INTRODUCTION
Differences in type of occupation may create disparities in health between sexes in Mexico, since, type of labor force activity is associated with physical and mental health risks in other countries [1–5]. This study makes use of the Mexican Health and Aging Survey (MHAS) to examine the relationship between labor force participation and gender differences in health the prevalence of arthritis, diabetes, and hypertension.
METHODS
Data
The Mexican Health and Aging Survey (MHAS) is a nationally representative panel survey of subjects and their spouses in Mexico in 2001 (N=15,186). For this analysis, cases were limited to sampled subjects and spouses where not included. Individual level sample weights created by the principal investigators of the data sets was used for the statistical analysis in order to adjust for sampling design and to represent the Mexican population 50 years of age and older [6].
Variable Measurement
The outcome variable for this project was self-reported physician-diagnosed arthritis, diabetes, and hypertension. We used three measures for our employment variables. First, years in the labor force was measured as a continuous variable. Second, the type of work that subjects reported doing for most of their lives was classified into seven categories based on general description of job classification. Finally, we accounted for pension eligibility as an indicator of whether subjects worked in the formal labor market. We adjusted for whether subjects migrated to the United States in their lifetime, years of education (continuous), age (continuous), and urban/rural residence.
Analysis
Descriptive statistics were generated to illustrate variations in employment participation and patterns in chronic disease by gender. Binomial logistic regression models were then estimated to predict the odds of having arthritis, diabetes, or hypertension by sex. A total of three models were run. The first model adjusted for demographic characteristics. In the second model, employment history was included, and in the final model interactions between gender and type of work, years worked, and pension eligibility were assessed.
RESULTS
Table 1 shows that older Mexican women were less likely to have previously migrated to the United States (3.5 versus 15.1 percent), worked fewer years (16.0 versus 41.4 years), and more likely to have never worked at all (30 percent). Because of these trends, it is not surprising to see that only 15.2 percent of older Mexican women are pension eligible compared to 45.7 percent of men.
Table 1.
Odds of reporting arthritis, diabetes, or hypertension in older Mexican women compared to men. (n = 7,792)
| Frequency Distributions | Arthritis | Diabetes | Hypertension | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Female | Male | Model 1 | Model 2 | Model 3 | Model 1 | Model 2 | Model 3 | Model 1 | Model 2 | Model 3 | |
| Arthritis (%) | 1,119 (25.9) | 535 (15.4) | |||||||||
| Diabetes (%) | 757 (17.2) | 443 (12.8) | |||||||||
| Hypertension (%) | 1,958 (45.3) | 963 (27.8) | |||||||||
| Gender | |||||||||||
| Women | 1.7*** | 2.0*** | 2.3* | 1.6** | 1.4* | 2.4† | 2.5*** | 2.6*** | 4.8*** | ||
| Men | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | ||
| Age (s.e.) | 62.4 (.14) | 61.5 (.16) | 1.0*** | 1.0*** | 1.0*** | 1.0 | 1.1 | 1.0 | 1.0** | 1.0** | 1.0* |
| Years of Education (s.e.) | 4.1 (.07) | 5.2 (.10) | .96* | .98 | .98 | 1.0 | .98 | .98 | 1.0 | 1.0 | 1.0 |
| Married (%) | |||||||||||
| Yes | 1,705 (39.5) | 2,380 (68.6) | 1.0 | 1.1 | 1.0 | .98 | .99 | .96 | 1.0 | 1.1 | 1.0 |
| No | 2,617 (60.6) | 1,090 (31.4) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| Migration to the US (%) | |||||||||||
| Former Migrant | 151 (3.5) | 523 (15.1) | 1.4* | 1.4* | 1.4† | 1.1 | 1.1 | 1.0 | 1.3† | 1.3† | 1.3 |
| Never Migrated | 4,171 (96.5) | 2,947 (84.9) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | |||
| Type of Work (%) | |||||||||||
| Sales/Vending | 436 (10.1) | 305 (8.8) | 1.1 | 1.0 | 2.1*** | 1.8† | 1.4† | 2.6*** | |||
| Professional | 708 (16.4) | 674 (19.4) | 1.0 | .85 | 1.7* | 2.1* | 1.2 | 1.7† | |||
| Services/Product Provider | 285 (6.6) | 488 (14.1) | 1.3 | .88 | 1.9** | 2.3** | 1.4* | 1.2 | |||
| In Client’s Home | 820 (19.0) | 418 (12.1) | 1.1 | .56† | 1.2 | .88 | 1.5** | 1.2 | |||
| Other | 326 (7.5) | 124 (3.6) | .86 | .93 | 2.0* | .76 | 1.7** | 1.4 | |||
| Did Not Work | 1,293 (29.9) | 23 (.66) | .68† | .37 | 1.8** | 1.6 | 1.3 | 1.4 | |||
| Physical Labor | 454 (10.5) | 1,438 (41.4) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | |||
| Years Worked (s.e.) | 16.0 (.27) | 41.4 (.22) | 1.0 | 1.0† | 1.0 | 1.0 | 1.0 | 1.0* | |||
| Receives Pension (yes=1) (%) | |||||||||||
| Yes | 658 (15.2) | 1,587 (45.7) | .65** | .75 | 1.3 | 1.6* | 1.1 | 1.3 | |||
| No | 3,664 (84.8) | 1,883 (54.3) | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | |||
| Significant Interaction Effects | |||||||||||
| Type of Work ( Physical Labor ) | |||||||||||
| Females | |||||||||||
| Sales/vending | 4.3 | ||||||||||
| Services/Product | 13.8 | ||||||||||
| In Client’s Home | 1.7 | ||||||||||
| Males | |||||||||||
| Sales/vending | .88 | ||||||||||
| Services/Product | 2.5 | ||||||||||
| In Client’s Home | 1.7 | ||||||||||
≤.10
≤ .05
≤ .01
≤ .001
Table 1 presents logistic regression results for the prevalence of arthritis, diabetes, hypertension, and obesity. In the first model for each condition, older Mexican women are at significantly greater odds of the chronic condition than men. When including the employment history variables in model 2, with the exception of diabetes, the odds ratios for female increases for all conditions. For example, for arthritis, when including work history, the odds ratio for female increases by 31.2 percent (model 1 OR = 1.726; model 2 OR = 2.038). Also, when including work history to model 2 for hypertension the odds increases by 11.3 percent (model 1 OR = 2.487; model 2 OR = 2.600). However, for diabetes, the odds ratio is reduced by 10.9 percent (model 1 OR = 1.554; model 2 OR = 1.445). These results, therefore, suggest that in the same type of working conditions, women are at greater risk for these chronic conditions than men.
To examine this relationship further, interactions between female and type of work, years worked and pension eligible are included in model 3. Significant interaction effects are observed for arthritis and hypertension. Individual equations were estimated to determine the effect of occupation type for men and women to have arthritis or hypertension. Estimates were obtained by adding the regression coefficient for sex (β1 sex) with the coefficient for the given occupation (β2 occupational class) with the interaction term (β3 sex*occupational class) [7]. The product was then exponentiated to determine the effect for each sex and these odds ratios are presented in the bottom of the table. These estimates show that women who worked in the services or product provider industry have odds of 13.8 of having arthritis, whereas men have odds of 2.5 relative to men who worked in physical labor occupations. However, for older Mexican women and men the odds ratio for having arthritis is the same (O.R. =1.7) for those working in the client’s home compared to men who worked in physical labor occupations. Finally, women who worked in the sales or vending professions have odds of 4.3 to self-reported hypertension, but, men are at a reduced risk of having hypertension (O.R. = .88) in relation to men who worked in physical labor occupations.
DISCUSSION
This study confirms findings from other studies that determined that Mexican women are more susceptible to chronic conditions than men [8]. Although type of labor force participation differs quite a bit between older Mexican men and women, findings from this study suggest that within the same occupational classification, women suffer from the damaging effects on health to a greater extent than men. Interaction effects show that women who provided services or products are particularly susceptible to arthritis compared to men. Moreover, women who work in sales were at a significantly greater risk of hypertension than men. Thus, although women in Mexico may have worked in the same occupation, they may have different responsibilities that have a greater lasting effect on health than for men.
The major limitation of this study is that since the analysis was based on cross-sectional data, we were only able to obtain a snap-shot of health in the later years of life. It may be that the effect of labor force participation on health fluctuates over the life course, and, based on this data, we would have no way of measuring those changes. Conversely, since the participants are 50 years old and over, we are able to determine the cumulative effects of labor force activity that would not be observable if subjects were younger. Despite the shortcomings, these findings demonstrate the need for public awareness campaigns targeted towards women to prevent chronic disease.
Footnotes
Publisher's Disclaimer: 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 citable 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.
Contributor Information
Jennifer J. Salinas, Sealy Center on Aging, University of Texas Medical Branch.
M. Kristen Peek, Preventive Medicine and Community Health, University of Texas Medical Branch.
References
- 1.Väänänen A, Toppinen-Tanner S, Kalimo R, Mutanen P, Vahtera J, Pieró JM. Job characteristics, physical and psychological symptoms, and social support as antecedents of sickness absence among men and women in the private industrial sector. Soc Sci Med. 2003;57:807–824. doi: 10.1016/s0277-9536(02)00450-1. [DOI] [PubMed] [Google Scholar]
- 2.Ala-Mursula L, Vahtera J, Kouvonen A, Väänänen A, Linna A, Pentti J, Kivimäki M. Long hours in paid and domestic work and subsequent absence: does control over daily working hours matter? Occup Environ Med. 2006;63:608–616. doi: 10.1136/oem.2005.023937. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Stallones L, Beseler C. Pesticide illness, farm practices, and neurological symptoms among farm residents in Colorado. Environ Res. 2002;90:89–97. doi: 10.1006/enrs.2002.4398. [DOI] [PubMed] [Google Scholar]
- 4.Verbrugge LM. Gender and health. J Health Soc Behav. 1985;26(3):156–182. [PubMed] [Google Scholar]
- 5.Evans O, Steptoe A. The contribution of gender-role orientation, work factors and home stressors to psychological well-being and sickness absence in male and female-dominated occupational groups. Soc Sci Med. 2002;54:481–492. doi: 10.1016/s0277-9536(01)00044-2. [DOI] [PubMed] [Google Scholar]
- 6.Aguilar-Salinas C, Velazquez-Monroy O, Gomez-Perez FJ, Gonzalez Chavez A, Lara Esqueda A, Molina Cuevas V, Rull-Rodrigo JA, Tapia Conyer R. Characteristics of patients with type 2 diabetes in Mexico: results from a large population-based nationwide survey. Diabetes Care. 2003;26(7):2021–2027. doi: 10.2337/diacare.26.7.2021. [DOI] [PubMed] [Google Scholar]
- 7.Agresti A. An Introduction to Categorical Data Analysis. New York: John Wiley & Sons, Inc.; [Google Scholar]
- 8.Wong R, Peláez M, Palloni A, Markides KS. Survey data for the study of aging in Latin America and the Caribbean: selected studies. J Aging Health. 2006;18(2):180–206. doi: 10.1177/0898264305285655. [DOI] [PubMed] [Google Scholar]