (pg. 
979
-
987
)
11.
Navaneethan
SD
Aloudat
S
Singh
S
A systematic review of patient and health system characteristics associated with late referral in chronic kidney disease
BMC Nephrol
 , 
2008
, vol. 
9
 pg. 
3
  
(doi: 10.1186/471-2369-9-3)
12.
Lanting
LC
Joung
IMA
Mackenbach
JP
, et al.  . 
Ethnic differences in mortality, end-stage complications, and quality of care among diabetic patients
Diabetes Care
 , 
2005
, vol. 
28
 (pg. 
2280
-
2288
)
13.
Wolf
G
Müller
N
Hunger-Battefeld
W
, et al.  . 
Hemoglobin concentrations are closely linked to renal function in patients with type 1 or 2 diabetes mellitus
Kidney Blood Press Res
 , 
2008
, vol. 
31
 (pg. 
313
-
321
)
14.
Wolf
G
Müller
N
Busch
M
, et al.  . 
Diabetic foot syndrome and renal function in type 1 and 2 diabetes mellitus show close association
Nephrol Dial Transplant
 , 
2009
, vol. 
24
 (pg. 
1896
-
1901
)
15.
Levey
AS
Bosch
JP
Breyer-Lewis
J
, et al.  . 
A more accurate method to estimate glomerular filtration rate from serum creatinine: a new equation
Ann Intern Med
 , 
1999
, vol. 
130
 (pg. 
461
-
470
)
16.
National Kidney Foundation
K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification
Am J Kidney Dis
 , 
2002
, vol. 
39
 
Suppl 2
(pg. 
S1
-
S266
)
17.
Dulon
M
Bardehle
D
Blettner
M
Assessing social inequality in microcensus data and German National Health Examination Survey
Gesundheitswesen
 , 
2003
, vol. 
65
 (pg. 
629
-
635
)
18.
Müller
UA
Femerling
M
Reinauer
KM
, et al.  . 
Intensified treatment and education of type 1 diabetes as clinical routine. A nationwide quality-circle experience in Germany. ASD (the working groups on structured diabetes therapy of the German diabetes association)
Diabetes Care
 , 
1999
, vol. 
22
 
Suppl 2
(pg. 
B29
-
B34
)
19.
Ward
MM
Access to care and the incidence of end-stage renal disease due to diabetes
Diabetes Care
 , 
2009
, vol. 
32
 (pg. 
1032
-
1036
)
20.
Shoham
DA
Vupputuri
S
Kshirsagar
AV
Chronic kidney disease and life course socioeconomic status: a review
Adv Chronic Kidney Dis
 , 
2005
, vol. 
12
 (pg. 
56
-
63
)
21.
Icks
A
Moebus
S
Feuersenger
A
, et al.  . 
Diabetes prevalence and association with social status-widening of a social gradient? German national heath surveys 1990–1992 and 1998
Diabetes Res Clin Pract
 , 
2007
, vol. 
78
 (pg. 
293
-
297
)
22.
Divers
J
Freedman
BI
Susceptibility genes in common complex kidney disease
Curr Opin Nephrol Hypertens
 , 
2010
, vol. 
19
 (pg. 
79
-
84
)
23.
Hanson
RL
Millis
MP
Young
NJ
, et al.  . 
ELMO1 variants and susceptibility to diabetic nephropathy in American Indians
Mol Genet Metab
 , 
2010
, vol. 
101
 (pg. 
383
-
390
)
24.
Bruan
S
Prezler
A
Mittendorf
T
, et al.  . 
Appraisal of resource use in the German heath-care system from the perspective of the statutory health insurance
Gesundheitswesen
 , 
2009
, vol. 
71
 (pg. 
19
-
23
)
25.
Stevens
LA
Stoycheff
N
Standardization of serum creatinine and estimated GFR in the kidney early evaluation program (KEEP)
Am J Kidney Dis
 , 
2008
, vol. 
51
 (pg. 
S77
-
S82
)
26.
Myers
GL
Miller
WG
Coresh
J
, et al.  . 
Recommendations for improving serum creatinine measurement: a report from the laboratory working group of the national kidney disease education program
Clin Chem
 , 
2006
, vol. 
52
 (pg. 
5
-
18
)
27.
Mackenbach
JP
Stirbu
I
Roskam
AJR
, et al.  . 
Socioeconomic inequalities in health in 22 European Countries
N Engl J Med
 , 
2008
, vol. 
358
 (pg. 
2468
-
2481
)
28.
Roskam
AJR
Kunst
AE
Van Oyen
H
, et al.  . 
Comparative appraisal of educational inequalities and obesity among adult in 19 European countries
Int J Epidemiol
 , 
2009
, vol. 
39
 (pg. 
392
-
404
)
29.
Wolf
G
After all those fat years: renal consequences of obesity
Nephrol Dial Transplant
 , 
2003
, vol. 
18
 (pg. 
2471
-
2474
)
30.
McLaren
L
Socioeconomic status and Obesity
Epidemiol Rev
 , 
2007
, vol. 
29
 (pg. 
29
-
48
)
31.
Kuntz
B
Lampert
T
Socioeconomic factors and Obesity
Dtsch Arztebl Int
 , 
2010
, vol. 
107
 (pg. 
517
-
522
)
32.
Mühlhauser
I
Cigarette smoking and diabetes: an update
Diabet Med
 , 
1994
, vol. 
11
 (pg. 
336
-
343
)
33.
Orth
SR
Smoking and the kidney
J Am Soc Nephrol
 , 
1002
, vol. 
13
 (pg. 
1663
-
1672
)
34.
Wolf
G
Ritz
E
Diabetic nephropathy in type 2 diabetes prevention and patient management
J Am Soc Nephrol
 , 
2003
, vol. 
14
 (pg. 
1396
-
13405
)
35.
Hagen
B
Altenhofen
L
Haß
W
, et al.  . 
Qualitätssicherungsbericht 2007 Disease-Management-Programme in Nordrhein. Herausgeber Nordrheinische Gemeinsame Einrichtung Disease-Management-Programme GbR, Düsseldorf (Stand November 2008) Eigenverlag KV Nordrhein
 , 
2008
 
36.
Yokoyama
H
Sone
H
Oishi
M
, et al.  . 
Prevalence of albuminuria an renal insufficiency and associated clinical factors in type 2 diabetes: the Japan Diabetes Clinical Data Mangement study (JDDM15)
Nephrol Dial Transplant
 , 
2009
, vol. 
24
 (pg. 
1212
-
1219
)
37.
Kramer
HJ
Nguyen
QD
Curhan
G
, et al.  . 
Renal insufficiency in the absence of albuminuria and retinopathy among adults with type 2 diabetes mellitus
JAMA
 , 
2003
, vol. 
289
 (pg. 
3272
-
3277
)
38.
Thomas
MC
MacIsaac
RJ
Jerums
G
, et al.  . 
Nonalbuminuric rebal impairment in type 2 diavetic patients and in the general population (national evaluation of the frequency of renal impairment coexisting with NIDDM [NEFRON] 11)
Diabetes Care
 , 
2009
, vol. 
32
 (pg. 
1497
-
1502
)
39.
Uzu
T
Kida
Y
Shirahashi
N
, et al.  . 
Cerebral microvascular disease predicts renal failure in type 2 diabetes
J Am Soc Nephrol
 , 
2010
, vol. 
21
 (pg. 
520
-
526
)

Comments

0 Comments