Correlation between studying strategies, personal and psychological factors with academic achievement and intelligence in health sciences university students: a cross-sectional study

Abstract

Introduction

To date, there are no sufficient studies aimed to determine a correlation between personal, academic, and psychological variables with academic achievement, measured with the grade point average (GPA) and intelligence in university students according to each sex.

Study aim

To determine the correlation between studying strategies, personal and psychological factors with GPA and intelligence in a sample of health sciences university students.

Methods

Health Sciences university students, were invited to participate, those who accepted were cited in a computer room where they signed an informed consent and filled an electronic questionnaire with sociodemographic, behavioral, psychological variables and studying strategies (from the MLSQ instrument) afterwards they performed a verbal and non-verbal intelligence test (Shipley-2).

Results

A total of 439 students were included, from which 297 (67.7%) were women. The mean of age was 20.34 ± 2.61 years old. We found that no differences in GPA where observed between sexes. We detected a higher correlation between combined intelligence and GPA in women than in men. In addition, most studying strategies showed a higher correlation with GPA than intelligence scores in men´s sample. All these findings coincide with the fact that preparatory GPA was the most correlated variable with university GPA in both sexes. Finally, women showed higher levels of the sum of diseases, somatization, anxiety, depression and academic stress than men, and all these variables showed low significant correlations with the combined intelligence score only in women´s sample.

Conclusion

Verbal and non-verbal intelligence scores show a lower association to GPA in men than in women, while studying strategies showed a higher association with GPA in men than in women.

Introduction

Academic achievement is usually measured with the Grade Point Average (GPA), and it has been correlated with a variety of, sociodemographic, health, psychological and intellectual factors in different populations [1,2,3,4,5,6,7,8,9]. Among the psychological factors it has been observed a positive correlation between GPA and emotional intelligence, academic motivation, task-oriented copping strategies and specific personality traits as well as negative correlations with stress, depression, stressful life events and avoidance coping [2,3,4,5,6,7,8,9]. Sociodemographic and behavioral variables correlated with higher GPA are female sex, sleep quality and physical activity [5, 6, 8, 10]. While intellectual variables correlated with higher GPA are verbal and non-verbal intelligence [5, 6, 11]. In addition, many socio-demographic, psychological and health-related variables have been correlated with verbal and non-verbal intelligence, including the maternal schooling, specific personality traits, psychosocial adversities and low income [12, 13]. However, to date, there are no studies which have searched the correlation between personal and psychological factors, along with studying strategies with GPA and intelligence in university students, in order to determine which factors are correlated with these two dependent variables by adjusting by the rest of the variables. In this sense, in a previous report of the research team we found some personal and psychological variables correlated with GPA and non-verbal intelligence in university students with low academic performance. Among these associations we found a higher association between non-verbal intelligence and GPA in the women´s sample than in men´s sample, while different variables were correlated with intelligence in each sex, including positive correlations between non-verbal intelligence with anxiety, depression, somatization and the number of diseases in the women´s sample but not in the men´s sample [14]. Therefore, the objective of our study was to corroborate these findings in a sample of university students with normal academic achievement, and additionally, we decided to add more variables to the study, including academic procrastination and studying strategies. Therefore the objective of the study was to determine the correlation between personal and psychological factors along with learning strategies with GPA and intelligence (verbal and non-verbal) in a sample of health sciences university students with normal academic performance. The hypotheses of our study are: (a) Women´s sample have a higher correlation between verbal and non-verbal intelligences with GPA than men´s sample; (b) health-related variables (including anxiety, depression, somatization and the number of diseases) are more correlated with intelligence in women´s sample than in men´s sample, (c) studying strategies differ in each sex and correlate differently with GPA and intelligence in each sex.

Subjects and methods

Ethical considerations

The study was conducted according to the guidelines of the Declaration of Helsinki and was approved by the ethical committee of the Health Sciences University Center, with the registration number CI-06022 (Approval date September 30, 2022). All personal and health-related data of participants were handled with strict confidentiality, being used only for research purposes.

Informed consent statement

All the participants signed an informed consent form.

Subjects

The inclusion criteria of the study were: (a) Active students of the Health Sciences University Center of the University of Guadalajara (all of them of mestizo ethnical origin), (b) Students older than 18 years old, (c) Those who accepted to participate in the study and signed an informed consent. The elimination criterium was: a) Those students who did not complete the measurement of all the instruments and tests.

Study design

This is a cross-sectional design by reason that no experiments nor follow ups in the measurements were performed. The independent variables were the sociodemographic, psychological ones and the studying strategies, while de dependent variables were the GPA and intelligence.

Sample size

The sample size was calculated with the formula for bivariate correlations [15], this formula was performed to detect a very low correlation as significant (r = 0.2) with a statistical confidence of 95% and a statistical power of 80%. The result of the formula yielded a total of 67 subjects; however, the minimum intended sample size consisted in 100 individuals per sex.

Procedures

Health sciences university students were invited to participate directly in their classrooms. Those who accepted were cited in a computational room of the Health Sciences University Center of the University of Guadalajara, where they signed an informed consent and filled an electronic questionnaire with sociodemographic, behavioral and psychological variables, afterwards they performed a verbal and non-verbal intelligence test. The measurement of these variables lasted around 1 h. The study was performed from October to December of 2022.

Variables

In order to diminish the confusion bias, we included a wide range of possible confounding variables including sociodemographic, psychological, academic ones and studying strategies.

Sociodemographic and health variables

The following sociodemographic variables were measured: age, sex, schooling, maternal and paternal schooling, number of siblings, daily physical activity minutes, whether they had a romantic partner, whether they had a job, whether they had children, monthly money available for they personal expenses, measured with five categories which ranged from nothing to more than 180 USD. We also measured the frequency of alcohol and smoking consumption, as well as the frequency of consumption of illegal drugs including: marijuana, hashish, ecstasy, cocaine/crack, heroin, amphetamines, LSD and hallucinogenic mushrooms, the consumption of any of these legal or illegal drugs ranged from never to many times in the week. The presence of the following acute or chronic diseases was also measured: diabetes, hypertension, kidney/bladder problems, cancer, respiratory infections (including COVID), depression that requires medication, anxiety that requires medication, gastritis/gastric ulcer, overweight, allergies/asthma, migraines, skin problems (acne, neurodermatitis), rheumatic diseases (arthritis, systemic lupus erythematosus, ankylosing spondylitis), thyroid problems, colitis/irritable colon, sinusitis, anorexia/bulimia, intestinal infections, hearth attack/angina pectoris, brain stroke, high cholesterol and any additional one. Sleep satisfaction was measured with the first item of the OVIEDO sleep questionnaire, with the answer options ranging from 1 (very unsatisfied) to 7 (very satisfied) [16], and sleep quality was measured with the second item (consisting in 5 items) of this scale, which ranged from 1 to 5 (low quality to high quality) [16].

Psychological variables

The following psychological variables were measured: depression with the instrument PHQ-9 [17], anxiety with the instrument GAD-7 [18], somatization with the instrument PHQ-15 [19], academic stress with the academic stressor´s subscale of the SISCO scale [20, 21], active coping was measured with the 4 items of active coping of the Mini-COPE scale [22], the subscales of the emotional intelligence questionnaire TEIQUE (emotional perception, emotional management, and self-motivation) were measured with four to five items for each subscale [23] (Supplementary file 1). Finally, we measured the subscale of positive relations with others of the shortened version of the psychological well-being scale [24].

Academic variables and intelligence

We obtained the following academic variables from the University records: GPA, preparatory GPA and the admission exam text, which consisted in the college board test, which is composed from math and verbal reasoning and indirect redaction. Finally, the verbal (vocabulary) and non-verbal (abstraction) intelligence was measured with the Spanish translated version of the Shipley-2 instrument [25]. This last scale was measured with the standard results of the verbal and abstraction subscales, as well as with the combined standard result of both subscales.

Studying strategies

The following studying strategies form the Motivation Strategies for Learning Questionnaire (MSLQ) instrument [26]: 7 learning strategies scales, including: time and study environment, critical thinking, rehearsal, effort regulation, self-regulation, elaboration and organization; and 2 motivation scales: task value and intrinsic goal orientation, were translated to the Spanish and measured to the sample (Supplementary file 2). In addition, academic procrastination was measured with the short version of the academic procrastination scale [27], this scale was also translated to the Spanish (supplementary file 3).

Statistical analysis

In order to describe the data, we used mean and standard deviations for quantitative variables with parametric distribution and median and ranges for quantitative variables with non-parametric distribution. Frequencies and percentages were used to describe qualitative variables. In order to compare qualitative variables between sexes we used Chi square test and Fisher exact test and to compare quantitative variables between sexes we used T-test for independent samples and Mann-Whitney U test, depending on the parametric or non-parametric of the data. In order to perform correlations between the studied variables with the two dependent variables: GPA and the combined intelligence score, we used Pearson and Spearman correlation tests depending on the parametric or non-parametric distribution of the data. Finally, to determine the independent variables more correlated with both dependent variables: GPA and the combined intelligence score, we used the multiple linear regression analysis with the step-wise method, this method was used to detect models with all the variables within the model being significant, which permitted us to identify the variables significantly correlated (from the most to the least correlated ones) with the dependent variables after adjusting for the rest of the variables (confounders), which diminishes the confusion bias.

For the adjusted models of the global sample, we included around 45 independent variables, which is accepted according to the sample size (439 students), because around 10 individuals per independent variable were included. Although, for adjusted models in each sex the sample size diminished in each sex, we consider that the performance of a linear regression is acceptable by considering that this analysis could permit us to detect the main correlated variables with the dependent ones, by performing an adjusting of the confounders in a reliable way, this is sustained by the concordance of the most correlated variables (in the adjusted models) with the bivariate analysis.

Additionally, for all scales and subscales used in the study, the Cronbach alpha test was performed, in which a value above 0.6 was considered acceptable. Effect sizes were calculated with the JASP program [28], by obtaining the Cohen´s r value and rank biserial correlation, for parametric and non-parametric distributions respectively. The effect size was classified as small < 0.4, moderate: from 0.4 to 0.6, and high > 0.6. The participants that did not competed any scale of the questionnaire or an intelligence test were eliminated from the database.

All the analyses were performed in SPSS program v. 25 and a p value < 0.05 was considered as significant.

Results

The Cronbach alpha test of all scales and subscales used in the study were above 0.6, which means that they are reliable for analysis.

A total of 440 students accepted to participate, which represents around 60% of invited students, those who did not want to participate mentioned the lack of time as the main reason; however, after eliminating the participants with incomplete scales or tests, a total of 439 students were included, from whom 297 (67.7%), were women, the mean ± SD of age was: 20.34 ± 2.61 years old, with a range form 18–54 years. The participants were studying one of 6 different Bachelor´s programs in health sciences, including: nursing, physical culture and sport, medicine, psychology, dental prothesis and higher university technician in emergencies, work safety and rescues.

The descriptive data of all studied variables and their comparison between sexes are shown in Table 1. There we can observe that women´s sample had higher preparatory GPA than men´s sample, men´s sample showed higher values of verbal, non-verbal intelligence and combined intelligence scores than women´s sample as well as higher scores in the admission exam test; however, all these differences had a small effect size. In relation with health and psychological variables, we observed that women´s sample showed higher levels of: sum of diseases, daily physical activity minutes, somatization, academic stress, depression and anxiety than men´s sample, from these comparisons, all the variables had a small effect size. While men´s sample showed higher levels of emotion perception, emotion management and sleep quality. With respect to the studying strategies, we observed that women´s sample showed higher levels of task value, rehearsal and organization, while men´s sample showed higher levels of critical thinking. Likewise, all these comparisons had a small effect size (Table 1).

Table 1 Descriptive data of the studied variables and their comparison between sexes

The possible value of GPA, preparatory GPA and admission exam test is ≤ 100. The possible value for all intelligence scores is ≤ 145. Psychological scales and studying strategies were scored as follows: Active coping (Mini-COPE), depression (PHQ-9) and anxiety (GAD-7): from 1 (low) to 4 (high); somatization (PHQ-15): from 1 (low) to 3 (high); emotional intelligence subscales from TIEQUE scale: emotion perception, self-motivation and emotion management: from 1 (totally disagree) to 7 (totally agree). For academic stress (SISCO scale), positive relations with others (PWB scale), sleep quality (OVIEDO scale) and all the studying strategies (time and study environment, critical thinking, task value, rehearsal, effort regulation, self-regulation, elaboration, intrinsic goal orientation, organization (MSLQ questionnaire) and academic procrastination): from 1 (low) to 5 (high).

Bivariate correlations

Global sample for GPA

In the bivariate correlations in the global sample we observed that the variables most correlated with GPA were: the preparatory GPA (with a moderate positive correlation), the admission exam text and the verbal intelligence, these last with a low positive correlation; additionally, all the studying strategies also showed low positive correlations (Table 2). Very low positive correlations were also found between GPA and self-motivation, positive relations with others and active coping. In addition, very low negative significant correlations were observed with: daily free hours, having a job, academic procrastination, smoking, alcohol and illegal drugs consumption frequency and daily physical activity minutes.

Table 2 Bivariate correlations between the studied variables with GPA and combined intelligence in the combined sample and segmented by sex

Global sample for intelligence

In the case of intelligence, we observed a moderate positive correlation with the admission exam test, followed by a low positive correlation with GPA, and very low positive correlations with male sex, age, paternal and maternal schooling, emotion perception, positive relations with others, active coping, academic procrastination and critical thinking; while very low negative correlations were found between GPA and daily free hours and daily physical activity minutes (Table 2). Combined intelligence also showed high positive correlations with verbal and non-verbal intelligence, this by considering that combined intelligence is composed by these 2 types of intelligence. In addition, the correlation between verbal and non-verbal intelligence in the global sample was: 0.369, p < 0.01.

Sex-specific correlations with GPA

In the women´s sample, we observed that the most correlated variable was the preparatory GPA, followed by the verbal intelligence, the admission exam test and the combined intelligence score, with positive moderate correlations. While in the case of men´s sample, we found that the most correlated variable was the preparatory GPA, with a moderate-high positive correlation, followed by the admission exam test and the verbal intelligence (Table 2). Additionally, in the women´s sample we observed low positive correlations with the following variables: paternal schooling, self-motivation, positive relations with others, active coping and the studying strategies: task value, rehearsal, time and study environment, effort-regulation, self-regulation and organization; and low negative correlations with daily free hours, academic procrastination and illegal drugs consumption frequency.

In the men´s sample, we also observed positive correlations between GPA and positive relations with others, active coping and the studying strategies: task value, time and study environment, effort regulation, critical thinking, self-regulation and elaboration, while very low negative correlations were found with: having an employment, daily physical activity minutes, smoking consumption frequency and academic procrastination (Table 2).

Sex-specific correlations with intelligence

In the women´s sample, the variable most correlated with intelligence was the admission exam test, with a moderate positive correlation, and the following variables showed a low positive correlation: paternal and maternal schooling, age, academic procrastination, positive relations with others, the sum of diseases, somatization, depressive and anxiety symptoms. In addition, low negative correlations were found with: daily free hours, self-motivation and sleep quality.

In the men´s sample, the variable most correlated with intelligence was also the admission exam text, with a moderate positive correlation, followed by positive relations with others, academic procrastination and maternal schooling, while having a job and the daily physical activity minutes showed a low negative correlation (Table 2).

The correlation between verbal and non-verbal intelligence in women´s sample was 0.327, p < 0.01, while in men´s sample it was 0.431, p < 0.01.

Multivariate correlation analyses for GPA

In the multivariate regression analyses for GPA in the global sample, women´s and men´s sample, we observed that preparatory GPA was the most correlated variable with university GPA. The effort regulation, was the second variable most correlated positively with GPA in the global sample and women´s sample, while the studying strategy time and study environment was the second variable most correlated with GPA in men´s sample. In addition, the combined intelligence also correlated positively with GPA in both, the global and women´s sample but not in men´s sample. The studying strategy “rehearsal” showed positive correlations with GPA in the global and women´s sample, as well as academic procrastination which correlated negatively with GPA (Tables 3, 4 and 5). Having an employment and self-motivation showed a negative correlation with GPA only in the global sample, while the daily physical activity minutes showed negative correlations with GPA in the global and men´s sample. The sum of diseases and the monthly money for necessary expenses showed a negative correlation with GPA only in women´s sample, while the socioeconomic level showed a positive correlation with GPA only in women´s sample. Finally, the maternal schooling showed a negative correlation with GPA only in men´s sample (Tables 3, 4 and 5).

Table 3 Multivariate regression analysis for GPA in the global sample

Table 4 Multivariate analysis for GPA in women

Table 5 Multivariate analysis for GPA in men

Multivariate regression analysis for intelligence

In the multivariate regression analysis for the combined intelligence score (excluding the admission exam test), we observed that paternal schooling was the variable most correlated with intelligence in the global and women´s samples, followed by academic procrastination, which was the most correlated variable with intelligence in men´s sample. The emotion perception was a positive psychological variable that correlated positively with the three samples (Tables 6, 7 and 8). The positive relations with others also correlated positively with intelligence in the global and women´s samples, while active coping correlated positively with intelligence only in the global sample. The self-motivation, the daily free hours and smoking consumption frequency correlated negatively with intelligence in the global and women´s samples, and the sum of diseases showed a positive correlation with intelligence in the global and women´s samples (Tables 6, 7 and 8). Finally, the variables: time and study environment, maternal schooling and male sex showed a positive correlation with intelligence in the global sample (Table 6).

Table 6 Multivariate regression analysis for global intelligence in the global sample

Table 7 Multivariate regression analysis for combined intelligence in women

Table 8 Multivariate regression analysis for combined intelligence in men

With respect to sex-specific correlations, we observed that academic stress and the number of siblings showed a negative correlation with intelligence, while age and the studying strategy “task value” showed a positive correlation with intelligence only in women´s sample, and having a job and the physical activity minutes showed a negative correlation with intelligence only in men´s sample (Tables 7 and 8).

Discussion

With respect to the first hypothesis of the study, we found that indeed, women had a higher correlation between verbal and non-verbal intelligence with GPA than men, so this hypothesis was accepted. With reference to the second hypothesis, we also found that women´ sample showed a small positive correlation between health-related variables, including anxiety, depression, somatization and the number of diseases, and the combined intelligence score, while no significant correlation was not found in men´s sample, therefore this hypothesis was also accepted. Finally, with respect to the third hypothesis, we found that 4 from the 9 studied studying strategies included, significantly differed between sexes, and 6 studying strategies showed different correlations with GPA and intelligence in each sex, therefore, this hypothesis is partially accepted.

Sex differences in intelligence and studying strategies and their correlation with GPA

In the present study we observed that although most sociodemographic variables were similar between sexes, there were many statistical differences in intellectual, academic and psychological variables. With respect to the intellectual variables, we observed that men´s sample showed slightly higher intelligence scores (including verbal, non-verbal and combined intelligence) than women´s sample, with a small effect size, and these differences were more evident in non-verbal intelligence. These results coincide with previous reports showing that males present higher scores in mathematics or abstract intelligence than females [29, 30]; however, the results do not coincide with these same reports showing that females outperform males in verbal abilities (including reading comprehension, perceptual speed and associative memory) [29, 30]. It has also been shown that these sex differences present age and cultural variations [31, 32], with female sex showing higher mathematics scores in early childhood [32] and Latino´s populations [31], which suggests an influence of biological and cultural factors in these differences. With respect to our results, it is possible that due to the small sample size studied, the reported advantage in verbal abilities in other studies, were not observed in the studied women´s sample. Another contributing factor to this observation, is that in this study verbal intelligence was measured with a scale mainly focused in vocabulary abilities (Shipley-2) and not in other verbal abilities, as reading comprehension, perceptual speed and associative memory, where women have better performed than men [29]. However, despite these low although significant differences in intelligence scores, there were no significant differences in GPA between sexes, these observations can be explained by the fact that non-verbal intelligence showed a very low correlation with GPA in both sexes (Table 2), while verbal intelligence showed a higher, but still low, correlation with GPA in both sexes. In addition, women´s sample showed a higher correlation between all intelligence scores (mainly for no-verbal and global intelligence) with GPA than men´s sample. These results coincide with those of our previous report, where non-verbal intelligence (measured with the BETA-4 intelligence test), showed a higher correlation with GPA in women´s sample than in men´s sample [14], which suggests that women effort more than men in order to obtain a higher GPA. This explanation is supported by the higher values of the studying strategies: task value and organization in women than in men (Table 1), which showed low but significant correlations with GPA in both sexes (Table 2). The studying strategy rehearsal also showed higher values in women than in men and significantly correlated with GPA only in women´s sample, while the only studying strategy with higher values in men was critical thinking which correlated with GPA only in men´s sample (Tables 1 and 2). All these observations suggest that women show higher GPA scores than men, although non-significant in this study for GPA but for preparatory GPA, as well as previously reported [8], mainly due to the higher use of studying strategies positively correlated with GPA when compared with men. Additional studying strategies showed positive correlations with GPA in both sexes: elaboration, self-regulation, effort regulation and time and study environment (Table 2), although these variables did not show sex differences, they showed low to moderate positive significant correlations with GPA, with variations in each sex. In this sense, men showed slightly higher correlations between GPA and these studying strategies than women. Interestingly, the observed correlations between all studying strategies with GPA in men overcame the correlation between all intelligence scores and GPA in this sex, while verbal and combined intelligence scores showed higher correlations with GPA in women than all the studying strategies correlated with GPA in this sex. These observations suggest that specific studying strategies (elaboration, time and study environment and critical thinking) have a greater influence in GPA in men than all the intelligence scores; while verbal and combined intelligence scores have a greater influence in GPA in women than all the studying strategies most correlated with GPA in this sex. The sex differences observed between the studying strategies used and their correlation with GPA can be compared with a previous report where differences in motivational profiles were found in male and female medical students, where females showed a higher frequency of the motivational profile correlated with higher GPA scores [8]. Nevertheless, studies addressing these specific studying strategies and their correlation with GPA in each sex were not found.

Psychological variables correlated with GPA

With respect to other psychological variables correlated with GPA, we found that the psychological variables positive relations with others and active coping showed a low positive correlation with GPA in both sexes and academic procrastination showed a low negative correlation with GPA in both sexes (Table 2). This last finding coincide with a previous report showing that different styles of academic procrastination correlate with lower GPA when compared with non-academic procrastination [33], although this previous report did not perform sex-comparisons, in the present study we observed that the negative correlation between academic procrastination and GPA, although low in both sexes, it was slightly higher in men than in women, which added to the fact that men´s sample showed higher scores of academic procrastination than women´s sample (with a borderline p value), also contributes to the explanation of lower GPA scores in men than in women. The positive correlation between GPA and active coping coincides with a previous report showing a low positive correlation between these two variables [4], and the positive correlation between GPA and positive relation with others, coincides. with our previous report performed in students with low academic performance, where we also observed a tendency towards a positive correlation between positive relations with others and GPA in women [14], however, in the present study that correlation was low but significant in both sexes (Table 2), which suggests that emotional support positively contributes to GPA.

Multivariate analysis for GPA

In the multivariate analyses of the three samples studied (Global, women´s and men´s samples), the most correlated variable with GPA was preparatory GPA (Tables 3, 4 and 5), which coincides with the bivariate analyses (Table 2), where preparatory GPA was the most correlated variable with GPA in the three samples. This observation coincides with the positive significant correlations of most studying strategies with GPA in the women´s and men´s samples, this by considering that preparatory GPA is also correlated with these studying strategies which may have been acquired in the early childhood and may be also related with personality factors, which remains constant along long periods of time.

We also observed other variables significantly correlated with GPA in each sex, including the combined intelligence scores in women´s sample, which was not correlated in men´s sample, also coinciding with bivariate analyses, where intelligence scores showed a higher correlation with GPA in women than in men. In addition, different studying strategies were correlated with GPA in each sex, where effort regulation and rehearsal correlated in women´s sample and time and study environment in men´s sample, which suggests that these specific strategies are the most correlated with GPA in each sex after adjusting by the rest of the variables and strategies. We also observed that additional personal factors correlated with GPA in the multivariate analysis, which included the negative correlation of the sum of diseases, monthly money and academic procrastination with GPA in women´s sample and the negative correlation of physical activity minutes and smoking consumption frequency with GPA in men´s sample. These findings coincide with the negative correlation between academic procrastination with GPA in the three samples in the bivariate analysis. With respect to physical activity, the results also coincide with our previous report [14], where physical activity negatively correlated with GPA only in men´s sample, a finding that could be related with a possible negative correlation between physical activity minutes and the time spent in studying and/or other studying strategies in this sex, however further studies and explanations should be explored. However, the R of the three multivariate models were moderate, which suggests that many personal, psychological and other unmeasured factors also contribute with GPA in each sex.

Sex differences in health and psychological variables and their relationship with intelligence

On the other hand, we also observed sex differences in psychological and health-related variables, where women showed higher levels of the sum of diseases, somatization, academic stress, anxiety and depressive symptoms as well as lower levels of emotion perception and emotion management (Table 1). These results coincide with previous reports of the research group, where higher levels of somatization, sum of diseases, stress, depression and anxiety, as well as lower levels of sleep quality were found in women when compared with men [34], and where lower levels of emotional intelligence were found in women than in men [35]. In this report we also found that academic stress was significantly higher in women than in men, which could be related with the higher frequency of anxiety in this sex. With respect to the relationship between these variables with intelligence, we found low positive correlations between the combined intelligence and the sum of diseases, somatization, depression and anxiety only in women´s sample, in addition, sleep quality showed a negative correlation with combined intelligence in this sex. These correlations coincide with our previous report [14], where significant positive correlations were found between the sum of diseases, somatization and anxiety with non-verbal intelligence and tendencies towards a positive correlation were found for academic stress and depression with non-verbal intelligence only in women [14]. Although the correlations were lower in this study, which can be due to the larger sample size and to the different test used to measure intelligence in the present study, this relationship can be explained by the higher values of these variables (sum of diseases, somatization, anxiety, depression and academic stress) in women than men, and to the lower values of sleep quality in women than in men. In addition, these correlations can be explained by the proposed intellectual overexcitabilities in people with high intelligence scores [36], which could lead to an increased stress perception and disease development, which could be more evident in women.

Sociodemographic and psychological variables correlated with intelligence

Other sociodemographic variables correlated with the combined intelligence were the paternal and maternal schooling, which showed a low positive correlation in both sexes, which coincides with previous reports [11] although in this case both parental schooling correlations with intelligence were similar, different to the previously reported higher correlation between intelligence and maternal schooling [11]. We also found that daily free hours showed a low negative correlation with intelligence in women, while having a job and daily minutes of physical activity showed a low negative correlation with intelligence in men (Table 2), correlations that need to be further explored in order to corroborate them. Finally, age also showed a low positive correlation with intelligence in the global and women´s sample, coinciding with our previous report [14], and which suggests that intelligence tends to decline in older stages.

The psychological variables positively correlated with intelligence were: academic procrastination and positive relations with others in women and active coping and emotion perception in men. All these were low but significant correlations. These findings coincide with our previous report showing positive correlations between non-verbal intelligence and emotion perception, emotion regulation and positive relation with others in both sexes [14], findings that suggest that verbal and non-verbal intelligences are correlated with emotional intelligence although with a low strength. The positive correlation between active coping and intelligence in men suggests that a higher intelligence contributes with active coping, although no similar reports were found in order to perform comparisons. Finally, the positive correlation between intelligence and academic procrastination can be due to the fact that students with higher intelligence have more confidence in their abilities and procrastinate more. This relationship could also explain the low negative correlation between combined intelligence and self-motivation in women. However, more studies are required in order to confirm these relationships, as no similar reports were found.

Multivariate analysis for intelligence

In the multivariate regression analyses for combined intelligence, we excluded the admission exam test by considering that this variable was the most correlated one with intelligence in both sexes because it also measures intellectual abilities. In this analysis, we observed that paternal schooling and academic procrastination where the variables most correlated with intelligence in the global and women´s samples, while academic procrastination was the variable most correlated with intelligence in men´s sample (Tables 6, 7 and 8). Although a positive correlation between intelligence and academic procrastination in the bivariate analysis was only observed in women´s sample, academic procrastination was positively correlated with intelligence in the multivariate analysis also for men´s sample, which as previously suggested may be a consequence of intelligence rather than a predictor of it. Additional variables significantly correlated with intelligence in the multivariate analyses were the positive correlations with the studying strategies time and study environment and task value in the global and women´s samples, respectively; and the negative correlations with smoking consumption frequency in the global sample and between the number of siblings and academic stress in the women´s sample. These results reveal that studying strategies are also correlated with intelligence after adjusting for confounding variables, which is mainly observed in women. In addition, the negative correlation between the number of siblings with intelligence in the global and women´s sample coincides with our previous report in students with low academic performance [14], where this variable negatively correlated with intelligence in all the samples. Nevertheless, the negative correlation between intelligence and smoking consumption frequency does not coincide with our previous report and deserves further exploration. Finally, the negative correlation between academic stress and intelligence in the multivariate analysis for women´s sample may also be a consequence of higher values of intelligence.

Study limitations

The main limitations of the study are the small sample size which diminishes the representativeness of the population and generalization of the results, and increases the random error. In addition, the lack of measurement of a more complete intelligence test including other intellectual abilities as memory or reading comprehension diminished the probability of finding new and interesting data including additional correlations and sex-differences in these abilities. Finally, the cross-sectional nature of the study impedes to perform causal relationships, therefore, there is a potential for a time-reversal relationship between the variables.

Conclusions

In conclusion, this study evaluated a comprehensive number of variables, including intelligence and studying strategies, and its relation with GPA and intelligence in health sciences university students and performed sex-separated analyses. We found that despite the slightly higher intelligence scores in men´s than in women´s sample, no differences in GPA where observed between sexes, which can be explained by the higher correlation between the combined intelligence and GPA in women than in men, and by the higher values of studying strategies positively correlated with GPA in women than in men. Additionally, women´s sample showed positive correlations between intelligence and the studying strategy “task value” in the multivariate analysis for intelligence that was not shown in men´s sample. With respect to the psychological and health variables we observed that women showed higher values of anxiety, depression, academic stress, somatization and the number of diseases than men, and all of which were low but significantly correlated with intelligence in this sex. Globally these results emphasize sex-differences in intellectual abilities and studying strategies that showed a differential correlation with GPA and intelligence in each sex.