C-Reactive Protein in Relation to the Development of Microalbuminuria in Type 1 Diabetes

RESEARCH DESIGN AND METHODS—

High-sensitivity CRP (hs-CRP) was measured in 329 blood samples collected longitudinally from 49 subjects (31 male) with MA (MA+) (means ± SD age 15.3 ± 3.6 years, diabetes duration 5.6 ± 2.8 years, and age at diagnosis 9.5 ± 4.1 years) and in 49 normoalbuminuric subjects (MA−) matched for age, sex, and diabetes duration from the Oxford Regional Prospective Study, an incipient cohort study established in 1986 to document the natural history of MA during childhood (8). The study was approved by regional ethics committees with written consent from the parents and assent from the children.

Annual assessment included measurements of height, weight, and three consecutive early-morning urine samples for determination of albumin-to-creatinine ratio. Annual blood samples were also collected for the central measurement of A1C, and additional blood samples were stored. MA was defined as albumin-to-creatinine ratio between 3.5 and 35 mg/mmol in men and 4.0 and 47 mg/mmol in women in two of three consecutive early-morning urine collections (8).

hs-CRP was measured in plasma samples by latex-enhanced nephelometry (CardioPhase hsCRP assay; Dade Behring, Milan, Italy) on a BN II Nephelometer (Dade Behring). The lower limit of detection of this assay was 0.18 mg/l. hs-CRP levels >10 mg/l were excluded from the analysis, as they are likely related to infections or other acute inflammatory processes (1,9).

BMI SD scores (SDS) were calculated from the U.K. growth reference charts (8). Statistical analyses were performed using SPSS version 11.5. Data are expressed as means ± SD or median (range). Differences between groups were tested by independent t test. Analysis of covariance was used to assess trends in hs-CRP. Associations between variables are expressed as regression coefficient B ± SE.

RESULTS—

Mean A1C levels were significantly higher in MA+ than in MA− subjects (10.9 ± 2 vs. 10.2 ± 1.4%; P = 0.03), whereas no differences were detected in demographic characteristics, BMI SDS, and insulin dose.

In univariate analysis, hs-CRP was significantly related to age (B ± SE 0.06 ± 0.01; P < 0.001), duration of diabetes (0.06 ± 0.01; P < 0.001), BMI SDS (0.26 ± 0.07; P < 0.001), and logA1C (−0.93 ± 0.40; P = 0.02), whereas there was a borderline relationship with insulin dose (0.28 ± 0.14; P = 0.05). In a multivariate model, the only factors independently related to hs-CRP were age (0.04 ± 0.006; P = 0.008) and BMI SDS (0.17 ± 0.08; P = 0.03). In a separate model including duration of diabetes instead of age, duration (0.04 ± 0.01; P = 0.009) and, again, BMI SDS were the only factors independently related to hs-CRP.

Overall mean hs-CRP levels were not different between MA+ and MA− subjects (median 1.0 mg/l [range 0.2–6.1] vs. 0.9 mg/l [0.2–4.0]). However, differences were detected when hs-CRP levels were analyzed in relation to time of MA onset. For this analysis, mean levels of hs-CRP before MA onset, at time of MA onset, and after onset of MA were calculated for MA+ and, at corresponding years, for MA− subjects. A progressive rise in hs-CRP levels was observed in the MA+ group (P = 0.003) (Fig. 1), with significant higher levels after the onset of MA compared with those before its onset (P = 0.003; age-adjusted P = 0.06). In contrast, no significant changes over time were detected in the MA− group (P = 0.3). After the onset of MA, hs-CRP levels were significantly higher in MA+ than in MA− subjects (1.9 mg/dl [0.2–9.8] vs. 1.1 mg/dl [0.2–6.4]; P = 0.02). These differences persisted after adjusting for possible confounders, such as age, diabetes duration, BMI SDS, and A1C (P = 0.036).

CONCLUSIONS—

The main finding of this study was that in young individuals with type 1 diabetes, levels of hs-CRP were significantly raised after the development of MA. This is compatible with reported data showing that MA is associated with a state of subclinical inflammation and endothelial dysfunction (10). In subjects with type 1 diabetes, hs-CRP levels are increased compared with those of healthy controls (2–6) and are related to signs of subclinical atherosclerosis (2) and to other markers of endothelial dysfunction (3–5). However, the role of CRP in relation to the development of diabetic nephropathy is not clearly defined. Data from previous studies have raised the question as to whether CRP is a cause or consequence of the vascular damage associated with MA (4,5,7). The results of the present study support the second option: inflammation is related to MA onset rather than a predicting risk for its development. A previous study conducted in adults with type 2 diabetes has shown increased CRP levels associated with macroalbuminuria but not with MA (11). In other populations with type 1 diabetes, a significant difference in CRP levels was detected between subjects with overt proteinuria and those with normoalbuminuria but not between microalbuminuric and normoalbuminuric subjects (4,6).

hs-CRP levels increased with age irrespective of MA status; this finding has been previously reported among children and adolescents (9). In our study, the relationship of hs-CRP levels with age, together with the significant relationship between hs-CRP and diabetes duration, might indicate an effect of the chronic metabolic derangements of diabetes on CRP production. A positive relationship was found between BMI and hs-CRP levels consistent with previous data from large pediatric populations (12). A possible explanation for this relationship is that the hepatic production of CRP is regulated by interleukin-6 and tumor necrosis factor-α, two cytokines produced by adipose tissue (1). However, our observed association between the development of MA and hs-CRP was independent of BMI, suggesting that the two are linked by a general state of inflammation.