Method

Participants

A non-experimental, cross-sectional, instrumental design (^{Carretero-Dios & Pérez, 2005}) was used through non-probabilistic convenience sampling. The participants were invited to participate in public areas or spaces, as well as public and private universities in the city and the state of Mexico. A total of 254 participants were included, most of them women, single, with undergraduate studies, and residents of Mexico City (Table 1). Regarding the number of subjects needed in the evaluation and validation of psychometric instruments, the recommendation is that there be at least 10 participants for each item to be tested. In our study, the sample was 254, which exceeds that suggested by experts (^{Nunnally & Berstein, 1994}).

Table 1: Sociodemographic and clinical data from a sample of 254 participants 

^Note: The sum of the FACT-GP is expressed as medians (25th-75th percentile).

Compliance with ethical standards

The procedures of this research complied with the provisions of the Declaration of Helsinki regarding research in humans. Each participant agreed to participate by understanding and signing the informed consent.

Participation criteria

Participants were selected based on the following criteria. Inclusion: 1) know how to read and write, 2) Mexican nationality, 3) minimum age of 18 years. Exclusion: 1) with cognitive impairment, 2) visual and/or hearing impairment that prevents him from answering the instruments. Elimination: 1) that during or after filling out the instruments they decide not to continue participating.

Results

The reliability analysis of the FACT-GP obtained a global Cronbach's alpha of .81. The KMO value was 0.71 (p= .001), confirming that the sample used was suitable for this analysis. For the confirmatory analysis (see Figure 1), the following procedure was conducted (^{Byrne, 2010}):

Identification. The maximum likelihood method allowed to over-identify several parameters lower than the total number of non-redundant elements of the matrix, while no correlations were observed in the residuals, indicating a recursive model.

Parameter estimation: In the AMOS program, the maximum likelihood method was applied, with standardized estimators, R2 estimator (squared multiple correlations), covariances of the estimators, indices to be modified, and critical proportions for the differences.

Evaluation of the fit. The objective was that the estimated coefficients did not exceed the acceptable limits (excluding negative or non-significant error variances, standardized coefficients greater than 1, or excessively high standard errors related to any estimated coefficient). At this point, no collinearity was observed in the measured variables, since the correlations were less than 0.3, nor extreme scores, neither univariate nor multivariate. Additionally, excellent asymmetry was observed in all variables (values not greater than ± 1.00).

The parsimony and efficiency indices confirm an acceptable goodness-of-fit statistic (AGFI = 0.893), while the root mean square residual of approximation (RMSEA = 0.048, 95 % CI = 0.035-0.060), which penalizes the increase in the complexity of the model, indicates a very acceptable fit to the data as it is almost equal to or less than 0.08, reaffirmed by the value of the mean residual error (RMR = 0.051), which is clearly below the limit value (less than 0.08), indicating almost zero error and almost perfect fit, with practically zero difference between the observed and predicted covariance matrix, and with this, it is possible to assume almost zero discrepancy between the proposed model and the analyzed data. Finally, the Hoelter test N= 254 (p= .01) allows us to accept the hypothesis that the sample used is sufficient for the analysis as it theoretically does not exceed the number of real data.

Figure 1:  CFA model of 4 first-order factors for FACT-GP in the Mexican population (N = 254) 

Discussion

The purpose of this study was to determine the psychometric properties of the FACT-GP in the general Mexican population. The factorial structure identified is similar to those reported in the international literature (^{Janda et al., 2009}) with four components: Physical well-being, Emotional well-being, Family and Social well-being; Couple and Sexual Life similar to the original version.

The physical component denotes symptoms associated with conditions of persistent work activity or illness, mainly of a chronic degenerative type and limitations previously reported in other studies (^{Alanis et al., 2015}; ^{Janda et al., 2009}), as well as components of anxiety before the death, which has been identified in the general population (^{Magaña et al., 2018}), suggesting the need to identify this population group and refer to health care programs.

A factor that remains constant is the emotional component that represents mental health in different areas of a person's life, expressed in satisfaction and enjoyment. This is particularly relevant since positive mental health contributes instrumentally to having an adequate standard of living, including, but not limited to, well-being (^{Keller, 2020}). Mental health is an essential component associated with quality of life in different general and clinical populations, which emphasizes the postulate: that without mental health there is no health (^{Gálvez Olivares et al., 2020}; ^{Ruiz Martínez et al., 2018}).

The family and social environment factor, on the other hand, corroborates what was identified in the original version. In this context, the family is a basic learning component for people. As an institution, it is the source of the most lasting relationships and the first social sustenance of the person, with relationship guidelines that are maintained throughout life, which makes it the instance with the greatest resources to produce changes (^{Fishman, 1988}). The Mexican family has among its functions the care and satisfaction of the basic needs of its members, procreation, the development of potentialities of each of the members, the establishment of affective bonds, as well as providing socially accepted models (Ruiz- Cardenas et al., 2018). Family functioning intervenes in a positive family environment, in which communication, harmony, respect, and demonstrations of affection are intrinsic elements in relationships, favoring an adequate quality of life (^{García-Méndez et al., 2006}).

Finally, the partner and sexual life component can be explained by the fact that the majority of the population is young and may be characterized by having an emphasis on the partner associated with the stage of life they are in. Although in young Mexicans differences by sex are identified in the age of the beginning of their sexual life, in the characteristics of the couple, and in the meanings given at the beginning of their sexual life (^{Menkes-Bancet et al., 2019}), the range of age of the studied sample can explain this component. Although most of the participants were young and healthy, an increase in morbidity in the Mexican population is expected.

Developers and researchers of cultural adaptation of scales must ensure that the adaptation process fully takes into account the linguistic and cultural differences between the populations for which the adapted versions of the instrument are intended (^{Prakash et al., 2019}). The real value of existing and potential new measures is determined by the trade-off between the quality and quantity of information that individualized measures offer, and the time, energy, and costs involved in collecting this information (^{Dijkers, 2003}), even more so in Latin America, where the evaluation of the quality of life is a relevant issue to identify the well-being conditions of the population.

HRQoL is a significant measure in the evaluation of living conditions, especially in low- and middle-income countries such as Mexico, because it provides valuable information on people's satisfaction and their living conditions concerning accessibility, organization, and the living conditions of a society. HRQoL is influenced by factors such as employment, housing, access to public services, communications, urbanization, crime, environmental pollution, and others that make up the social environment and influence the human development of a community (^{Velarde-Jurado & Avila- Figueroa, 2002}). The relevance of the findings of this study proposes the use of a valid and reliable tool for its measurement.

The FACT-GP, being an easy-to-apply, brief and reliable instrument for clinical practice and research, is relevant for determining the quality-of-life levels. The internal consistency and explained variance identified in the version in the Mexican population corroborates the psychometric background. The indices from the confirmatory analysis indicate that the proposed model is parsimonious and balanced in its adjustment to the empirical data used, with an error close to zero to estimate CV in the general population.

Among the limitations and future perspectives of this work is the lack of concurrent validity with other psychological and clinical variables that would be correlated. And additionally, establishing the stability over time of the FACT-GP scale is an objective to be determined. Finally, another aspect to consider is the possible bias due to the type of population of productive age, with a low percentage of comorbidity and with a productive occupation that, although it is similar to other studies that use the FACT-GP, it is recommended to have a larger sample. heterogeneous in future studies.

Conclusion

The FACT-GP can be used in the general Mexican population in procedures for diagnostic, clinical, and research care. Being an easy-to-apply, brief and reliable instrument, it is relevant to determine the magnitude of quality of life in different social circumstances. The confirmatory analysis indicates that the instrument is parsimonious, with a structure similar to the original and balanced version to measure QoL in the general Mexican population.