Abstract
Background and aims. The risk of diabetic nephropathy (DN) increases with increase in intraglomerular pressure, which may partly be regulated by nitric oxide (NO). NO‐production can be affected by polymorphisms in the endothelial NO‐synthase gene (NOS3), hyperglycaemia and smoking. We therefore studied association between DN and two polymorphisms in NOS3, Glu298Asp and NOS4ab, in Caucasian type 1 diabetes (T1D) patients.
Patients and methods. A total of 1510 Finnish and Swedish T1D patients were included in a cross‐sectional case‐control study. Incipient DN was defined as an albumin excretion rate (AER) of 20–200 µg/min (n = 336). Overt DN = AER>200 µg/min or renal replacement therapy (n = 619). All patients with DN were considered as cases. The controls were T1D patients with diabetes duration ⩾20 years, AER<20 µg/min and without antihypertensive treatment (n = 555). The genetic markers studied were a 27 bp repeat (NOS4ab) and Glu298Asp (rs1799983).
Results. Age at onset of diabetes, male sex, duration of diabetes, HbA1c, blood pressure and smoking were assessed as possible confounders in the logistic regression analysis, which showed that homozygosity for the Glu‐allele of the Glu298Asp‐polymorphism was independently associated with increased risk of DN (OR = 1.46; 95% CI = 1.12–1.91). The variables smoking (OR = 2.13; 95% CI = 1.63–2.78), male sex (OR = 1.61; 95% CI = 1.23–2.10), HbA1c (OR per % increase above upper limit of the normal reference range = 1.02; 95% CI = 1.02–1.03), systolic (OR = 1.05; 95% CI = 1.04–1.06) and diastolic blood pressure (OR = 1.04; 95% CI = 1.02–1.05) also significantly and independently increased the risk of DN when taking age at diabetes onset and diabetes duration into account. The NOS4 a‐allele was not associated with DN.
Conclusions. The Glu/Glu‐genotype of the NOS3 Glu298Asp polymorphism may increase the risk of developing DN independently of other known risk factors.
| Abbreviations | ||
| T1D | = | type 1 diabetes |
| DN | = | diabetic nephropathy |
| NO | = | nitric oxide |
| eNOS | = | endothelial nitric oxide synthase |
| NOS3 | = | endothelial nitric oxide synthase gene |
| OR | = | odds ratio |
Introduction
One of the most serious complications of type 1 diabetes (T1D) is diabetic nephropathy (DN). DN depends on the diabetic milieu, but from family studies it has become clear that genetic factors contribute Citation1–3. Lifestyle‐related factors, such as smoking, may also be of importance Citation4,5. In T1D there is an increase in glomerular capillary pressure and plasma flow resulting in elevation of the glomerular filtration rate (GFR), which in turn can damage the glomerular cells and contribute to the initiation and progression of DN Citation6,7. Genes involved in the regulation of GFR are therefore plausible candidates for the susceptibility to develop DN. The vasodilating nitric oxide molecule (NO) may be important in the regulation of intraglomerular pressure since it is produced in vascular beds of the glomeruli. NO is produced by endothelial nitric oxide synthase (eNOS) Citation8, coded by the NOS3 gene, located on chromosome 7q36. This chromosomal region has been linked to hereditary variations in urinary albumin excretion in type 2 diabetes patients and their relatives Citation9. In experimental diabetes there is increased eNOS content in the afferent arterioles of the glomeruli, but normal eNOS content in the efferent arterioles Citation10. This can lead to increased blood flow into the glomeruli but unchanged outflow, which will induce glomerular hyperfiltration.
In intron 4 of the NOS3 gene there is a polymorphism, NOS4ab, that consists of four or five 27 bp repeats (designated a and b respectively). This polymorphism has been associated with variations in plasma NO levels Citation11,12 and some, but not all studies, showed association with DN Citation13–16. In exon 7, a single nucleotide polymorphism leads to a Glu‐Asp amino acid change in position 298 (rs1799983). Notably this polymorphism has been associated with essential hypertension, myocardial infarction and end stage renal disease (ESRD) in Japanese populations Citation17–19. However, no association with DN was found in American Caucasian T1D patients Citation14.
Smoking affects regulation of vascular tone Citation20,21 and is associated with DN Citation4,5. In addition, serum from smokers induces production of eNOS but at the same time reduces its activity Citation22. Since not all smoking T1D patients are affected by DN, the individual response to smoking may be genetically determined.
In the present study we examined a possible association between two polymorphisms in the NOS3 gene and DN, in a large group of T1D patients from Finland and Sweden. We also wanted to test the hypothesis that the effect of smoking on DN development can be increased by these polymorphisms.
Key messages
The Glu/Glu‐genotype of the Glu298Asp‐polymorphism in the NOS3 gene independently increases the risk of diabetic nephropathy in a Finnish and Swedish type 1 diabetic population.
The risk of diabetic nephropathy is also increased by smoking, but no significant interaction between the NOS3 gene and smoking could be detected.
Patients and methods
The study was approved by the Ethics Committees in both countries. All 1510 patients were of Caucasian origin, had T1D with onset before 30 years of age and were insulin‐dependent from onset. In Sweden (S) a total of 347 T1D patients, from the University Hospital in Umeå, and the Karolinska Hospital in Stockholm, participated in the study. The 1163 Finnish (F) T1D patients were collected nationwide by the Finnish Diabetic Nephropathy Study group (FinnDiane). An albumin excretion rate (AER) of 20–200 µg/min in at least two out of three consecutive overnight samples was defined as microalbuminuria (incipient nephropathy) and AER>200 µg/min as macroalbuminuria (overt nephropathy). Microalbuminuria was present in 72 Swedish and 264 Finnish participants, and macroalbuminuria in 32 Swedish and 360 Finnish participants. Patients with end stage renal disease (dialysis treatment or kidney transplantation) were also included as overt nephropathy cases (S: n = 46; F: n = 181). The control group included T1D patients of at least 20 years' duration, normal AER (<20 µg/min) and no antihypertensive treatment (S: n = 197; F: n = 358). AER was measured using either nephelometric or immunoturbidometric methods of comparable sensitivity. Smoking habits (current or previous smoking, smoking at least one cigarette per day in at least one year) were obtained from questionnaires. Patients stating that they were current smokers or that they had been smokers were combined into one group (ever smoking). Patients that were not current or previous smokers were considered as non‐smoking (never smoking). Participants that did not answer the question on smoking habits in the questionnaire, 9.5%, were excluded from all smoking‐related calculations. The latest known HbA1c and blood pressure values were obtained from hospital records or from measurements at the inclusion visit. HbA1c was measured with high pressure liquid chromatography (HPLC) and reference values for healthy individuals were in Sweden <5.2% and in Finland 4.0%–6.0%. To compensate for the different normal values, the HbA1c measures were transformed to per cent of upper normal limit in both countries. In the Swedish sample set an HbA1c value of 7.0 equals 134.6% of reference and in the Finnish sample set 7.0 equals 116.7% of reference.
DNA was isolated from peripheral blood samples preserved in ethylenediamine tetra‐acetic acid (EDTA). Genotyping of the NOS4ab polymorphism was performed using primers described previously Citation23. Reactions were performed using a PTC‐200 thermocycler (MJ Research, SDS Biosciences, Falkenberg, Sweden). Final volume was 15 µl, containing 0.2 mM deoxyribonucleotide‐tri‐phosphate mix (dNTP), 1.5 mM MgCl2, 0.4 U AmpliTaq Gold DNA‐polymerase with provided buffer (Roche, Applied Biosystems), 0.2 mM of each primer, 20 ng DNA. Amplification for 40 cycles consisted of 30 sec at 94°C, 45 sec at 56°C and 30 sec at 72°C with initial denaturation at 94°C for 10 min. Polymerase chain reaction (PCR) products were separated on 3.0% agarose gel and stained with ethidium bromide. The Glu298Asp polymorphism was genotyped using the ABI PRISM 7000 Sequence Detection System (Applied Biosystems, Foster City, CA, USA) and allele‐specific fluorescent probes. PCR reactions were performed according to the manufacturer's recommendations.
Statistical methods
Genotype distributions were tested for Hardy‐Weinberg equilibrium. Linkage disequilibrium (LD) measures (D′ and r2) were determined using the Haploview 3.2 software Citation24. Power calculations were performed using a program developed by Purcell et al. (available at http://statgen.iop.kcl.ac.uk/gpc/cc2.html) Citation25. Allele and genotype distributions were compared between cases and controls using standard chi‐square analyses. The Finnish and Swedish sample sets were considered to be similar and they were combined. Most of the analyses were performed using T1D patients without DN as controls and T1D patients with DN as cases (incipient and overt DN grouped together); when overt and incipient DN cases were separated this is stated in the text. To evaluate differences in clinical characteristics between cases and controls, Pearson chi‐square or the non‐parametric Mann‐Whitney U tests were used as appropriate. A two‐sided P‐value <0.05 was considered significant.
As an estimate of the crude risk of developing DN associated with smoking or genotype, we calculated odds ratios and 95% confidence intervals. To assess the independent contribution of the two NOS3 polymorphisms we used logistic regression analyses, which included the known risk factors: age at diabetes onset, diabetes duration, HbA1c (transformed value), male sex, blood pressure and smoking. The combined effect of genotype and smoking was assessed using logistic regression analyses with stratification according to smoking status (never and ever smoked) and genotype. Never smoking and having low‐risk genotype was considered as reference. There is no sign of interaction (departure form additivity) if the combined effect (relative excess risk due to interaction) is 0. This is calculated as follows; 1+ORA+B+−ORA+B−−ORA−B+ where 1 is the effect of the reference (background) Citation26.We did not correct for multiple testing since we only tested for two polymorphisms in one candidate gene. All statistical analyses were performed with the statistical program SPSS 13.0 for Windows (SPSS Inc., Chicago, Illinois, USA).
Results
We calculated the power of our sample size to replicate the association of the Glu298Asp and NOS4ab with DN, using recessive models of the common alleles. In this case control setting the prevalence of DN was assumed to be 60%. Our sample provided >80% power, at the P = 0.05 level, to find an OR of 1.2 for both polymorphisms. As seen in our sample, frequencies of 0.70 for the Glu allele and 0.83 for the NOS4b‐allele were used.
Clinical characteristics and smoking
shows the clinical characteristics of the participating patients. The controls had lower HbA1c and lower blood pressure than the cases. Controls were older at diabetes onset and had longer diabetes duration than DN cases. The risk of having DN was significantly increased by ever smoking (OR = 2.00; 95% CI = 1.60–2.50).
Table I. Clinical characteristics of the participants.
The Finnish and Swedish sample sets were similar regarding the clinical characteristics, but the Swedish participants were older than the Finnish participants (mean 46.6 years, compared to 39.6, P<0.001), had longer duration of diabetes (median 31.5 years, compared to 27.4, P<0.001) and had lower blood pressure (mean 134/76 mmHg, compared to 140/82, P<0.001). However, since there were no significant differences in genotype distributions () the two sample sets were combined in the main analyses.
Table II. Genotype and allele frequencies of NOS3 polymorphisms in Finnish and Swedish cases and controls.
NOS3 polymorphisms, associations with DN and combined effects with smoking
Both polymorphisms were in Hardy‐Weinberg equilibrium in the Swedish and Finnish sample sets together, as well as separately, so the two sample sets were combined into one. The allele and genotype frequencies are summarized in . Linkage disequilibrium (LD) between the two markers were calculated to D′ = 0.92 and r2 = 0.07 in the control group. There were no associations between the polymorphisms and HbA1c or blood pressure.
We chose to analyse a recessive model of the common alleles (Glu298Asp, Glu/Glu versus Glu/Asp+Asp/Asp and NOS4ab, bb versus ab+aa). In an unadjusted calculation none of the polymorphisms were associated with a statistically increased risk of having DN (OR = 1.14; 95% CI = 0.93–1.41, P = 0.21 and OR = 1.04; 95% CI = 0.83–1.31, P = 0.75, respectively). When using logistic regression analyses to control for the known risk factors (age at diabetes onset, duration of diabetes, HbA1c, male sex, blood pressure and smoking), we found that the Glu298Asp Glu/Glu genotype showed significant association with DN, OR = 1.46; 95% CI 1.12–1.91, P = 0.005. The other included variables were also significant predictors (). Since a fraction of the cases (49 of 955) had longer duration of diabetes than 30 years before onset of DN they might belong to a group of patients that are not enriched for risk genes, but when removing these patients there was no change in associated risk. When analysing incipient and overt cases separately, Glu/Glu genotype was significantly associated with incipient DN (OR = 1.59; 95% CI = 1.15–2.20), as well as overt DN (OR = 1.44; 95% CI 1.06–1.96).
Table III. Logistic regression with diabetic nephropathy (DN) as outcome versus no DN controls, including Glu298Asp (Glu/Glu genotype versus Glu/Asp+Asp/Asp) and other contributing factors.
The NOS4ab polymorphism was not independently associated with DN.
In the analysis of additive interaction the effects of the polymorphisms were not enhanced by smoking. In the Swedish sample set there was, however, about two‐fold increased risk of DN associated with having the NOS4a allele and ever smoking, compared to having only one of the factors. The combined effect of the NOS4a allele and smoking was 1.7, which indicates departure from additivity in the Swedish sample set. This was not found in the Finnish sample set so no definite conclusions can be drawn from this finding.
Discussion
The present study, consisting of 1510 T1D patients, shows that the Glu/Glu genotype of the Glu298Asp polymorphism in NOS3 may increase the risk of developing diabetic nephropathy when taking other known risk factors into account. Due to the inclusion criteria of ⩾20 years of diabetes duration for the controls, and that age at diabetes onset and having diabetes at puberty have been suggested as contributing factors in the development of diabetic renal complications Citation27,28, age at diabetes onset and diabetes duration were controlled for in the logistic regression analyses, as well as blood pressure, HbA1c, smoking and male sex. Despite the long diabetes duration in the control group, it is possible that some may develop DN later in life. This will, however, only lead to a more conservative estimate of the results.
The Glu298Asp polymorphism has previously been tested for association with DN in American Caucasian T1D patients Citation14, but none was found. Our study shows that the effect of the Glu/Glu genotype is noticeable only when we control for other known risk factors, which demonstrates that confounding effects of environmental factors may influence the results of genetic studies. In Japanese populations, the Asp allele has been associated with hypertension, myocardial infarction and ESRD Citation17–19. The contradictory findings in Japanese and Caucasian populations may be accounted for by different genetic backgrounds of these populations. Since the genotype distributions of the polymorphisms in this study were not different between the Finnish and Swedish sample sets, the samples were combined to increase power in the analyses.
We were not able to replicate the previous association findings of the NOS4ab polymorphism with DN Citation14, although our study was well powered to detect an association Citation25, even assuming a somewhat smaller OR for the risk genotype than has been reported. Previous studies on this polymorphism have however shown contradictory results. Zanchi et al. found association between the NOS4a‐allele and DN in an American population Citation14, but it was not confirmed in other Caucasian T1D populations Citation16,Citation29. One possible explanation for the inability to replicate the previously reported association is that this association could have been a statistical fluctuation due to the small number of patients tested. Alternatively, inconsistency of results can be explained by true variability between populations due to either different disease alleles specific to some ethnic groups or different effect sizes because of modifying genetic or environmental factors Citation30.
An increase in GFR in early diabetes has been associated with development of nephropathy Citation6,Citation31,32, and NO may be involved in the regulation of GFR. There is, however, evidence that high glucose concentration inhibits eNOS and thereby NO formation Citation33, but if the expression of eNOS is higher in the afferent arterioles than in the efferent Citation10, there will be afferent vasodilation not compensated for by equal efferent vasodilation, which can cause an increase in GFR and contribute to DN development. The NOS3 genotype may modify the production and function of eNOS and thereby NOS3 contributes to the glomerular damage leading to DN.
We could not verify, in the large Finnish sample set, the interaction between smoking and the NOS4ab polymorphism previously found in a Swedish population Citation34. The effects of smoking and NOS3, on DN risk, thus seem to act separately although population differences may exist. Since DN is a complex disease depending on both genetic and environmental factors, we cannot expect one single gene to show very strong association to DN, but this study indicates that NOS3 might be involved in determining renal outcome of T1D patients. We found that, in this group of Swedish and Finnish Caucasian T1D patients, the Glu298Asp Glu/Glu genotype independently influences DN risk. This could indicate a true association with DN, although the small increase in risk must be interpreted with caution in a clinical setting.
Acknowledgements
This project was supported by grants from the Swedish Medical Research Council (project number 07531), Swedish Diabetes Association, Swedish Diabetes Foundation, Samaritan Foundation, Swedish Society of Medicine, Swedish Society of Nephrology, Children's Diabetes Foundation, Sven Jerring Foundation, Oskar Foundation, Ronald McDonalds Child Foundation and Family E. Persson Foundation, Folkhälsan Research Foundation, Wilhelm and Else Stockmann Foundation, the Research Funds of the Helsinki University, Sigrid Juselius Foundation, the Academy of Finland (00213 to MW) and Liv och Hälsa.
We wish to thank Dr E‐L Stattin, Nurse E Sjöström‐Fahlén, Professor JW Eriksson, Umeå University hospital, Biomedical Laboratory Technologists E‐L Forsberg, A Skogholm and A Reinholdsson, Karolinska Hospital, for their kind help with collecting the samples.
We warmly thank all the patients participating in the FinnDiane Study and gratefully acknowledge the assistance of the members of the Finnish Diabetic Nephropathy Study Group [35].
References
- Borch‐Johnsen K., Norgaard K., Hommel E., Mathiesen E. R., Jensen J. S., Deckert T., et al. Is diabetic nephropathy an inherited complication?. Kidney Int 1992; 41: 719–22
- Doria A., Warram J. H., Krolewski A. S. Genetic susceptibility to nephropathy in insulin‐dependent diabetes: from epidemiology to molecular genetics. Diabetes Metab Rev 1995; 11: 287–314
- Quinn M., Angelico M. C., Warram J. H., Krolewski A. S. Familial factors determine the development of diabetic nephropathy in patients with IDDM. Diabetologia 1996; 39: 940–5
- Scott L. J., Warram J. H., Hanna L. S., Laffel L. M., Ryan L., Krolewski A. S. A nonlinear effect of hyperglycemia and current cigarette smoking are major determinants of the onset of microalbuminuria in type 1 diabetes. Diabetes 2001; 50: 2842–9
- Christiansen J. S. Cigarette smoking and prevalence of microangiopathy in juvenile‐onset insulin‐dependent diabetes mellitus. Diabetes Care 1978; 1: 146–9
- Mogensen C. E., Christensen C. K. Predicting diabetic nephropathy in insulin‐dependent patients. N Engl J Med 1984; 311: 89–93
- Rudberg S., Osterby R., Dahlquist G., Nyberg G., Persson B. Predictors of renal morphological changes in the early stage of microalbuminuria in adolescents with IDDM. Diabetes Care 1997; 20: 265–71
- Zatz R., de Nucci G. Effects of acute nitric oxide inhibition on rat glomerular microcirculation. Am J Physiol 1991; 261((Pt 2))F360–3
- Krolewski A. S., Poznik G. D., Placha G., Canani L., Dunn J., Walker W., et al. A genome‐wide linkage scan for genes controlling variation in urinary albumin excretion in type II diabetes. Kidney Int 2006; 69: 129–36
- Sugimoto H., Shikata K., Matsuda M., Kushiro M., Hayashi Y., Hiragushi K., et al. Increased expression of endothelial cell nitric oxide synthase (ecNOS) in afferent and glomerular endothelial cells is involved in glomerular hyperfiltration of diabetic nephropathy. Diabetologia 1998; 41: 1426–34
- Wang X. L., Mahaney M. C., Sim A. S., Wang J., Blangero J., Almasy L., et al. Genetic contribution of the endothelial constitutive nitric oxide synthase gene to plasma nitric oxide levels. Arterioscler Thromb Vasc Biol 1997; 17: 3147–53
- Tsukada T., Yokoyama K., Arai T., Takemoto F., Hara S., Yamada A., et al. Evidence of association of the ecNOS gene polymorphism with plasma NO metabolite levels in humans. Biochem Biophys Res Commun 1998; 245: 190–3
- Neugebauer S., Baba T., Watanabe T. Association of the nitric oxide synthase gene polymorphism with an increased risk for progression to diabetic nephropathy in type 2 diabetes. Diabetes 2000; 49: 500–3
- Zanchi A., Moczulski D. K., Hanna L. S., Wantman M., Warram J. H., Krolewski A. S. Risk of advanced diabetic nephropathy in type 1 diabetes is associated with endothelial nitric oxide synthase gene polymorphism. Kidney Int 2000; 57: 405–13
- Shimizu T., Onuma T., Kawamori R., Makita Y., Tomino Y. Endothelial nitric oxide synthase gene and the development of diabetic nephropathy. Diabetes Res Clin Pract 2002; 58: 179–85
- Rippin J. D., Patel A., Belyaev N. D., Gill G. V., Barnett A. H., Bain S. C. Nitric oxide synthase gene polymorphisms and diabetic nephropathy. Diabetologia 2003; 46: 426–8
- Miyamoto Y., Saito Y., Kajiyama N., Yoshimura M., Shimasaki Y., Nakayama M., et al. Endothelial nitric oxide synthase gene is positively associated with essential hypertension. Hypertension 1998; 32: 3–8
- Shimasaki Y., Yasue H., Yoshimura M., Nakayama M., Kugiyama K., Ogawa H., et al. Association of the missense Glu298Asp variant of the endothelial nitric oxide synthase gene with myocardial infarction. J Am Coll Cardiol 1998; 31: 1506–10
- Noiri E., Satoh H., Taguchi J., Brodsky S. V., Nakao A., Ogawa Y., et al. Association of eNOS Glu298Asp polymorphism with end‐stage renal disease. Hypertension 2002; 40: 535–40
- Kugiyama K., Yasue H., Ohgushi M., Motoyama T., Kawano H., Inobe Y., et al. Deficiency in nitric oxide bioactivity in epicardial coronary arteries of cigarette smokers. J Am Coll Cardiol 1996; 28: 1161–7
- McVeigh G. E., Lemay L., Morgan D., Cohn J. N. Effects of long‐term cigarette smoking on endothelium‐dependent responses in humans. Am J Cardiol 1996; 78: 668–72
- Barua R. S., Ambrose J. A., Eales‐Reynolds L. J., DeVoe M. C., Zervas J. G., Saha D. C. Heavy and light cigarette smokers have similar dysfunction of endothelial vasoregulatory activity: an in vivo and in vitro correlation. J Am Coll Cardiol 2002; 39: 1758–63
- Wang X. L., Sim A. S., Badenhop R. F., McCredie R. M., Wilcken D. E. A smoking‐dependent risk of coronary artery disease associated with a polymorphism of the endothelial nitric oxide synthase gene. Nat Med 1996; 2: 41–5
- Barrett J. C., Fry B., Maller J., Daly M. J. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005; 21: 263–5
- Purcell S., Cherny S. S., Sham P. C. Genetic Power Calculator: design of linkage and association genetic mapping studies of complex traits. Bioinformatics 2003; 19: 149–50
- Rothman K. J. Modern Epidemiology. Little, Brown and Company, Boston 1986, 1st ed.
- Schultz C. J., Konopelska‐Bahu T., Dalton R. N., Carroll T. A., Stratton I., Gale E. A., et al. Microalbuminuria prevalence varies with age, sex, and puberty in children with type 1 diabetes followed from diagnosis in a longitudinal study. Oxford Regional Prospective Study Group. Diabetes Care 1999; 22: 495–502
- Svensson M., Eriksson J. W., Dahlquist G. Early glycemic control, age at onset, and development of microvascular complications in childhood‐onset type 1 diabetes: a population‐based study in northern Sweden. Diabetes Care 2004; 27: 955–62
- Degen B., Schmidt S., Ritz E. A polymorphism in the gene for the endothelial nitric oxide synthase and diabetic nephropathy. Nephrol Dial Transplant 2001; 16: 185
- Colhoun H. M., McKeigue P. M., Davey Smith G. Problems of reporting genetic associations with complex outcomes. Lancet 2003; 361: 865–72
- Rudberg S., Persson B., Dahlquist G. Increased glomerular filtration rate as a predictor of diabetic nephropathy—an 8‐year prospective study. Kidney Int 1992; 41: 822–8
- Brenner B. M. Hemodynamically mediated glomerular injury and the progressive nature of kidney disease. Kidney Int 1983; 23: 647–55
- Chu S., Bohlen H. G. High concentration of glucose inhibits glomerular endothelial eNOS through a PKC mechanism. Am J Physiol Renal Physiol 2004; 287: F384–92
- Mollsten A., Svensson M., Kockum I., Rudberg S., Brismar K., Dahlquist G. Increased risk for diabetic nephropathy in smoking patients carrying a polymorphism in the endothelial cell nitric oxide synthase gene. Diabetologia Supplement 2004; 47(Suppl 1): A48
- Pettersson‐Fernholm K., Karvonen M. K., Kallio J., Forsblom C. M., Koulu M., Pesonen U., et al. Leucine 7 to proline 7 polymorphism in the preproneuropeptide Y is associated with proteinuria, coronary heart disease, and glycemic control in type 1 diabetic patients. Diabetes Care 2004; 27: 503–9