Accepted Manuscript Risk factors, prevalence trend, and clustering of hypospadias cases in Puerto Rico Luis A. Avilés, Laureane Alvelo-Maldonado, Irmari Padró-Mojica, José Seguinot, Juan Carlos Jorge PII:

S1477-5131(14)00132-6

DOI:

10.1016/j.jpurol.2014.03.014

Reference:

JPUROL 1690

To appear in:

Journal of Pediatric Urology

Received Date: 13 September 2013 Accepted Date: 6 March 2014

Please cite this article as: Avilés LA, Alvelo-Maldonado L, Padró-Mojica I, Seguinot J, Jorge JC, Risk factors, prevalence trend, and clustering of hypospadias cases in Puerto Rico, Journal of Pediatric Urology (2014), doi: 10.1016/j.jpurol.2014.03.014. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Risk factors, prevalence trend, and clustering of hypospadias cases in Puerto Rico

RI PT

Luis A. Avilés a, Laureane Alvelo-Maldonado b, Irmari Padró-Mojica c, José Seguinot c, Juan Carlos Jorge d,*

Department of Social Sciences, School of Public Health, University of Puerto Rico

b

Birth Defects Prevention and Surveillance System, Department of Health of Puerto Rico

c

Department of Environmental Health, School of Public Health, University of Puerto Rico

d

Department of Anatomy and Neurobiology, School of Medicine, University of Puerto Rico

M AN U

SC

a

* Corresponding author. Department of Anatomy and Neurobiology, School of Medicine, University of Puerto Rico, PO Box 365067, San Juan, Puerto Rico, 00936 5067. Tel.: +1 787 758 2525, ext. 1506; fax: +1 787 767 0788.

KEYWORDS

TE D

E-mail address: [email protected] (J. C. Jorge).

Hypospadias; Prevalence; Risk factors; Geographic information systems Abstract Objective: The aim was to determine the distribution pattern of hypospadias cases across a

EP

well-defined geographic space.

AC C

Materials and methods: The dataset for this study was produced by the Birth Defects Prevention and Surveillance System of the Department of Health of Puerto Rico (BDSS-PR), which linked the information of male newborns of the Puerto Rico Birth Cohort dataset (PRBC; n = 92 285) from 2007 to 2010. A population-based case–control study was conducted to determine prevalence trend and to estimate the potential effects of maternal age, paternal age, birth-related variables, and health insurance status on hypospadias. Two types of geographic information systems (GIS) methods (Anselin Local Moran’s I and Getis-Ord G) were used to determine the spatial distribution of hypospadias prevalence.

1

ACCEPTED MANUSCRIPT

Results: Birthweight (< 2,500 g), age of mother (40+ years), and private health insurance were associated with hypospadias as confirmed with univariate and multivariate analyses at 95% CI. A cluster of

(p ≤ 0.05).

RI PT

hypospadias cases was detected in the north-central region of Puerto Rico with both GIS methods

Conclusions: The clustering of hypospadias prevalence provides an opportunity to assess the underlying causes of the condition and their relationships with geographical space.

SC

Introduction

Hypospadias is traditionally defined by the anatomical position of the urethral meatus on the ventral

M AN U

surface of the penis other than close to the tip of the glans. This anatomical variation results from incomplete fusion of the urethral folds of the urethral spongiosum during intrauterine development and can range in position from the glans corona up to the perineum. Aside from variations in the localization of the urethral opening, hypospadias can also present with atypical shape of the urethral opening, of the

TE D

glans, and of penile skin, various degrees of penile curvature, unilateral or bilateral cryptorchidism, among other variations in genital appearance. Severe hypospadias cases can also present with penile– scrotal transposition and/or enlarged prostatic utricle. In severe cases, clinical protocols for the

EP

management of intersexuality are commonly deployed. Given that hypospadias is not a monolithic clinical entity, we hypothesized that hypospadias prevalence is not uniformly distributed across space.

AC C

The study of prevalence across space can provide valuable information about plausible gene/environment interactions that may underlie distinct distributions of cases. Moreover, the identification of hypospadias spatial clusters, if any, can be used to optimize the delivery of health services to manage the condition. An impressive body of work on the etiology of hypospadias suggests that genetic and environmental factors are the main contributors to the high incidence of this male congenital condition. In fact, it has been suggested that, in the majority of cases, the etiology of this condition can be

2

ACCEPTED MANUSCRIPT

explained by a “two-hit hypothesis,” whereby genetic susceptibility plus environmental exposure increases the risk for having offspring with hypospadias [1,2]. Early epidemiologic studies on incidence rates reported an increase of hypospadias in certain regions of the world including the United States

RI PT

[3,4]. However, more recent studies do not support such a view; for a recent review see Fisch et al. [5]. One of these early studies showed a higher incidence of hypospadias cases in the eastern and central regions of the United States than the western region [4]. But this study, based on a nationwide

SC

surveillance program, the Birth Defects Monitoring Program, is limited by the fact that such a database was not built with a random sample of US births by representative geographic populations [5,6].

M AN U

We used the Puerto Rican archipelago as a case study to assess prevalence trend and to estimate the potential effects of maternal age, paternal age, birth-related variables, health insurance status, and the spatial distribution pattern of hypospadias prevalence. We found that birthweight (< 2,500 g), age of the mother (40+ years), and private health insurance were associated with

detected. Materials and methods

TE D

hypospadias. In addition, clustering of confirmed hypospadias cases in the mainland of Puerto Rico was

EP

The dataset for this analysis was produced by the Birth Defects Prevention and Surveillance System of the Department of Health of Puerto Rico (BDSS-PR), which linked the information of male newborns of

AC C

the Puerto Rico Birth Cohort dataset (PRBC; n = 92,285) with a dataset of confirmed cases of hypospadias from 2007 to 2010. Initial diagnosis of hypospadias was made by a US board-certified physician in the birthing hospital and the birthing hospital reported each case to BDSS-PR routinely, as required by local and US law. A group of five BDSS-PR trained nurses in congenital defects visited the birthing hospital and reviewed the medical charts to confirm cases. Whenever necessary, BDSS-PR nurses interviewed medical staff in the birthing hospital if medical information was missing in the charts. In addition, BDSS-PR staff contacted the parents by phone to provide health information, genetic

3

ACCEPTED MANUSCRIPT

counseling, and/or referral to medical services when appropriate. Case follow-up by phone provided another opportunity to confirm diagnosis. Given the geographical area of Puerto Rico (9,104 km²), it was possible for BDSS-PR nurses to monitor birthing hospitals across the entire Island. Out of hospital births

RI PT

and home deliveries, which represented less than 1% of total births for the years 2009 and 2010 (Department of Health of Puerto Rico, Division of Statistical Analysis, personal communication), are required by law to provide birth information to the Puerto Rico Health Department before a birth

SC

certificate can be issued to the parents. There is a system in place where BDSS-PR is notified of births with congenital defects even when born outside the traditional health-care system. Therefore, case

M AN U

ascertainment following the established protocol identified 279 cases of hypospadias for the years 2007–10. Case–control study

We conducted a population-based case–control study to estimate the potential effects of maternal age,

TE D

paternal age, birth-related variables, and health insurance status on hypospadias. Cases were defined as confirmed cases of hypospadias for the years 2007–10 by BDSS-PR (n = 279), with or without another congenital condition. Control subjects were defined as those male newborns with no hypospadias and

EP

no other congenital condition according to the PRBC dataset for the years 2007–10 (n = 91,615). We performed univariate analyses to estimate the prevalence odds ratio for hypospadias for the

AC C

following variables: weeks of gestation (25–37 or 38+), birthweight (< 2,500 g or ≥ 2,500 g), previous births (0 or 1+), father’s age (< 20, 20–24, 25–29, 30–34,35–39, 40+ years), mother’s age (< 20, 20–24, 25–29, 30–34,35–39, 40+), and health insurance status (private, including pay out-of-pocket cases, vs. government) with 95% confidence interval (CI). Age of the father and age of the mother category of 40 years or older was aggregated due to small sample size. Eligibility for the government health insurance plan in Puerto Rico is based on the relation of family income to family size, which is similar to US poverty guidelines [7]. Therefore, health insurance status was used as the criterion to provide an insight about

4

ACCEPTED MANUSCRIPT

socioeconomic status at the household level [8], which according to the American Community Survey Briefs, United States Census Bureau, 45% of the population in Puerto Rico in 2010 lived under the poverty line. A multivariate analysis to estimate the joint prevalence odds ratio for hypospadias was

RI PT

performed with those variables which their 95% CI for odds ratio did not include 1. Statistical analyses were performed with SPPS v. 17 (IBM). Spatial distribution of hypospadias’ prevalence

SC

The annual prevalence at birth of hypospadias for the Puerto Rican archipelago (the islands of Puerto Rico, Vieques, and Culebra) was calculated by dividing the number of hypospadias cases by the

M AN U

corresponding number of male births for each year (2007–10). The overall 4-year prevalence at birth for hypospadias for the study period according to the mother’s municipality of residence was used to determine patterns in the spatial distribution of hypospadias prevalence. In order to statistically assess the existence of a series of neighboring municipalities with similar or dissimilar prevalence of

TE D

hypospadias beyond what would be expected by chance, cluster analyses were conducted using two types of geographical information systems methods using the place of residence of the mother: (a) Cluster and Outlier Analysis (Anselin Local Moran’s I), to test for the plausible existence of clusters; and

EP

(b) Getis-Ord G, to test for the plausible clustering of hot spots (high prevalence) and cold spots (low prevalence) of hypospadias. The algorithms for cluster analyses were based on a comparison of each

AC C

individual region (municipalities within the Puerto Rican archipelago) with its neighboring municipalities. Moran's I value, z-score, p ≤ 0.05, and a code representing the cluster type for each prevalence

per municipality were calculated, where z-scores and p-values represent the statistical significance of the computed index values. A positive value for Moran’s I indicates that a given municipality has neighboring municipalities with a similarly high (or similarly low) prevalence of hypospadias which defines a spatial cluster. A negative value for Moran’s I indicates that a given municipality has neighboring municipalities with dissimilar values, which defines a spatial outlier. The output field,

5

ACCEPTED MANUSCRIPT

cluster/outlier type (COType), distinguishes between statistically significant cluster of high values (HH) at p ≤ 0.05, cluster of low values (LL), outliers in which a high value is surrounded primarily by low values (HL), and outliers in which a low value is surrounded primarily by high values (LH). Prevalence data were

RI PT

aggregated to detect hot and cold spots. Cluster analyses were performed with ARCGIS v10 (ESRI). Results

As in other countries, hypospadias is the most common urogenital tract condition in Puerto Rico. For the

SC

time period 2007–10, we found that hypospadias prevalence remained relatively constant with an average prevalence of 30.2/10 000 male live births (Fig. 1A). Analysis by comorbidity showed that 251

M AN U

out of 279 hypospadias cases did not have another congenital condition, whereas 14 cases had one additional condition and 14 cases had two or more additional congenital conditions (Fig. 1B). Those cases with one comorbidity were associated with cardiovascular (n = 6), musculoskeletal (n = 3), central nervous system (n = 1), oral (n = 1), eyes and ears (n = 1), chromosomal–genetic (n = 1), or genital

Case–control study

TE D

ambiguity (n = 1) conditions.

Table 1 shows that there is an increased odds ratio of having a newborn with hypospadias according to

EP

gestational age (25–37 weeks), birthweight (< 2,500 g), age of the mother (40+ years), age of father (30– 34 years), and health insurance status (private insurance). A positive linear correlation for parental age

AC C

was noted (Pearson correlation coefficient = 0.69; p < 0.01). Therefore, the age of the father was not included in our multivariate logistic regression analysis as it showed the lowest OR value between these two variables and it did not meet the assumption of absence of multicollinearity. Similarly, a positive linear correlation between gestational age and birthweight was noted (Pearson correlation coefficient = 0.78; p < 0.001). Therefore, gestational age was not included in our multivariate logistic regression analysis as it showed the lowest OR value between these two variables and it did not meet the assumption of absence of multicollinearity.

6

ACCEPTED MANUSCRIPT

Multivariate analyses confirmed statistical significance as follows: birthweight less than 2,500 g, OR = 3.84 (95% CI 2.99–4.95); age of the mother of 40+ years, OR = 2.63 (95% CI 1.19–5.81); and private health insurance, OR = 1.41 (95% CI 1.07–1.86).

RI PT

Spatial distribution of hypospadias prevalence

The Puerto Rican Archipelago is divided into 78 municipalities, which include the mainland of Puerto Rico with a population of 3,725,778, the island municipality of Vieques with a population of 9,301, and

SC

the island municipality of Culebra with a population of 1,818. Therefore, the population of mainland Puerto Rico constitutes 99.7% of the total population of the archipelago [9]. Figure 2A shows those

M AN U

municipalities below the average value of hypospadias prevalence and those above the average value of hypospadias prevalence for the years 2007–10. Cluster and outlier analysis (Anselin Local Moran’s I) detected six contiguous municipalities in the north-central region of the Island (from west to east: Vega Baja, Vega Alta, Dorado, Toa Baja, Cataño, and Bayamón) with a similar high prevalence of hypospadias (Fig. 2B, shown in black). Therefore, this cluster of municipalities is statistically different from the rest of

TE D

the mainland with regard to hypospadias prevalence (p ≤ 0.05). Two municipalities (from north to south: San Sebastián and Maricao) were identified as spatial outliers for hypospadias prevalence where a high

EP

value is surrounded by low values (HL). Three municipalities (from north to south: Ceiba, San Lorenzo, and Naguabo) were identified as having low values (LL) for hypospadias prevalence. By using a second

AC C

geographical information system, Getis-Ord G analysis, clustering of a hot spot (high hypospadias prevalence) in the north-central region was confirmed (Fig. 2C). Discussion

The prevalence of hypospadias in Puerto Rico has remained relatively constant during the years 2007– 10. The calculated prevalence was 30.2/10 000 male live births for this time period. This low prevalence allows us to interpret the odds ratio as a measure of risk. We found that 90% of these cases did not exhibit comorbid congenital conditions. Birthweight (< 2,500 g), age of the mother (40+ years), and

7

ACCEPTED MANUSCRIPT

private health insurance were found to be risk factors for having a newborn with hypospadias. A cluster of hypospadias cases was detected in the north-central region of the mainland. Given the protocol for case ascertainment in the Island, the clustering of cases cannot be a statistical artifact related to a

RI PT

regional bias in the medical competency of the health professionals related to the surveillance system. Health-care provider density does not explain the clustering of cases either, as the highest densities for providers (San Juan, Mayaguëz, and Ponce metropolitan municipalities) fell outside the hot spot for

SC

hypospadias prevalence.

The prevalence and epidemiologic profile of hypospadias in Puerto Rico further supports the

M AN U

notion that cases are not particularly different from other regions of the world. Our study is also consistent with most recent studies, which report stable or declining prevalence of hypospadias [5,10– 15]. Australia [16] and Denmark [17] are noted as exceptions. We found an association between hypospadias and low birthweight, as has been shown particularly for severe hypospadias [18–20]. But

TE D

the strong association between birthweight and gestational age in our sample, and the lack of information on severity for low birthweight cases, does not allow us to determine which of the two is the proxy variable that underlies hypospadias cases. Nevertheless, in agreement with recent studies, we

EP

also found that gestational age is associated with hypospadias [20,21]. Strong evidence for early delivery as a risk factor for congenital urogenital conditions is provided by Jensen et al. [20], suggesting that

AC C

prenatal factors that affect relative fetal growth restriction can also produce hypospadias. However, it is important to note that global growth deficits among hypospadiac boys do not seem to persist through infancy [22]. Other reports have shown that younger [23] or older [24] paternal age is associated with increased hypospadias risk. In our sample, we found a strong positive linear correlation between maternal and paternal age, which confounds the plausible contribution of each as a risk factor for the condition. According to several studies, older maternal age has been identified as a risk factor for hypospadias [10,25,26], but it has been difficult to pinpoint the relevance of this association with

8

ACCEPTED MANUSCRIPT

hypospadias etiology. For instance, is age associated with underlying genetic defects associated with aging or with subfertility? We did not conduct analyses for multiple births (n = 6) or gestational diabetes mellitus (n = 6) due to the small number of cases in the sample.

RI PT

We found that private health insurance is associated with the condition. Our multivariate logistic regression analysis indicates that the effect of health insurance status persists after controlling for age of mother. Therefore, it will be advantageous for further lines of inquiry to determine social class

SC

differences following a social theory-based model beyond traditional socioeconomic status measures [27,28]. In this context, it will be important for future studies to inquire about the use of assisted

M AN U

reproductive technologies among women with private health insurance. In addition, further analyses by type and place of employment may reveal patterns of significant population mobility on a daily basis, which adds to the complexity of plausible environmental factors associated with the condition. Our study indicates that hypospadias prevalence according to the mother’s residence is not

TE D

distributed uniformly across the island of Puerto Rico as identified with prevalence values. The cluster of hypospadias cases in the north-central region of Puerto Rico was detected by using two separate approaches: Moran’s Global I and Getis-Ord G. The first one identifies statistically significant spatial

EP

clusters and spatial outliers, while the second one detects locations with high or low prevalence (hot and cold spots, respectively). While Anselin Local Moran’s I analysis identified six contiguous

AC C

municipalities in the north-central region of Puerto Rico with high hypospadias prevalence, Getis-Ord G identified a gradient pattern across the mainland from north to south with the hot spot in the northcentral region as well. Getis-Ord G is based on inferential statistics where results of the analysis are interpreted within the context of the null hypothesis, which states that there is no spatial clustering of feature values. But if the z-score is positive or negative, it indicates that high values or low values are clustered in the study area, respectively. The use of Global Moran’s I becomes particularly important when both high and low values produce clusters as these may cancel each other out because Global

9

ACCEPTED MANUSCRIPT

Moran’s I provides a spatial autocorrelation analysis which identifies the spatial distribution of high values when surrounded by low values, and vice versa. In this study, the spatial autocorrelation refers to the relationship of computed I values with neighboring municipalities. Taken together, both GIS

RI PT

methods do not require data filtering and both methods assume complete spatial randomness; that is, values are randomly distributed among the features in the dataset which assumes random spatial processes at work. In our case, the total number of hypospadias cases and the total number of male live

SC

births without a congenital condition per municipality per year were submitted to built-in analytical algorithms of each method without making assumptions about frequency of cases nor space. The fact

M AN U

that both methods detected spatial clustering of hypospadias provides strong evidence that such spatial distribution was not produced by an artifact introduced by the number of hypospadias cases nor from the geographical units that were used in the analyses.

According to four local and federal government reports (data not shown), the following

TE D

environmental contaminants are found in discrete regions of the island: chlorinated pesticides, polychlorinated biphenyls, phthalates, and dioxins; all of which have been associated with hypospadias [29]. It is of interest for future studies to determine whether there is a correlation between the spatial

EP

distribution of these contaminants and hypospadias clusters according to severity of the condition. It remains as a tenet in the field that most hypospadias cases are idiopathic. In spite of great

AC C

research efforts, it is plausible that the etiology of the condition has remained elusive because studies that screen for risk factors are not conducted according to severity [19]; for a recent review see [30]. Therefore, it remains as a challenge for future studies to determine the distribution of hypospadias prevalence for the Puerto Rican archipelago according to severity type. Conclusion The clustering of hypospadias prevalence provides an opportunity to assess the underlying causes of the condition and their relationships with geographical space.

10

ACCEPTED MANUSCRIPT

Conflict of interest None. Funding

RI PT

None. Ethical approval

This study was approved by the Institutional Review Board (IRB) Committee.

SC

References

M AN U

[1] Baskin LS. Can we prevent hypospadias? J Pediatric Urol 2007;3:420–5.

[2] Willingham E, Baskin LS. Candidate genes and their response to environmental agents in the etiology of hypospadias. Nat Clin Pract Urol 2007;4:270–9.

[3] Paulozzi LJ. International trends in rates of hypospadias and cryptorchidism. Environ Health Perspect

TE D

1999;107:297–302.

[4] Paulozzi LJ, Erickson JD, Jackson RJ. Hypospadias trends in two US surveillance systems. Pediatrics 1997;100: 831–4.

AC C

39.

EP

[5] Fisch H, Hyun G, Hensle TW. Rising hypospadias rates: disproving a myth. J Pediatric Urol 2010;6:37–

[6] Edmonds LD, Layde PM, James LM, et al. Congenital malformations surveillance: two American systems. Int J Epidemiol 1981;10:247–52. [7] Krieger N, Williams DR, Moss NE. Measuring social class in the US public health research: Concepts, methodologies, and guidelines. Annu Rev Public Health 1997;18:341–78

11

ACCEPTED MANUSCRIPT

[8] U.S. Department of Health & Human Services. (January 24). Annual update of the HHS Poverty Guidelines. Federal Register 2013;78(16):5182–3.

RI PT

[9] United States Census Bureau. American Fact Finder. Retrieved July 2, 2013 from http:factfinder2.census.gov/faces/nav/jsf/pages/community_facts.xhtml.

prevalence trends. Pediatrics 2005;115:e495–e499.

SC

[10] Porter MP, Faizan MK, Grady RW, et al. Hypospadias in Washington State: maternal risk factors and

[11] Abdullah N, Pearce M, Parker L, et al. Birth prevalence of cryptorchidism and hypospadias in

M AN U

northern England, 1993–200. Arch Dis Child 2007;92:576–79.

[12] Correa A, Cragan JD, Kucick JE, et al. The Metropolitan Atlanta Congenital Defects Program 40th anniversary edition surveillance report. Birth Defects Res A Clin Mol Teratol 2007;79:1–186. [13] Ghirri P, Scaramuzzo RT, Bertolloni S, et al. Prevalence of hypospadias in Italy according to severity,

TE D

gestational age, and birthweight: an epidemiological study. Ital J Pediatr 2009;35:18. [14] Elliott CS, Halpern MS, Paik J, et al. Epidemiologic trends in penile anomalies and hypospadias in the

EP

state of California, 1985–2006. J Pediatr Urol 2011;7:294–298. [15] Loane M, Dolk H, Kelly A, et al. Paper 4: EUROCAT statistical monitoring identification and

AC C

investigation of ten year trends of congenital anomalies in Europe. Birth Defects Res A Clin Mol Teratol 2011;91(Suppl 1):531–543.

[16] Nassar N, Bower C, Barker A, et al. Increasing prevalence of hypospadias in Western Australia, 1980–2000. Arch Dis Child 2007;92:580–584. [17] Lund L, Engebjerg MC, Pedersen L, et al. Prevalence of hypospadias in Spanish boys: a longitudinal study, 1977–2005. Eur Urol 2009;55:1022–1026.

12

ACCEPTED MANUSCRIPT

[18] Carlson WH, Kisely SR, MacLellan DL, et al. Maternal and fetal risk factors associated with severity of hypospadias: a comparison of mild and severe cases. J Pediatr Urol 2009;5:283–6.

RI PT

[19] Brouwers MM, van der Zanden LF, de Gier RP, et al. Hypospadias risk factor patterns and different phenotypes. BJU Int 2010;105:254–62.

[20] Jensen MS, Wilcox AJ, Olsen J, et al. Cryptorchidism and hypospadias in a cohort of 934, 538 Danish

SC

boys: the role of birth weight, gestational age, body dimensions, and fetal growth. Am J Epidemiol 2012;175:917–25.

M AN U

[21] Bang JK, Lyu SW, Choi J, et al. Does infertility treatment increase male reproductive tract disorder? Urology 2013;81:644–8.

[22] Hsieh MH, Alonzo DG, Gonzales ET, et al. Expremature infant boys with hypospadias are similar in size to age-matched, ex-premature infant boys without hypospadias. J Pediatr Urol 2011;7:543–7.

Epidemiology 1995;6:282–8.

TE D

[23] McIntosh GC, Olshan AF, Baird PA. Paternal age and the risk of birth defects in offspring.

[24] Materna-Kiryluk A, Wiśniewska K, Badura-Stronka M, et al. Parental age as a risk factor for isolated

EP

congenital malformations in a Polish population. Paediatr Perinat Epidemiol 2009;23:29–40.

AC C

[25] Fisch H, Golden RJ, Libersen GL, et al. Maternal age as a risk factor for hypospadias. J Urol 2001;165:934–936.

[26] Carmichael SL, Shaw GM, Nelson V, et al. Hypospadias in California: trends and descriptive epidemiology. Epidemiology 2003;14:701–706. [27] Kreiger N. Epidemiology and the people’s health: Theory and context. Oxford University Press: New York; 2011.

13

ACCEPTED MANUSCRIPT

[28] Wright EO. Classes. Verso: New York; 1989. [29] Kalfa N, Philibert P, Baskin LS, Sultan C. Hypospadias: interactions between environment and

RI PT

genetics. Mol Cell Endocrinol 2011;30:89–95 [30] Carmichel SL, Shaw GM, Lammer EJ. Environmental and genetic contributors to hypospadias: a review of the epidemiologic evidence. Birth Defects Res A 2012;94:499–510.

SC

Figure 1 Profile of hypospadias prevalence for the Puerto Rican archipelago, 2007–10. (A) Hypospadias prevalence per 10,000 male live births is shown for years 2007–10. The dotted line indicates the average

M AN U

hypospadias prevalence for this time period. (B) Most of hypospadias cases are isolated (90%), whereas 5% exhibit one comorbid congenital condition and 5% exhibit two or more congenital conditions (see text for details).

Figure 2 Spatial distribution of hypospadias prevalence for the Puerto Rican archipelago, 2007–10. (A)

TE D

Municipalities below the average value of hypospadias prevalence are shown in white whereas municipalities above the average value of hypospadias prevalence are shown in grey and black. Puerto Rico is the largest island of the Puerto Rican archipelago followed by the island municipality of Vieques.

EP

Culebra is the smallest island municipality of this archipelago within the Caribbean region. North cardinal point applies to (A–C). Geographic scale (size in km) and legend for hypospadias prevalence are

AC C

shown (see text for details). (B) Six contiguous municipalities in the north-central region of the mainland with the highest prevalence of hypospadias are shown in black. Two municipalities identified as spatial outliers (High Low) are shown in the darkest grey tone. Three municipalities identified as spatial outliers (Low Low) are shown in the lightest grey tone. Geographic scale (size in km) and legend for Anselin Local Moran’s I values are shown (see text for details). (C) Cold spots for low hypospadias prevalence are shown in white or hatched lines. Hot spots for high hypospadias prevalence are shown in black.

14

ACCEPTED MANUSCRIPT

Municipalities between cold and hot spots are shown in grey tones. Geographic scale (size in km) and legend for Getis–Ord G values are shown (see text for details).

Hypospadias

Controls

Total

RI PT

Table 1 Risk factors for hypospadias in Puerto Rico.

Prevalence

a

279

91,615

91,894

30.36

Gestational age

275

90,948

91,239

30.14

25–37 weeks

140

34,074

34,214

38–44 weeks

135

56,874

56,384

Birthweight

279

91,576

91,855

30.37

90

10,166

10,256

87.75

189

81,410

81,599

23.16

279

91,601

91,880

> 2,500 g

1.73

23.77

1

3.81

95% CI

1.36–2.19 b

2.96–4.90 b

273

90,701

90,974

30.01

1

6

900

906

66.23

2.22

0.98–4.99

268

88,479

88,747

19

64,81

65,00

29.23

1.19

0.71–2.00

77

23,804

23,881

32.24

1.31

0.94–1.84

25–29

60

24334

24394

24.60

1

30–34

64

17,858

17,922

35.71

1.45

1.02–2.07 b

35–39

25

9,566

9,591

26.07

1.06

0.66–1.70

40–44

15

4,005

4,020

37.31

1.52

0.86–2.68

8

2,431

2,439

32.80

1.33

0.64–2.80

1+ Paternal age < 20

AC C

20–24

EP

0

TE D

Prior births

40.92

M AN U

< 2,500 g

SC

Total

OR

45 +

15

ACCEPTED MANUSCRIPT

91,586

91,865

< 20

46

16,231

16,277

28.26

1.11

0.63–1.96

20–24

80

29,511

29,591

27.04

1.06

0.62–1.82

25–29

73

23,453

23,526

31.03

1.22

0.71–2.10

30–34

54

14,716

14,770

36.56

1.44

0.82–2.51

35–39

16

62,65

6,281

25.47

1

≥ 40

10

1,410

1,420

70.42

2.77

Health insurance

279

91,383

91,662

Government

169

62,551

62,720

27.12

1

Private

110

28,832

28,942

38.01

1.41

M AN U

SC

1.26–6.13 b

1.11–1.80 b

TE D

OR = odds ratio.

RI PT

279

Maternal age

Per 10 000 male live births,

b

Significant at 95% confidence interval (CI).

EP

a

names

AC C

AQ1: refs 6, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 24, 25, 26: please provide up to 6 author

AQ2: please check that the figures have been correctly cited in the text

16

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT