Spss program version 19

We calculated the sample size on the basis of 80% power and a 5% two-tailed significance level. We assumed that we would need to explore modest differences of a continuous variable between two groups of unequal size. We therefore calculated the sample size on the basis of detecting a difference of half a standard deviation between two groups the smallest of which comprises 10 of the totals. This yielded a required sample size of 350 (https://epitools.ausvet.com.au/twomeanstwo). We then doubled this to take into account factors such as refusals and non-response.
Data were analyzed using the SPSS program version 19 (IBM Corp. Released 2010. IBM SPSS Statistics for Windows, Version 19.0. Armonk, NY: IBM Corp). Besides standard descriptive and basic analytical statistics such as Chi-square tests, two multivariate statistical methods, linear regression to assess the determinants of vitamin D level and binary logistic regression to identify determinants of binary outcomes, were used. A significance level of 0.05 was used throughout.

A cross-sectional descriptive study was performed of all patients diagnosed clinically with 21OHD from January 2000 to December 2018 at the National Institute of Endocrinology, Havana, Cuba. The local ethical committee approved the study and informed consents were obtained.
For the molecular analysis of the CYP21A2 gene, a protocol designed and approved by the National Center for Medical Genetics was used. The first stage used a 2-phase Polymerase Chain Reaction (PCR) and in the second stage, 5 different point mutations were detected (Intron 2, 8 bp Deletion, G318X, I172N and P30L). Characteristics of the mutational analysis in the CYP21A2 gene are shown in Table 1.
Table 1

Characteristics of the mutational analysis in the CYP21A2 gene

Mutation	Primers	Product of PCR (bp)	Restriction enzyme	Normal	Mutated
Intron 2	P7 P8	378	HhaI	378	24, 354
Pro-30-Leu	P5 P6	249	HhaI	21, 228	249
I172N	P11P2		Taq I	416	394
Deletion 8-bp	P9 P10	89	–	89	81
Gln-318-Stop	P12 P13	136	PstI	25, 111	136

Statistical analysis was performed using the SPSS program (version 19). Frequency distributions of qualitative variables were obtained, as well as mean (or median) and standard deviation (or interquartile range) according to whether the distribution was normal (or not). A p-value of less than 0.05 was considered statistically significant.

We expressed normally distributed continuous variables as means ± standard deviation and when asymmetrically distributed as median (interquartile range). The Kolmogorov–Smirnov test examined the normality of the continuous variables. An unpaired Student’s t-test was used to test normally distributed continuous variables and the Mann–Whitney U-test in asymmetrically distributed variables. We compared discrete variables with the Pearson χ² test. Pearson χ² test was similarly used to test whether the genotype distribution deviates from Hardy–Weinberg equilibrium. A significant relationship between two categorical variables was determined by the Fisher’s Exact test when cells with expected frequencies < 5 were identifiedin a contingency table.
Moreover, a stepwise multiple logistic regression was used for all variables that showed significant deviations in univariate analysis.
Statistical significance was considered with a p-value of ≤0.05. SPSS program version 19 (SPSS Inc., Chicago, IL, USA) was used for performing statistical analysis.

Statistical analysis was performed with SPSS program version 19 (SPSS Inc., Chicago, IL, USA). Normally distributed continuous variables were presented as means ± standard deviation and in the case of asymmetrical distribution as median and interquartile range. The Kolmogorov–Smirnov test was used to examine the normality of the continuous variables. Normally distributed continuous variables were tested using unpaired Student’s t-test. Asymmetrically distributed variables were tested with the Mann–Whitney U-test. Discrete variables were compared with the Pearson χ² test. The Pearson χ² test was also used to determine if the genotype distribution deviates from Hardy–Weinberg equilibrium. Fisher’s Exact test was used to determine a significant relationship between two categorical variables if cells with expected frequencies lower than five were found in a contingency table. A stepwise multiple logistic regression was performed for all variables that showed significant deviation in univariate analysis. A p-value of <0.05 was considered to be of statistical significance.