Mecanismos sugeridos entre obesidad y el CM

76

Tables 12 summarizes the comparison of variables between subjects with and without clinical nephropathy. The BMI, mean arterial pressure, diastolic BP, serum creatinine, Blood Urea Nitrogen, glycated haemoglobin and duration of DM were significantly higher in subjects with clinical nephropathy compared to subjects without clinical nephropathy. The eGFR of subjects with clinical nephropathy was lower (68.46 ± 24.55ml/min/1.73m²) compared to subjects without clinical nephropathy (103.53 ± 24.93ml/min/1.73m²) p value = 0.001.

Table 12: Comparison between subjects with and without clinical nephropathy.

Variables Clinical

nephropathy Present

Clinical nephropathy Absent

P value

Mean ± SD Mean ± SD

Age in years 61.0 ± 12.8 58.6 ± 10.7 0.172

BMI (kg/m²) 28.4 ± 5.5 24.8 ± 5.9 0.001**

Waist -Hip- Ratio 0.95 ± 0.09 0.96 ± 0.07 0.552

Systolic BP (mmHg) 131.0 ± 13.5 128.6 ± 14.0 0.275

Mean Arterial Pressure (mmHg)

Current fasting plasma glucose (mmol/l) eGFR (MDRD) ml/min/1.73m²

98.5 ± 10.3 9.5 ± 4.4 68.46 ± 24.55

94.2 ± 9.47 9.3 ± 4.0 103.53 ± 24.93

0.005**

0.765 0.001**

Mean Rank Mean Rank

*Diastolic BP 169.9 129.8 0.001**

*Serum Creatinine 203.32 122.13 0.001**

*Blood Urea Nitrogen 192.40 124.98 0.001**

*Glycated Hb 166.55 130.37 0.003**

*Duration of DM 223.03 117.71 0.001**

BMI = Body mass index, BP = blood pressure, Hb = haemoglobin, eGFR = estimated glomerular filtration rate, Duration of DM = Time period between diagnosis of DM and recruitment for this study.

Footnote: *The mean rank of non-parametric variables was compared using Mann Whitney U

test. Significant p values were highlighted in red. **Variables with significant p values.

CHAPTER FIVE DISCUSSION

Clinical nephropathy is an advanced stage of DN. It implies that the disease had progressed from the stage of intraglomerular hypertension/glomerular hyperfilteration

77

through the stage of incipient nephropathy. Unlike microalbuminuria, patients with clinical diabetic nephropathy invariably develop hypertension, a progressive worsening in proteinuria, and a relentless decline in glomerular filtration rate over time without intervention.¹⁵ End stage renal disease in diabetic patients is associated with poorer outcome compared to non-diabetic patients because of the comorbid conditions in the diabetic population, and majority of these patients have type 2 DM.¹⁵ Moreover, DN is the diabetic specific complication with highest mortality, and the risk of cardiovascular death is higher by a factor of 12 in these patients when compared to renal diseases resulting from primary glomerular diseases.^5,15,21 Despite this dreaded complication, the prevalence of type 2 DM is increasing worldwide, particularly in the developing countries like Nigeria.^3,4 This hospital-based cross-sectional study was designed to investigate the frequency of occurrence, characteristics and correlates of clinical nephropathy in type 2 DM patients. This is against the backdrop of the facts that DN is the leading cause of ESRD in western world, the third leading cause of ESRD in Nigeria and in Ilorin, Nigeria.^5-7,106

Earlier study in this centre on type 2 DM patients did not investigate clinical nephropathy, but it focused on microalbuminuria.²⁰ Study on clinical nephropathy is important because clinical DN is potentially preventable and treatable through routine screening for early detection of microalbuminuria, optimal blood pressure and glycaemic control, and the use of certain agents like angiotensin ACEIs and/or ARBs that ameliorate the structural and haemodynamic changes implicated in the pathogenesis of DN. Therefore, the knowledge of the burden, characteristics and correlates of clinical DN in our local setting could reinforce the institution of effective preventive and management plan. This study will also contribute to existing literature

78

on this subject being the first study in this region of the country on clinical diabetic nephropathy.

5.1 THE PREVALENCE OF CLINICAL NEPHROPATHY IN TYPE 2 DM PATIENTS IN ILORIN.

The prevalence of clinical nephropathy as seen in patients with type 2 DM in this study was 18.3%. Although there have been several studies on diabetic nephropathy, most of these studies either focused on microalbuminuria or investigated DN as a whole.16-20,48,50,106 Relatively few studies specifically examined clinical nephropathy in type 2 DM patients.61,107,109,110

It is difficult to make a diagnosis of clinical DN from a snapshot cross-sectional study without establishing persistence of macroalbuminuria. Although, macroalbuminuria is the hallmark of clinical DN, the two terms are not synonymous.

Persistence of macroalbuminuria in a diabetic patient over a 3 to 6 months period in the absence of other renal disease should be established before a diagnosis of clinical DN can be made.^12-15 There are other factors like infection, fever, muscular exertion, heart failure and elevated blood pressure among others that could transiently increase urine albumin excretion. Establishing persistence of macroalbuminuria eliminates these transient confounders and reduces the proportion of subjects with macroalbuminuria that actually have clinical DN. Notwithstanding, the prevalence of macroalbuminuria in a diabetic population should mirror that of clinical nephropathy in that population. This underlies the rationale for comparison of prevalence of clinical DN in this study with that of macroalbuminuria in other studies. This study introduced a modification to the cross-sectional design by repeating the estimation of

79

urine ACR after at least 3 months in all patients with macroalbuminuria and also eliminated known confounders. This approach improved the sensitivity of the diagnosis of clinical DN as there were no ethical indications for renal biopsy in these patients.

The prevalence of macroalbuminuria found in a prospective cross-sectional study in Sudanese teaching hospital was 18.4%, this was the closest prevalence in literatures to that of clinical nephropathy found in this study.³⁵ The observed prevalence of clinical DN in this study is lower to the previous report by some authors in other centres in this country.^37,38,107 Alebiosu reported that the overall frequency of occurrence of clinical DN in diabetic patients (both type 1 & 2) in Olabisi Onabanjo Teaching Hospital, Shagamu, Nigeria, was 28.4%.¹⁰⁷ In Ibadan, Nigeria, Alebiosu et al in a clinical review of diabetic nephropathy commented that the prevalence of clinical DN increased from 19% in 1971 to 42.5% in 1988.³⁹ An increment in the prevalence of DN in diabetic population was also reported by Afifi et al in Egypt.³³ With the increasing prevalence of type 2 DM, the prevalence of complications like clinical nephropathy should theoretically be on the increase irrespective of locality. This may however not be so if there had been improvement in detection and quality of management of type 2 DM in the region, thereby forestalling concomitant increase in complications like clinical DN. There is however no current study that established such improvements despite the growing knowledge base on DM. Also, there are no previous studies in Ilorin, Nigeria to determine the prevalence of clinical DN before this present study, hence, the trend in prevalence over the years could not be ascertained. Although, studies by Osuntokun et al, Bella et al, and Alebiosu et al were not restricted to type 2 DM patients, this alone could not have accounted for the

80

disparity in prevalence.^37-39,107 Over 90% of diabetic patients have type 2 disease,^1,2,15 so a study restricted to type 2 DM should still mirror the overall prevalence of clinical DN.

In this current study, most of the study subjects were on various antihypertensive drugs except ACEI and ARB. These could very much explain the relatively low prevalence of clinical nephropathy recorded in this study, since lowering of blood pressure has been established to attenuate the progression of renal disease and reduce proteinuria.59,69,111,112 Eight patients with microscopic haematuria were also excluded from this study to improve the specificity of the diagnosis of clinical nephropathy.

This may also account for the relatively lower prevalence observed in this study, because It has been shown with renal histologic findings that microscopic haematuria may actually be a common feature of clinical DN especially in patients with nephrotic range proteinuria.¹⁰⁸ Although, none of the excluded patients with haematuria had nephrotic range proteinuria, it is possible that some of them may actually have clinical nephropathy.¹⁰⁸

Other studies have defined overt nephropathy as albuminuria ≥ 300µg/mg of creatinine or dipstick positive proteinuria in the presence of diabetic retinopathy. An example of such study is the Chennai Urban Rural Epidemiology Study (CURES 45) in which the prevalence of overt nephropathy was as low as 2.2%.¹¹³ The drawback of such definition is that, in type 2 DM, absence of retinopathy does not exclude clinical nephropathy.¹⁵

81

Immunologic based assay of urine albumin like Immunoturbidometry method used in this study may not detect non immunoreactive albumin. This could result in the underestimation of urine albumin concentration and the actual prevalence of clinical nephropathy.¹⁵

The variation in the prevalence of clinical DN, macroalbuminuria, or DN in type 2 DM patients in various studies could also be attributed to differences in diagnostic criteria, methods of assessment, period of study and ethnicity. The prevalence of 18.3% reported in this current study is still alarming considering enormity of the attendant morbidity and mortality, as well as the financial implication of renal care in a developing country setting.

5.2 GENDER DIFFERENCE IN THE PREVALENCE OF CLINICAL NEPHROPATHY IN TYPE 2 DM PATIENTS.

The prevalence of clinical DN in the male and female subgroups of type 2 DM patients were 15.3% and 20.6% respectively. The observed difference was not statistically significant. This may be due to high numerical strength of female subjects in this study with male to female ratio of 1:1.3. Although, most studies on subjects with diabetes mellitus did not specifically look into gender differences in the prevalence of clinical nephropathy, previous reported male to female ratios showed male preponderance.^39,107 The finding in this study is in accord with an earlier report from this centre on microalbuminuria in type 2 DM in which the prevalence was also higher in females.²⁰ This observed female preponderance should generate concern since a case control study in Lyon, France reported increased mortality only in type 2

82

diabetic females with ESRD.¹¹⁴ The impact of gender on mortality of patients with ESRD from type 2 DM is however yet to be investigated in this region.

5.3 CLINICAL AND SOCIODEMOGRAPHIC FEATURES OF TYPE 2 DM PATIENTS WITH CLINICAL NEPHROPATHY

Type 2 DM is essentially commoner in the adults above the age of 40 years.^1,2 Although maturity onset diabetes in the young (MODY) can occur before the age of 40 years, and latent autoimmune diabetes of adult onset (LADA) can occur above this age, most patents with the onset of DM after the age of 40 years have type 2 DM.^1,2 To improve the specificity of the diagnosis of type 2 DM, none of the patients used in this study were insulin dependent. The mean age of type 2 DM subjects with and without clinical nephropathy were 61.0 ± 12.8 and 58.6 ± 10.7 years respectively.

Although, the difference in the mean age between the 2 groups was not statistically significant, it does show that clinical DN is predominantly a disease of the middle age group. This is in consonance with previous studies.^32,107 Alebiosu et al reported mean age of type 2 DM patients with clinical nephropathy of 54.5 ± 9.7 years in Ibadan, Nigeria.³⁹ The study in

Shagamu titled; Clinical diabetic nephropathy in a tropical African population, recorded a mean age of 53.4 ± 6.3 years.¹⁰⁷ The clustering of patients with clinical DN in the middle age could be a reflection of the fact that most type 2 diabetic patient fall within this age group.

It is not surprising that all the subjects with clinical nephropathy were Yoruba, since 92.7% of the study population and majority of diabetic patients receiving care at the study location were Yoruba.

83

It was observed that 58% and 64% of the subjects with clinical nephropathy were petty traders and uneducated respectively. This reflects that it is this category of patients that are least likely to be able to access healthcare facilities adequately due to their socioeconomic status. Hence, they are more prone to delayed diagnosis, poor drug compliance, poor glycaemic control, and diabetic complications. This is further reinforced by the observation that majority of these subjects had inadequate BP and glycaemic control (76% each).

5.4 CORRELATES OF CLINICAL NEPHROPATHY IN TYPE 2 DIABETIC PATIENTS.

The age of the subjects, duration of DM, systolic blood pressure, current fasting plasma glucose, glycated hemoglobin and serum creatinine concentration correlated positively, while eGFR correlated negatively with clinical nephropathy in this study.

Out of all these correlates, the duration of DM, serum creatinine concentration and eGFR were the strongest (r = 0.779, 0.75, and 0.695 respectively). After controlling for confounders, the independent predictors of clinical nephropathy were duration of DM, eGFR, and glycated haemoglobin. Body mass index (BMI), waist -to- hip ratio, diastolic blood pressure, mean arterial pressure, and blood urea nitrogen did not correlate with clinical nephropathy. The correlation pattern seen in this study is in agreement with the established pathophysiology and clinical characteristics of DN in some earlier literatures.15,109,115,116

Although, plasma cystatin C based estimates of GFR has been shown to be more accurate than creatinine based estimate in early stages of DN, in established DN, there

84

is only marginal difference in these 2 estimation modalities.¹⁰⁵ Hence, the use of MDRD formula to estimate GFR in this study is still currently reasonable and acceptable.^15,105,115 The effect of antihypertensives and exclusion of subjects with poor blood pressure control in the study population could explain absence of correlation between clinical nephropathy and diastolic/mean arterial pressure. It is however difficult to explain why BMI and waist -to- hip ratio did not correlate significantly with clinical nephropathy. This contrasts with the observation of ‘the DEMAND study, which reported that the risk of progression of urinary albumin excretion increased by 7% for every 5cm increase in waist circumference from baseline value, and by 17% for every one unit increase in BMI.¹¹⁷ The DEMAND study was more powered (1289 patients), and it was a prospective design, therefore, expected to be more sensitive at observing the association between the two variables.¹¹⁷

Previous studies on DN showed varying pattern of correlates.1,2,11,15,109 The negative correlation of clinical nephropathy with eGFR is in agreement with earlier report by Alebiosu et al in Ibadan, Nigeria.³⁹ Dobronravov also reported high proteinuria in type 2 DM patients with chronic renal failure (i.e. low eGFR).¹¹⁰ Finding in this study is also in support of the observation in Dar es Salaam that duration of diabetes, systolic blood pressure and serum creatinine not only correlated, but actually predicted proteinuria in type 2 diabetic patients.¹¹ In related previous literature, poor glycaemic control correlated with the formation of AGEs, which in turn correlated with DM complication like nephropathy.^1,15,118 Some previous local studies reported no correlation between glycaemic control and microalbuminuria, ^20,106 this present study found that glycaemic control correlated with macroalbuminuria. This later observation may be a reflection that microalbuminuria does not necessarily indicate

85

nephropathy in type 2 DM.^1,2 Ballard et al in a population based study in Rochester, Minnesota, found that elevated initial fasting blood glucose and older age at onset of diabetes were associated with the development of persistent proteinuria in type 2 DM.¹⁰⁹ The finding of Ballard et al by projection is in consonance with the finding in this study that age, fasting plasma glucose and glycated haemoglobin correlated positively with clinical nephropathy. However, contrary to what was found in this present study, Ballard et al found no association between duration of type 2 DM and incidence of persistent proteinuria after controlling for attained age.¹⁰⁹ Although, the duration of disease has a strong correlation with the evolution of nephropathy in type l DM, the prolonged asymptomatic period that characterizes type 2 DM may distort this relationship.^1,2,15 The strong correlation between duration of DM and clinical nephropathy observed in this study is in consonance with earlier report by Nakhjavani et al and Ranjit et al.^113,119 The later 2 studies found that increased urine albumin excretion correlated with a longer duration of diabetes in type 2 diabetic population independent of other variables.^113,119 Afkhami-Ardekani et al reported that the duration of diabetes strongly correlated with microalbuminuria in type 2 diabetic patients.¹²⁰ Hence, by projection duration of diabetes would be expected to also correlate with clinical nephropathy in type 2 DM patients.

5.5 COMPARISON OF CHARACTERISTICS OF PATIENTS WITH AND WITHOUT CLINICAL NEPHROPATHY

The BMI, mean arterial pressure, diastolic blood pressure, serum creatinine, blood urea nitrogen, glycated haemoglobin and duration of DM were significantly higher in type 2 DM patients with clinical nephropathy compared to those without clinical nephropathy. This observation could be seen as a sequel to earlier observation on

86

microalbuminuria in Ilorin, Nigeria in which there was no significant difference in age, duration of DM, blood pressure, glycaemic control, and serum creatinine between type 2 DM patients with and without microalbuminuria.²⁰ The observed difference is a reflection that clinical nephropathy is a more advanced stage in the evolution of DN.⁴⁶ The study in Dar es Salaam, Tanzania, also found that the duration of type 2 DM and diastolic blood pressure were significantly higher in type 2 DM patients with abnormal urinary albumin excretion compared to those with normal urine albumin excretion.¹¹ However, in contrast to the present study, the study in Dar es Salaam did not find any significant difference in BMI and glycaemic control between the 2 groups.¹¹ The cut-off point of urinary albumin excretion used for comparison in these 2 studies were different, which could explain the disparity in the observations. Idogu et al found that overweight and obesity were more common in type 2 diabetic with nephropathy when compared with those without nephropathy.⁴⁹ Their observation is in consonance with the difference in BMI observed between type 2 DM patients with and without overt albuminuria in this present study. This observation also confirms the hypothetic association between BMI and urinary albumin excretion as postulated by Rossi et al in the DEMAND study.¹¹⁷ The effect of high BMI on nephropathy is due to insulin resistance, which has been shown to cause renal damage by independently inducing systemic hypertension, glomerular hypertension, hyperfiltration, glomerulosclerosis, endothelial dysfunction, increased mesengial extracellular matrix deposition and release of oxidative stress product.¹²¹

87

CHAPTER SIX

CONCLUSIONS, RECOMMENDATIONS AND LIMITATIONS

CONCLUSIONS

The prevalence of clinical nephropathy in type 2 DM in University of Ilorin Teaching Hospital (UITH) in this study was 18.3%. The gender specific prevalence of clinical nephropathy were 15.3% and 20.6% in males and females respectively, however, these observed difference was not statistically significant. Patients with clinical nephropathy were mainly uneducated petty traders, overweight or obese, and had inadequate BP and glycaemic control

The age of the subjects, duration of DM, systolic blood pressure, current fasting plasma glucose, glycated haemoglobin and serum creatinine correlated positively, while the eGFR correlated negatively with clinical nephropathy in type 2 DM subjects.

88

Type 2 DM subjects with clinical nephropathy were more likely to have higher BMI, mean arterial pressure, diastolic blood pressure, serum creatinine, blood urea nitrogen, glycated haemoglobin, longer duration of DM and lower eGFR compared to those without clinical nephropathy.

RECOMMENDATIONS

1. Assessment of urinary albumin excretion should be incorporated in the evaluation of type 2 DM patients at diagnosis and at regular interval during follow up for early detection of abnormal urinary albumin excretion.

2. Type 2 DM patients with abnormal urinary albumin excretion should be treated appropriately even if glycaemic control is optimal to forestall rapid progression to end stage renal disease.

3. Blood pressure should be adequately controlled in all type 2 DM patients.

4. Serum creatinine concentration and eGFR should be monitored in all type 2 DM patients with abnormal urinary albumin excretion.

5. Renal care should be subsidized in Nigeria to ensure that people with clinical DN receive adequate care.

89

LIMITATIONS

1. Immunoturbidometry method does not detect nonimmunoreactive albumin. This could have resulted in underestimation of the actual prevalence of clinical diabetic nephropathy in this study.

2. The use of antihypertensives other than ACEIs and ARBs can also attenuate the progression of renal disease and reduce proteinuria (especially calcium channel blockers). These could have resulted in underestimation of the actual prevalence of clinical nephropathy in this study.

3. Diluting urine sample with isotonic saline for quantitative assessment of urine albumin concentration above 150mg/l with HemoCue Albumin 201 analyzer might affect the precision of the result.

4. Symptoms and urinalysis parameters were used to exclude urinary tract infection instead of the more sensitive urine microscopy and culture because of logistics and financial reasons.