Effects of sex, age, and visits on receipt of preventive healthcare services: a secondary analysis of national data

The two overarching goals of Healthy People 2010 are to increase quality and healthy years of life and eliminate health disparities. Recognizing that the health disparity between men and women is not simply biological, the agenda of Healthy People 2010 calls for greater attention to and scientific exploration of the health differences between men and women [1].

Several studies have identified factors associated with disparities in the use and/or receipt of preventive services [2‐20]. Many of these studies have examined the influence of enabling resources (e.g., insurance, usual source of care) on preventive services use [4, 8, 10, 11]. Others have focused on differences in receipt of preventive services by characteristics such as race/ethnicity or by community (rural vs urban) [5, 6, 4, 15, 17, 18, 20]. Few studies have addressed the disparity in receipt of preventive services between men and women.

Like sex, age influences the receipt of preventive services, with younger individuals less likely to receive services [21]. In the United States in 2002, women 25 to 44 years of age accounted for 17 percent of all office-based visits to healthcare providers [22]. Visits for routine Pap smears and other reproductive needs, including pregnancy, provide multiple opportunities for health professionals to incorporate other preventive services as deemed appropriate. Younger to middle-aged men lack similar impetus to visit healthcare providers. We therefore hypothesized that sex and age would exert a combined effect on receipt of preventive services and that this combined effect would be accounted for by the difference between men and women in their number of ambulatory visits.

In the year 2000, approximately 48 percent of the nation's adult population was men, almost two-thirds of whom were between the ages of 20 and 49 (Table 1). Men were more likely to be uninsured and lack a usual source of care. Men were also slightly more likely to be college graduates and report their health as excellent.

Table 2 shows unadjusted and adjusted odds ratios (and 95% confidence intervals) for the receipt of the selected preventive services by men compared to women. Men were significantly less likely than women to receive each of the selected services with the exception of colorectal cancer screening. Table 2 also shows that the adjusted odds ratios differed little compared to the unadjusted odds ratios, suggesting minimal confounding due to the covariates.

Figures 1, 2, 3, 4, 5 are graphs of unadjusted weighted percentages of men and women stratified by age group who received the selected preventive services. Unadjusted percentages are shown because given the large nationally representative sample used for this study, these measures provide the closest estimates of actual receipt of services. There was no significant difference in receipt of colorectal cancer screening when stratified by age group. All other differences between men and women in receipt of the preventive services within each age stratum were statistically significant with the exception of cholesterol checks in those 60 years and older and smoking cessation counseling in those 60 years and older. For blood pressure checks, the difference between men and women was most pronounced for 20–29 year olds (65% vs. 85%, respectively), whereas this discrepancy narrowed for older ages (87% vs. 91%, for age 60+). Similarly, the difference in receipt of checkups was most pronounced for the youngest group and narrowed for older ages (52% vs. 75% for 20–29 year olds and 86% vs. 89% by age 60+). The graphs for receipt of cholesterol screening and smoking cessation counseling appear to follow a similar pattern.

Table 3 shows adjusted odds ratios for receipt of the selected preventive services by men compared to women stratified by age groups. Although the graphs in the figures suggest interaction between sex and age for all services with the exception of colorectal cancer screening, in adjusted models, there was statistically significant interaction only for blood pressure checks and checkups. For blood pressure checks, the difference between men and women was most pronounced for 20–29 year olds (OR = .38), whereas this discrepancy narrowed for older ages (OR = .68 for age 60+). There was a similar pattern for checkup within two years (OR = .42 for 20–29 and 30–39 year olds vs. OR = .72 for age 60+).

Number of ambulatory care visits in the previous year was added to the original models and interaction between sex and age was again tested. The interaction between sex and age for blood pressure checks was mitigated (p = .206) but overall, men still had lower odds of receiving the service (OR = .58; 95% CI .51–.66). There was no effect on the sex-age interaction for checkups when number of visits was added to the model in this manner.

As shown in previous studies, [6, 17] this study confirms that, compared to women, men receive fewer blood pressure measurements and cholesterol measurements. By excluding men and women already diagnosed with hypertension in our analyses of blood pressure measurements and excluding those with hypercholesterolemia in our analyses of cholesterol measurement, we specifically have focused on screening for these health problems. In addition to these findings, we also found that compared to women, a lower percentage of men receive smoking cessation counseling and health checkups. The rates of smoking cessation counseling are notably low in both sexes. Also, as shown elsewhere, colon cancer screening rates are low in both sexes with men receiving colorectal cancer screening at a similar, or slightly higher percentage than women [27, 28]. This slight difference may be due to more fear and embarrassment by women about colon endoscopy [29].

The presence of interaction in the adjusted models between sex and age group for blood pressure checks and health checkups indicates that men in the younger age groups are even less likely to receive these services than expected solely by virtue of sex, even after accounting for potential confounders. One might expect the same effect for cholesterol screening, but in combination with adjustment for covariates, the interval of five years may mitigate the effect of interaction by age group. The generally low rates of reported receipt of smoking cessation counseling by both men and women of all ages may have contributed to the lack of interaction on this outcome in the adjusted models. The rates of colorectal cancer screening may not be as affected by age because such screening does not begin until age 50. Men and women 50 years and over may have more similar patterns of health service use than younger men and women. Future research could explore this hypothesis.

It appears that some of the discrepancy in receipt of preventive services between men and women younger than 50 results from the difference in the number of visits to healthcare providers. Previous work supports this finding. For example, one study demonstrated that the rate of visits for annual examinations and preventive services by women was 100% higher than for men [30]. Even after excluding pregnancy-related visits, women were 33% more likely than men to visit a physician, a difference that decreased with age [30]. However, visits should not be thought of as a confounder because, in the current predominant model of preventive healthcare, a visit is a necessary step along the causal pathway to receiving preventive services. In order to increase men's adherence to preventive care recommendations, future research should explicate the motivating factors behind men's decisions to seek outpatient care. Analytically, our findings underscore the importance of using a carefully considered conceptual model to guide analyses of care-seeking behavior. Future research in this area should also consider the use of structural equation modeling techniques to gain a more complete understanding of the complex underlying relationships between variables.

With the exception of colorectal cancer screening (for which both men and women have low rates), fewer men receive the preventive services considered in this analysis than women. This disparity is further affected by age group for certain services, such that younger men are particularly vulnerable. This effect may be partly explained by the difference between younger men and women in the number of ambulatory care visits.

Two potential strategies to improve preventive healthcare for young to middle-aged men emerge from this study. One strategy would be to increase visits by these men for ambulatory care. Presumably, such visits would be health maintenance visits. However, this strategy is not ideal because even if men could be persuaded to increase their visits, such a strategy would not only be costly and resource-consuming, but because of disparities in access to care, would also not likely reach those men most likely to benefit. A second strategy would be to develop alternative methods of health promotion and preventive services delivery to younger and middle-aged men. Such strategies might include worksite programs or community programs situated in settings where men live and gather, with appropriate mechanisms to insure contact with health professionals when health problems (e.g., hypertension) are identified.