Association Between Temperature and Inpatient Stone Admission in a Pediatric Population

Introduction

The prevalence of nephrolithiasis has been increasing worldwide for the past 30 years. ¹ Recently, the incidence of this condition has been disproportionately higher in the pediatric population. ^2,3 Aside from the immediate costs of treatment, which is estimated at $2 billion per year in the United States, patients with nephrolithiasis often experience a greater lifetime burden of disease because of a higher rate of recurrent events. ⁴ With the needle pushing ever younger for a condition traditionally associated with adults, there has been an increased focus to investigate factors that may contribute to pediatric nephrolithiasis.

It was once believed that stones formed during childhood are because of rare genetic causes or infection. Although monogenic causes of nephrolithiasis are certainly more likely in children than in adults, it remains a rare phenomenon overall. ³ As such, the prevailing theory has shifted more toward likening pediatric stone disease to adult stone disease. To this end, several factors have been associated with nephrolithiasis, including obesity, metabolic syndrome, fluid intake, and geography. However, these studies are based almost exclusively on the adult population, with extrapolations of the data being made to inform care for the pediatric population.

The lack of robust data in children provides an opportunity to confirm that these extrapolations remain valid among our younger patients. To bridge this gap in knowledge, we sought to examine associations between ambient temperature and inpatient hospitalizations for nephrolithiasis in the pediatric population. As a frame of reference, we also examined adult inpatient hospitalizations for nephrolithiasis during this same time span.

Methods

This population-based cross-sectional study utilized the Kids' Inpatient Database (KID) for calendar year 2016 to identify nationwide pediatric inpatient admissions related to nephrolithiasis. The design and objective of KID have been previously described. ⁵ In brief, it is part of the family of databases developed for the Healthcare Cost and Utilization Project (HCUP), sponsored by the Agency for Healthcare Research and Quality (AHRQ), designed to obtain estimates of hospital inpatient stays for pediatric patients in the United States. These data are collected every 4 years, most recently in 2016.

Temperature data were obtained from the National Oceanic and Atmospheric Administration (NOAA). Although NOAA provides daily temperature data for >4000 locations throughout the United States, the highest level of granularity provided by KID is at the level of the month and region, as defined by the United States Census Bureau (Fig. 1). Therefore, the average monthly temperature was recorded for each of the four regions and appropriately linked to each admission in KID. Any patients with incomplete data on month and/or region of admission were excluded from analysis. Furthermore, any admissions listed for a newborn birth were excluded.

OPEN IN VIEWER

FIG. 1.

United States Census Bureau Regions. Color images are available online.

Patients with a stone-related admission were identified based on hospital billing codes (Supplementary Table S1). Months were categorized into seasons as follows: winter (December–February), spring (March–May), summer (June–August), and fall (September–November). Comparative statistics were performed to detect differences in patient and admission characteristics between the stone admission group and the nonstone admission group. Logistic regression was performed to determine associations between stone-related admissions and temperature, as well as patient factors to control for potential confounding variables.

To establish a frame of reference, this analysis was replicated using the National Inpatient Sample (NIS) from 2016 to evaluate associations in the adult population. Similar to KID, NIS is also a part of HCUP and provides estimates of hospital inpatient stays for adult patients in the United States.

All statistical analyses were conducted using STATA v15 (StataCorp LLC, College Station, TX) and with significance set at α = 0.05. As this study is based on publicly available databases, institutional review board approval was not required.

Results

Discussion

The current understanding of temperature on the development of childhood nephrolithiasis is largely gleaned from the adult population. As such, there is no evidence to date that suggests whether this extrapolation holds true in the pediatric cohort. This study demonstrates that temperature is indeed significantly associated with stone-related admissions in the pediatric population, consistent with what is seen in the adult population. The magnitude of effect is nearly the same in both groups, with a 2% increased odds of stone admission for every 10°F increase in temperature.

The association between higher temperatures and increased presentation for stone-related events in adults has been noted for >60 years. ^6,7 Since then, multiple studies have demonstrated similar findings worldwide, and this link was validated in a systematic review of the literature in 2017. ⁸ Although factors such as humidity, atmospheric pressure, and sunlight-induced increases in vitamin D have been theorized as potential causes, ambient temperatures have thus far demonstrated the strongest link with stone presentation. ^9,10 Indeed, a time-series model found that high temperatures >30°C (86°F) were associated with clinical presentations for nephrolithiasis within 3 days. ¹¹

This suggests that although the true incidence of nephrolithiasis cannot be accurately captured, the use of inpatient admissions can be used as a proxy to examine trends between temperature and nephrolithiasis. In concordance with the current body of evidence, findings from this study demonstrate an association between increased temperatures and the downstream effects of developing nephrolithiasis—in this case inpatient admission—in the adult population. Using this same methodology, it is reasonable to conclude that increased temperatures are associated with a greater likelihood of pediatric nephrolithiasis events as measured by inpatient admissions.

Despite the repeatedly demonstrated link between temperature and nephrolithiasis, the pathogenesis still remains unclear. Studies of American soldiers deployed to desert conditions have shown not only an increase in episodes of renal colic but also dramatic changes in urinary composition despite increases in fluid intake. ^{12 –14} One of these changes involved a decrease in urinary pH, which is associated with formation of calcium oxalate stones in pediatric populations. ¹⁵ An analysis of British soldiers deployed to the Persian Gulf demonstrated increased rates of calciuria, another factor that has been implicated in pediatric stone formation. ^15,16

Notably, once soldiers returned from the desert conditions, the urinary composition reverted nearly to baseline. ¹⁴ Nevertheless, the mechanism by which temperature affects stone formation has yet to be comprehensively elucidated. It is important to consider that although warmer temperatures might be a precipitant for stone-related admissions, the lithogenic process responsible for stone formation likely occurred months to years before clinical presentation. Therefore, techniques such as increasing fluid intake during warmer temperatures may be a useful countermeasure to combat stone-related presentations but may not preclude development of the stone itself. However, it should be noted that of all the variables associated with stone-related admissions, temperature had a smaller effect than age, gender, and race.

In the search to identify factors associated with stone disease, a gender disparity has been noted in the literature. Among the pediatric population, girls demonstrate a higher incidence of nephrolithiasis compared with boys. ^5,17 Indeed, we show similar findings, with females comprising nearly 65% of our pediatric stone cohort. Furthermore, the risk of childhood nephrolithiasis was nearly 10% higher among females compared with males. Subsequently, proposed theories to explain this gender difference revolve around the differential effects of certain hormones, namely testosterone and estrogen. ⁵

However, even this remains a topic of debate because of the opposite effect noted in the adult population, where male presentations of nephrolithiasis predominate. Further complicating matters, the gender gap in adults has narrowed over time; a study from the Rochester Epidemiology Project demonstrated a drop in the male-to-female ratio of symptomatic stone presentations from 3.2 to 1.3 between 1970 and 2000. ¹⁸ Findings from our study estimate that adult men are associated with a 23% increased odds of stone presentation compared with adult women, further substantiating the closing gender disparity among adults.

Focusing on age in the pediatric cohort, we found that the median age of patients with nephrolithiasis was ∼17 years. Furthermore, there was a remarkable increase in the odds of admission once the patient crossed 15 years of age. Following the hormone theory previously mentioned, these findings suggest that postpubertal adolescents behave more like adults. Indeed, increasingly common phenomena such as childhood obesity and metabolic syndrome have been identified as potential causes of early-onset nephrolithiasis. ¹⁹ In contrast, younger children may be less likely to have a large stone burden or run into issues with spontaneous passage as opposed to adolescents who behave and are treated more like adults.

Finally, race is another factor that has been associated with differential rates of pediatric nephrolithiasis. White children tend to be at higher risk for stone development compared with their nonwhite counterparts. ^17,20 In particular, pediatric stone disease is most common among white children, followed by Hispanics and then blacks. Similarly, we show that white patients were significantly more likely to be admitted for a stone than all other races, with blacks demonstrating the lowest risk. Examination of metabolic abnormalities based on race have not yielded many explanations, suggesting that socioeconomic factors such as diet and environment may be just as important as race in the etiology of stone disease.

There are several limitations of this study, which should be noted. Because of privacy protection laws, the granularity of data regarding inpatient admissions for pediatric patients could only be ascertained at the level of a region, as defined by the United States Census Bureau. As there are only four regions, each region encompasses a large area of land with variable geography and climate, thus making precise temperature measurements difficult. Nevertheless, our findings in the adult population are consistent with the current literature, thus providing credence to application of this methodology in the pediatric population. In fact, clustering of patients into a region may have contributed to underestimation of the observed temperature trends; repeat analysis with more granular geographic data, and thus better temperature discrimination, may reveal an increase in the effect size.

Furthermore, the data used only capture inpatient admissions. Therefore, all outpatient clinic visits and surgeries were not accounted for in our analysis. In addition, it could not be determined whether nephrolithiasis or another concurrent diagnosis was the primary cause for admission. Because of the outpatient nature of most stone surgeries, it is likely that our results underestimate the true incidence of stone disease. In contrast, pediatric patients are perhaps more likely to be admitted after a procedure for monitoring, even if the procedure would be suitable to perform as an outpatient, thus introducing another potential confounder. Finally, other factors such as a change in dietary habits or activity during warmer months may be implicated with increased stone admissions but could not be examined in our analysis.

Conclusions

The current practice and understanding of pediatric nephrolithiasis is derived primarily from the adult population. The current analysis is the first to provide evidence linking the effect of higher temperatures to increased stone admissions in the pediatric population. However, the pathogenesis of stone formation is incredibly complex and the role of temperature on lithogenesis, beyond stone-related presentations, remains unclear. Further investigation into this association is prudent, especially with the increasing incidence of pediatric stone disease and concerns about climate change resulting in higher temperatures.