Longitudinal Decline in Cardiorespiratory Fitness With Age Among Male Firefighters in San Diego, California, 2005–2015

Abstract

Objectives. To determine age-associated changes in cardiorespiratory fitness (CRF) among firefighters.

Methods. Male firefighters (n = 1169) underwent annual evaluations from 2005 to 2015 in San Diego, California. We assessed CRF, the ability to deliver and use oxygen, during a maximal graded exercise test and recorded it in metabolic equivalents (METs; 1 MET = 3.5 ml/kg/min). We investigated differences in baseline CRF among 10-year age groups and patterns of change over follow-up using cross-sectional and longitudinal analyses, respectively.

Results. Both analyses support an inverse relationship between CRF and age. Mean baseline CRF decreased from 15.7 ±1.7 to 11.7 ±1.9 METs in the younger than 30 years and 50 years and older age groups, respectively. There was no interaction between body fat and age in predicting CRF (P_{age × body fat} = .09). Longitudinally, there was a nonlinear decrease in CRF, which differed across age groups: the youngest age group demonstrated the greatest decrease in CRF.

Conclusions. Results support a negative, nonlinear association between age and CRF without modification by body fatness.

Public Health Implications. Fire departments should recognize the health and safety risks of declining CRF and institutionalize programs to promote firefighter fitness.

Firefighters require high levels of physical fitness to meet the demands of a strenuous and hazardous occupation.^1–3 Cardiorespiratory fitness (CRF) is directly related to firefighter job performance² and is important for promoting safety and health.^4–9 Consequently, the National Fire Protection Association (NFPA) has adopted a minimum CRF of 42 milliliters per kilogram per minute as measured by maximal oxygen uptake (VO_2max; equivalent to 12 metabolic equivalents [METs]).¹⁰ Poplin et al.⁴ found that CRF for the average-aged firefighter (∼39 years) exceeded the NFPA requirement (mean VO_2max = 49.6 ml/kg/min) but did not take into consideration the expected decline in CRF with age. According to the Fitness Registry and the Importance of Exercise National Database (FRIEND), VO_2max was 48.0 milliliters per kilogram per minute for those aged 20 to 29 years and 28.3 milliliters per kilogram per minute for those aged 60 to 69 years; the average decrease was 10% per decade.¹¹ The decline in CRF with age is an important concern, as the energy demands and strenuous nature of firefighter work do not decrease with age. Several cross-sectional studies of CRF among male firefighters have reported age group mean CRF values below the NFPA threshold in those older than 40 years.^12–15 This age-related decline in CRF, however, may be attenuated by maintaining a body mass index (BMI) of less than 30 kilograms per meter squared and performing 150 minutes of combined “cardio” and weight training per week.¹² These results suggest that body composition may moderate the relationship between age and CRF.

To our knowledge, there have been no longitudinal analyses documenting change in CRF over time in firefighters. Our primary aim was to determine the rate and pattern of change in CRF with age in male firefighters with data collected across 11 years using cross-sectional (for baseline data) and longitudinal analyses. We also explored the modifying effect of body composition, specifically body fat, on the relationship between age and CRF. On the basis of previous literature, we hypothesized a negative association between age and CRF with modification by body fatness.

METHODS

Participants were currently employed, uniformed firefighters of the San Diego Rescue Fire Department who underwent annual medical and performance evaluations as participants in the San Diego Firefighters’ Regional Fire Wellness Program, a clinical evaluation program for firefighters between 2005 and 2015. All firefighters were eligible to participate, and there were no incentives for participation. This was an open cohort study, as firefighters could enter and leave the program at any time between 2005 and 2015. Firefighters hired after 2005 were also eligible to participate at any time. We excluded female firefighters from this analysis because they represented less than 10% of the study population (n = 96); additionally, the findings from a cross-sectional analysis of CRF in this group are described in a separate report.¹⁶ We also excluded participants without baseline CRF or body fat data (n = 19). Compared with the 1150 included participants, those excluded were slightly older (mean_excluded = 43.4 ±2.4 years; mean_included = 38.1 ±0.3 years; P = .03) but did not differ by BMI (mean_excluded = 28.9 ±2.8 kg/m²; mean_included = 27.7 ±3.8 kg/m²; P = .19).

Measures

We measured variables of interest during annual exams at the San Diego Sports Medicine and Family Health Center’s Regional Fire Wellness Center. The wellness officer scheduled appointments according to firefighter battalion and station each year. Trained exercise physiologists employed by the center obtained the measurements.

Anthropometrics.

We measured height and weight to the nearest half centimeter and 0.1 kilograms, respectively, using a stadiometer and scale. We calculated BMI as weight in kilograms divided by height in meters squared. We estimated body fatness from the sum of 7 skinfold sites using the Jackson–Pollock equation.¹⁷

Cardiorespiratory fitness.

CRF is the capacity of the respiratory, cardiovascular, and musculoskeletal systems to deliver and use oxygen during sustained dynamic exercise. The gold standard measure of CRF is maximal oxygen uptake (VO_2max), which we estimated in this study from the final stage of a maximal graded exercise test on either a motorized treadmill or a stationary cycle. Patients selected either treadmill or cycle on the basis of the mode they use during recreational activities. Approximately 10% chose cycle ergometer (5% competitive cyclists, 2%–3% injury/joint pain precluding walking/running, ∼2% dislike for treadmill). We estimated VO_2max from the final work rate during the graded exercise tests using well-established metabolic equations on the basis of the linear relationship between work rate and VO₂^18–20 and expressed it as metabolic equivalents (METs; 1 MET = VO_2max [ml/kg/min]/3.5 ml/kg/min).

Statistical Analyses

Cross-sectional analysis of baseline data.

We categorized firefighters into 4 age groups on the basis of age at initial visit: younger than 30 years, 30 to 39 years, 40 to 49 years, and 50 years and older. We chose 10-year age groups consistent with previously published studies to allow comparison with reference norms.¹¹ We computed descriptive statistics using baseline data for variables of interest (height, weight, BMI, body fat percentage, CRF, percentage not meeting CRF guideline). Analysis of variance (ANOVA) compared mean values for the variables of interest among age groups. Multivariate linear regression examined the association of CRF and age group while controlling for body fat and ethnicity (White, Latino, African American, Asian, other). We entered body fat as a categorical variable using quartiles. We chose these covariates a priori on the basis of previous literature suggesting associations between CRF, body composition, and ethnicity.^21–23 We added an age times body fat interaction term to the model to assess the moderating effect of body fat in the relationship between age and CRF. We considered P values of .05 significant for all analyses. We conducted statistical analyses with Stata version 14.2 (StataCorp LP, College Station, TX).

Longitudinal analysis.

We fit generalized linear mixed models to examine the association between baseline age and changes in CRF and to compare patterns of change in CRF over time across baseline age groups. We estimated model coefficients using the restricted maximum likelihood method. This method uses all available information to provide estimates and SEs without excluding participants with missing values at any time point. We used an unstructured covariance matrix that allowed any valid pattern of variances and covariances for the within-participant correlation between repeated CRF measures. The model included baseline age group, baseline body fat (entered as a continuous variable), ethnicity, visit year, and a baseline age group times visit year interaction term. We met assumptions of fixed effect modeling including independent observations for individual participants and normal distribution of dependent variables. We conducted statistical analyses with SAS version 9.4 (SAS Institute, Cary, NC).

RESULTS

The average age at program entry was 38.1 ±0.3 years. Most firefighters self-identified as White (77.1%); the remaining were Latino (10.8%), African American (5.0%), Asian (3.7%), and other (3.4%). More than 95% of firefighters in the department participated each year for an average of 6.8 ±3.7 years. Of the 718 firefighters who entered in 2005, 292 (40.7%) participated for 11 years. Overall 64.8% of firefighters (745/1150) who entered the program at any point participated until study completion in 2015. Most firefighters who did not participate through 2015 retired; a small number (1–4 each year) went on military leave.

Cross-Sectional Analysis of Baseline Data

Table 1 shows mean body composition and CRF for each age group at visit 1. ANOVA revealed significant differences among mean values for all variables at P < .05, except for height. Mean body fat increased from 14.5% to 24.3%, mean CRF decreased from 15.7 to 11.7 METs, and the percentage of firefighters not meeting the 12-MET guideline increased from 1.9% to 56.1% in the younger than 30 years and 50 years and older age groups, respectively. The largest decrease of 1.7 METs occurred between the 30 to 39 years and 40 to 49 years groups.

TABLE 1—

San Diego Firefighters’ Mean Body Composition and Cardiorespiratory Fitness at Initial Visit, by Age Group: San Diego, CA, 2005–2015

	Age Group, Years
Variable	< 30	30–39	40–49	≥ 50	Pa
No.	320	345	273	212
Height, cm, mean ±SD	179.19 ±6.42	178.97 ±6.74	178.89 ±7.08	177.79 ±6.18	.10
Weight, kg, mean ±SD	84.19 ±10.70	87.65 ±12.69	92.26 ±14.97	91.13 ±13.89	< .001b
BMI, kg/m2, mean ±SD	26.21 ±2.92	27.34 ±3.40	28.82 ±4.34	28.81 ±3.98	< .001c
Body fat, %, mean ±SD	14.50 ±5.17	17.96 ±6.46	22.29 ±6.42	24.27 ±6.18	< .001d
Cardiorespiratory fitness (METs), mean ±SD	15.70 ±1.65	14.60 ±2.01	12.91 ±2.11	11.67 ±1.93	< .001d
Cardiorespiratory fitness < 12 METs, %	1.87	10.72	28.57	56.13	< .001d

Note. BMI = body mass index; MET = metabolic equivalent.

^aP values for analysis of variance comparing means across age groups.

^bP < .05 for all comparisons except 40–49 y vs ≥ 50 y (P = .78).

^cP < .01 for all comparisons except 40–49 y vs ≥ 50 y (P > .99).

^dP < .01 for all comparisons.

After controlling for body fat and ethnicity in the regression model, CRF still significantly decreased with increasing age group (Table 2). Body fat significantly predicted CRF: for any particular age group and ethnicity, firefighters with body fat percentages in the upper quartile had lower CRF values than did firefighters in the lower quartile (b = −3.43; 95% confidence interval [CI] = −3.73, −3.12). However, there was no interaction between age and body fat (entered as continuous variables) in predicting CRF (b = 0.1; 95% CI = 0.0, 0.0). Lastly, for any particular age group and body fatness, African Americans and Asians had lower CRF values than did Whites (Africans Americans: b = –0.9; 95% CI = −1.3, −0.5; Asians: b = –0.8; 95% CI = −1.3, –0.3).

TABLE 2—

Association Between Cardiorespiratory Fitness and Age Group at Visit 1 for San Diego Firefighters, Controlling for Body Fat and Ethnicity: San Diego, CA, 2005–2015

		METs
Variable	No.	Unadjusted Mean ±SD	Adjusted Differences, 95% CI
Age group, y
< 30	320	15.70 ±1.65	0 (Ref)
30–39	345	14.60 ±2.01	−0.47 (−0.72, −0.22)
40–49	273	12.91 ±2.11	−1.37 (−1.66, −1.08)
≥ 50	212	11.67 ±1.93	−2.32 (−2.64, −2.00)
Body fat quartile, %
4.5–13.9	291	16.14 ±1.65	0 (Ref)
14.0–18.6	284	14.68 ±1.81	−1.17 (−1.43, −0.90)
18.7–24.0	298	13.34 ±1.78	−1.98 (−2.27, −1.70)
24.1–39.4	277	11.61 ±1.86	−3.43 (−3.73, −3.12)
Race/ethnicity
White	853	14.15 ±2.44	0 (Ref)
Latino	119	13.75 ±2.19	−0.09 (−0.39, 0.22)
African American	55	12.63 ±1.98	−0.91 (−1.34, −0.48)
Asian	41	13.60 ±2.02	−0.75 (−1.25, −0.26)
Other	38	13.94 ±2.34	−0.49 (−1.00, 0.03)

Note. CI = confidence interval; MET = metabolic equivalent.

Longitudinal Analysis

Table 3 shows the number of participants per visit year by age group: 735 (64%) firefighters had 5 consecutive years of data, whereas 276 (24%) had 11 years of data. Only 2 firefighters in the 50 years or older group had CRF and body fat data for visit year 11, and we therefore excluded them from the analysis.

TABLE 3—

Number of Participants per Visit Year for San Diego Firefighters, by Age Group: San Diego, CA, 2005–2015

	Visit No.
Age Group	1	2	3	4	5	6	7	8	9	10	11
< 30 y	320	269	241	214	210	203	185	165	135	97	65
30–39 y	345	314	287	245	236	238	221	206	187	168	127
40–49 y	273	256	238	237	218	210	192	175	158	126	82
≥ 50 y	212	167	131	95	71	31	23	12	10	7	2
Total	1150	1006	897	791	735	682	621	558	490	398	276

Mean body fat, mean CRF, and percentage of firefighters not meeting the 12-MET guideline for each visit by age group are presented in Tables A, B, and C (available as a supplement to the online version of this article at http://www.ajph.org), respectively. Mean body fat increased from a minimum of 14.5% in the younger than 30 years age group to a maximum of 23.6% in the 50 years and older age group, whereas mean CRF decreased from a maximum of 15.7 METs in the younger than 30 years age group to a minimum of 11.7 METs in the 50 years and older age group. The percentage of firefighters not meeting the 12-MET guideline increased with age (minimum 1.9% for those younger than 30 years; maximum 58.1% for those 50 years and older). Of those 50 years and older, 34.8% to 58.1% did not meet the 12-MET guideline.

The mixed model showed that the pattern of change in CRF over 11 year follow-up was nonlinear and differed across baseline age groups (P_{age group × visit year} = .01; Figure 1). The magnitude of the 11-year decrease in mean CRF from visit 1 to 11 decreased with increasing age group: the 11-year decrease in CRF was 1.3, 1.2, and 0.4 METs for the younger than 30, 30 to 39, and 40 to 49 years groups, respectively. Mean CRF increased by 0.9 METs for firefighters 50 years and older, although they had the lowest CRF. This group also demonstrated a 1.3% decrease in mean body fat over 10 years. For the 30 to 39 years, 40 to 49 years, and 50 years and older year groups, mean CRF tended to increase from visit 1 to 3, whereas it declined for the younger than 30 years group.

FIGURE 1—

San Diego Firefighters’ Cardiorespiratory Fitness Change Over Visit Year, by Age Group: San Diego, CA, 2005–2015

Note. MET = metabolic equivalent.

For all age groups, the greatest decrease in CRF occurred between visit years 5 and 6, ranging from a decrease of 0.4 METs in the younger than 30 years and 50 years and older groups to 0.6 METs in the 40 to 49 years group. CRF was relatively stable from visit 7 to 11 for all groups. The interaction between age group and visit year in predicting CRF remained significant after sequential adjustment for baseline body fat and ethnicity (P = .05 after adjustment for body fat; P = .03 after adjustment for body fat and ethnicity). However, the interaction was no longer significant after adjusting for baseline CRF (P = .17).

DISCUSSION

As expected, CRF significantly decreased with increasing age in our sample of firefighters, with the largest decline occurring between the 30 to 39 and 40 to 49 years groups. In addition, more than 50% of firefighters older than 50 years did not meet the 12-MET recommendation for CRF. Similar to the results of Baur et al,¹² our results support an inverse relationship between body fat (BMI reported in Baur study) and CRF. However, body fat did not moderate the relationship between age and CRF. Age may be directly related to declining CRF via reductions in and the maldistribution of cardiac output, as well as declines in the oxidative capacity of musculature²⁴ regardless of body fat. Lifestyle factors that may contribute to declining CRF with age include changes in physical activity, work-related responsibilities, life responsibilities, motivation to stay fit, and the accumulation of work-related injuries and exposures.

However, the CRF of our firefighters exceeded that previously reported: mean CRF in our sample fell below 12 METs only for firefighters 50 years and older (vs ∼40 years in previous studies^12–15). Furthermore, mean CRF for firefighters in our oldest age group corresponded to the 75th to 90th percentile of reference norms from the FRIEND database.¹¹ This comparatively higher CRF could be a function of the Wellness Program. It is also possible that firefighters in our sample were simply more fit and had lower body fatness throughout their lives: most of the firefighters in our study had body fat percentages corresponding to less than the 25th percentile of normative values.¹⁸ Still, our results support the negative association of body fat with CRF.^12,13,15

Longitudinal Analysis

Our longitudinal data also support a negative association between age and CRF because mean CRF was highest among the younger firefighters. In addition, the decline in CRF over 11 years differed among age groups and therefore provides evidence of a temporal relationship between age and CRF change. The 2 youngest groups (< 30, 30–39 years) experienced the largest overall decreases in CRF over time compared with the 2 oldest groups (40–49 years, 50 years and older), although this may partially be explained by higher baseline CRF values among younger firefighters. For the 50 years and older group, CRF increased by 0.9 METs over 10 years with a concomitant 1.3% decrease in body fatness. Regardless, up to 58% of firefighters 50 years and older did not meet the national 12-MET guideline. These results suggest that promoting fitness is important among all firefighters, both younger firefighters, who experienced the greatest rate of decline in CRF, and firefighters whose mean CRF was more likely to fall below the national threshold. Although older firefighters are more experienced and potentially more effective than are younger firefighters and are more likely to transition into more sedentary roles at work, maintaining fitness is still important for promoting cardiovascular health and firefighter safety, particularly because sudden cardiac death is the leading cause of on-duty firefighter mortality.⁹ Fire departments should be aware of the health risks associated with low CRF and consider instituting programs to encourage fitness among all firefighters.

The longitudinal analysis also revealed a nonlinear decrease in CRF across the 11-year follow-up. The greatest decline for all age groups occurred between the fifth and sixth visit years and was followed by a period of relative stability in CRF. This pattern of decline may be related to economic factors in the city that affected the hiring of new firefighters and created mandatory overtime. The excessive workload may have contributed to decreasing motivation or time to maintain fitness. Another possible factor was a change in the type of on-duty physical activities in which firefighters engaged. This finding may have important implications for future interventions targeted at improving CRF in firefighters because additional measures may be needed to promote fitness when resources are limited and demand is high.

Strengths and Limitations

This study was strengthened by a large sample size and 11 years of data from firefighters whose age and ethnic distribution were representative of the US firefighter population.²⁵ Therefore, our results are overall generalizable to US firefighters, although they may differ on the basis of setting (urban vs rural). In addition, participation and retention rates were relatively high. To our knowledge, this is also the first study to describe changes in CRF in male firefighters longitudinally, allowing better characterization of patterns of change over time while controlling for within-participant variation. Finally, we assessed CRF via maximal graded exercise tests rather than estimating it from submaximal or field tests.

There were several limitations to this study. First, because retired firefighters were not eligible to participate, sample size was small (range = 2–12) for firefighters in the 50 years and older group at visits 8 through 11. Second, measurement personnel who conducted physiological testing were not blinded to the purpose of the program because they were regular employees of the clinic assigned to the Fire Wellness Program. However, employees did not receive incentives for participants doing well and had no contact with participants outside the clinic. Another limitation included the use of skinfold measurements to estimate body fat, particularly without determining interrater reliability. To minimize interrater variability, technicians were trained by the same knowledgeable and experienced individual. We were also unable to control for several variables that may moderate the relationship between CRF and age, including physical activity, socioeconomic status, job classification, toxin exposure, call volume, and major life events. Lastly, this study excluded female firefighters. The results of a cross-sectional analysis of CRF in female firefighters from the same fire department are described in a separate report.¹⁶ Briefly, CRF also decreased with age in female firefighters; however, the absolute decrease from the youngest to oldest group was smaller for women (14.0 to 12.8 METs) than for men (15.7 to 11.7 METs).¹⁶

Public Health Implications

As expected, our results support a negative association between age and CRF. Younger firefighters had higher levels of fitness but also demonstrated the greatest rate of decline over 11 years. In addition, although our firefighters were overall more fit than were those in previous studies and when compared with age-matched reference data, up to 58% of firefighters 50 years and older did not meet the 12-MET national guideline. Therefore, fire departments should recognize the health and safety risks associated with decreasing CRF and institutionalize programs to promote and sustain fitness among firefighters of all ages throughout their careers.