The prevalence of ARF in the elderly remains highly variable depending on the region of the world. These differences are probably related to the demographic structure of the countries. In fact, in developing countries with large populations, the population is younger. This is the case in Senegal, where the average age is 19 [17] , which contrasts with a much higher average age in developed countries such as France, where 16.2% of the population is over 65 [18] . Multiple studies have demonstrated that incidence of ARF increases with age. In an Italian hospital cohort, the elderly (≥65 years) had 10 times the incidence rate of ARF compared to those <65 years of age [7] . The same trend is found in a Spanish hospital cohort where the incidence of ARF was 3.5 times higher in patients older than 70 years than in younger counterparts [19] . the increased incidence of ARF in the elderly is multifactorial, attributable in part to anatomic and physiologic changes in the aging kidney, to an increased burden of comorbidities impacting kidney function, to more frequent exposure to medications and interventions that alter renal hemodynamics or are nephrotoxic, and to alterations in drug metabolism and clearance associated with aging (6, 17, 43). In our study, 45 patients aged 60 or over presented with ARF, i.e. a prevalence of 3.34%. Diallo et al. [14] in Saint Louis in Senegal had found a higher prevalence than ours (29.5%). However, the studies by Ishani et al. [9] in the United States and that of Turgutalp et al. [20] in Türkiye reported that prevalence was close to our series (3.1% and 5.8% respectively). The variability of prevalence in the literature is strongly influenced by the variability of the threshold for defining the “elderly” subject. In our study, we chose the threshold of 60 years, corresponding to the WHO [4] references. Table 4 presents the prevalence in different studies.

Age is an important factor in the occurrence of ARF. This is because the kidneys undergo a number of important age-dependent changes. Among these changes, we have the renal mass which decreases with aging, reaching approximately 75% - 80% of young adulthood weight by the age of 80 to 90 years [21] [22] [23] . The kidneys have lost between 30% to 50% of their cortical glomeruli by the age of 70 years due to ischemic changes and a significant number of the remaining glomeruli manifest some degree of sclerosis [24] [25] . In addition, there has been a reduction in the number and size of tubules, increasing tubulointerstitial

fibrosis, a decrease in glomerular filtering surface area, thickening of glomerular and tubular basement membranes, arteriosclerosis and decreased afferent arteriolar luminal area [26] [27] . These structural changes contribute to functional alterations among which we have a reduction in renal blood flow (RBF) of up to 50% from age 20 to age 80 [3] and a decline in glomerular filtration rate (GFR) [3] . The decrease in RBF and GFR, taken together, represents a loss of renal functional reserve in elderly and contributes to an increased risk for development of ARF [26] . In our study, the average age was 70.31 ± 6.5 years. This average corresponds to the age of vulnerability to ARF as described in the literature. This result could be superimposed on the values found by Diallo et al. [14] in Senegal and Koffi [28] in Côte d’Ivoire, who found an average age of 70.94 years and 71.8 ± 7 years respectively. Our result was also close to the mean reported by Selmi et al. [29] , Ishani et al. [9] and Gorsane et al. [30] who respectively found average ages of 74.2 years, 78.34 years and 79.2 years in relation to the patient selection threshold in these series that were between 65 and 67 years old.

The implication of gender in the susceptibility of developing AKI remains a subject of debate. Female sex is listed among the shared susceptibility factors that confer a higher risk of ARF according to the Kidney Disease Improving Global Outcomes (KDIGO) clinical practice guidelines for acute kidney injury [5] . At present, there are no randomized controlled trials examining sex as a susceptibility variable for ARF. Most clinical cohort studies are confounded by the inclusion of women across all age groups including pre- and post-menopausal women in an uncontrolled way [31] . A meta-analysis by Neugarten et al. demonstrated that female sex is associated with protection against ARF and undermined the previously established belief that female sex is a significant risk factor for ARF. On the contrary, male sex seems to enhance the susceptibility to ARF [32] . The choice of classification would also have an impact on the results of studies on the susceptibility of developing ARF depending on gender. Srisawat et al. showed that the incidence of ARF was greater in men than women when KDIGO criteria were used to define ARF, but that sex-related differences in ARF disappeared when RIFLE criteria were used [33] . We did not look for statistical correlation between gender and ARF in our study. We just observed that the male gender was predominant. This same trend was found in the vast majority of studies. This male predominance would be related to an acceleration of renal aging in men secondary to the stimulation of androgen production [18] . Diallo et al. [14] in Senegal and Koffi [28] in Côte d’Ivoire reported a sex ratio of 3.5 and 4.55 respectively, which was close to that of our series.

The older patients present several comorbidities increasing the susceptibility to kidney injury. These comorbidities include hypertension, diabetes mellitus, heart disease, pre-existing chronic kidney disease (CKD), and cancer.

Hypertension: In our study, hypertension was one of the most common comorbidities (68.9%). This high frequency of elderly suffering from ARF is explained by the fact that hypertension affects more than 50% of the general elderly population. Continued high blood pressure will lead to increased smooth muscle and hyaline content of the media. This will result in thickening of arterial wall and stenosis of the lumen, which in turn induces reduction of the RBF. These alterations make the kidney more prone to develop pre-renal ARF in case of hypo-perfusion [34] .

Diabetes mellitus: In patients with diabetes mellitus, the high-glucose concentration damages blood vessels (microvascular damage) also in the glomerular structure, leading to several toxic effects and consequent micro-infarcts with reduced renal functional reserve. The persistence of hyperglycemia induces the formation of advanced glycation end-products (AGE), which lead to increased production of extracellular matrix, occlusion of the glomerular capillaries and cellular damage [35] [36] . Hyperglycemia also promotes serious infections. All these factors make the elderly diabetic patient predisposed to ARF. In a prospective case-control study of older adults, Mittalhenkle et al. evaluated the ARF incidence in a cohort of 5731 older individuals and found a significant association of diabetes mellitus, hypertension, and current smoking with incident acute renal failure (ARF) [37] . In our study, diabetes was the 3^rd most common comorbidity (31.1%).

Commonly, the coexistence of several comorbidities in elderly patients needs multiple drugs (polypharmacy), which may be nephrotoxic alone or in combination [38] . The drugs may induce renal damage with several mechanisms, often combined, such as acute hypersensitivity, chronic accumulation and intoxication given the reduced excretion of elderly kidney. In West Africa, traditional alternative medicine is the first resort for nearly 80% of patients [39] , polypharmacy therefore also includes the consumption of natural plants. In Dakar, Senegal, 46.7% of households use analgesic plants because they seem to be effective and less expensive [40] . The legitimate questions that must be asked are firstly, what are the harmful effects of these traditional medicines on the kidneys? And secondly, what are the interactions between these traditional medicines and the pharmaceutical products consumed for the management of comorbidities? Studies on the adverse effects of herbal medicine show that most of the harmful effects of medicinal plants are related not to the plants themselves, but to an error of identification, to involuntary contamination (by another plant, by heavy metals, by pathogenic microorganisms or by agrochemical residues), non-compliance with the adequate dose or an interaction with medications [41] . The error in identifying plants can be illustrated by the example of Belgium where more than 50 people suffered from kidney failure in 1996 after ingesting an herbal preparation containing Aristolochiafangchi (guang fang ji) a toxic plant, instead of Stephaniatetrandra (fang chi hang) following the confusion between these two species with very similar Chinese names [41] . Phytotherapy was found in the majority (68.9%) of patients in our study. However, we did not look for a correlation between the use of these traditional treatments and the occurrence of ARF.

In our study, functional acute renal failure predominated at 46.6%, followed by obstructive acute renal failure at 35.6%. These results are similar to those of Diallo et al. [14] and Selmi et al. [29] in Tunisia. On the other hand, Koffi [28] in Côte d’Ivoire had found a predominance of organic renal failure.

Aging decreases the body’s reserve capacities, which has the effect of making it more difficult for the body to adapt and making it more vulnerable in the event of aggression. In addition, thirst is particularly altered by aging. This could explain the predominance of functional acute renal failure in these studies.

The most common cause of ARF in our study was of tumoral origin (35.5%). It was followed by infection (20%). The series of Koffi [28] in Côte d’Ivoire and Gorsane et al. [30] in Tunisia found a predominance of infection among the causes of ARI. This difference could be justified by the fact that the urology department was one of the most represented referral departments.

It is currently the most widely used classification. In our study, Stage 3 was predominant with 66.67%, followed by Stage 2 with 17.78%. These results were similar to those of Koffi in Côte d’Ivoire [28] , and that of Selmi et al. [29] in Tunisia, where Stage 3 was also the majority. On the other hand, Li et al. in China [42] found a predominance of Stage 1, with 47.2% followed by Stage 3 with 27.6%. This difference could be explained by the late transfer to the nephrology department in our context.

The complications identified were mainly hyperkalemia, hyponatremia each found in 15 patients (33.3%) and poorly tolerated uremia in 7 patients (15.21%).

The evolution was favorable in 80% of the patients among whom 44.44% or 20 patients kept an alteration of the renal function, and unfavorable in 20% of patients with 4.44% of deaths.

In the literature, Diallo et al. [14] in Saint Louis in Senegal found a favorable evolution in 61.1% with recovery of renal function and unfavorable in 38.9% including 27.8% death. Koffi [28] in Côte d’Ivoire found similar results with an evolution that was favorable in 50.5% of cases; death in 46.7% of cases and among the patients who evolved favorably, impaired renal function persisted in 12 cases (11.2%).