Associations Between Bone Mineral Density, Grip Strength, and Lead Body Burden in Older Men Naila Khalil, MBBS, MPH, PhD,* Kimberly A. Faulkner, PhD,† Susan L. Greenspan, MD,‡ and Jane A. Cauley, PhD,† for the Osteoporotic Fractures in Men Research Group
OBJECTIVES: To study the association between blood lead concentration (BPb) and bone mineral density (BMD), physical function, and cognitive function in noninstitutionalized community-dwelling older men. DESIGN: Cross-sectional study. SETTING: University of Pittsburgh clinic, Pittsburgh, Pennsylvania. PARTICIPANTS: Non-Hispanic Caucasian men aged 65 and older (N = 445) recruited as a subset of a prospective cohort for the Osteoporotic Fractures in Men Study. MEASUREMENTS: BPb was measured in 2007/08. From 2007 to 2009, BMD (g/cm2) was measured using dualenergy X-ray absorptiometry. At the same time, physical performance was measured using five tests: grip strength, leg extension power, walking speed, narrow-walk pace, and chair stands. Cognitive performance was assessed using the modified Mini-Mental State Examination and the Trail-Making Test Part B. Participants were categorized into quartiles of BPb. Multivariate regression analysis was used to evaluate the independent relationship between BPb, BMD, and cognitive and physical function. RESULTS: Mean BPb standard deviation was 2.25 1.20 lg/dL (median 2 lg/dL, range 1–10 lg/dL). In multivariate-adjusted models, men in higher BPb quartiles had lower BMD at femoral neck and total hip (P-trend < .001 for both). Men with higher BPb had lower age-adjusted score for grip strength (P-trend < .001), although this association was not significant in multivariate-adjusted models (P-trend < .15). BPb was not associated with lumbar spine BMD, cognition, leg extension power, walking speed, narrow-walk pace, or chair stands. From the *Center for Global Health, Department of Community Health, Boonshoft School of Medicine, Wright State University, Dayton, Ohio; † Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, and ‡Division of Geriatric Medicine, Department of Medicine, Medical School, University of Pittsburgh, Pittsburgh, Pennsylvania. Address correspondence to Dr. Naila Khalil, Center for Global Health, Department of Community Health, Boonshoft School of Medicine, Wright State University, 3123 Research Blvd, Suite #200, Dayton, OH 45430. E-mail: [email protected] DOI: 10.1111/jgs.12603
JAGS 62:141–146, 2014 © 2013, Copyright the Authors Journal compilation © 2013, The American Geriatrics Society
CONCLUSION: Environmental lead exposure may adversely affect bone health in older men. These findings support consideration of environmental exposure in ageassociated bone fragility. J Am Geriatr Soc 62:141–146, 2014.
Key words: lead; elderly; men; bone; grip strength; cognition; physical function
A
ge-related osteoporosis and decline in physical function are significant public health concerns because they are associated with falls,1,2 fractures,2,3 frailty,2 and morbidity.4,5 In the U.S. National Health and Nutrition Examination Survey (NHANES) data completed in 2005 to 2008, 9% of adults aged 50 and older had osteoporosis at the femoral neck or lumbar spine, and half had low bone mass at one of these two sites.6 It is estimated that one-third of all osteoporotic fractures and 25% of osteoporosis-related costs are observed in men,6 with higher associated mortality than in women.6 The healthcare burden will increase with increasing longevity in the population,6 so it is important to identify determinants of bone strength and physical function in older men, including environmental exposures such as lead. Lead is a toxic element highly prevalent in the environment. Lead exposure has been associated with low bone mineral density (BMD),7,8 poor physical function,9 and deterioration in fine motor skills10 in communityexposed adults and occupational cohorts,11 but concurrent associations between blood lead concentration (BPb), BMD, and physical and cognitive function in environmentally exposed men remain unexplored. The NHANES survey from 2007/08 showed that older adults had a higher BPb (1.4 lg/dL) than the population mean of 1.3 lg/dL and that men had a higher BPb than women.12 The skeleton stores 95% of the lead in the body, with a half life of up to 27 years.13 During rapid bone turnover related to aging, lead can be released into the circulation, where it becomes bioavailable to exert toxic effects on target organs. The current study examined the relationship
0002-8614/14/$15.00
142
KHALIL ET AL.
between BPb, BMD, and physical and cognitive performance in a cohort of older men.
METHODS Study Population The Osteoporotic Fractures in Men Study (MrOS) is a prospective cohort study of community-dwelling noninstitutionalized men in the United States. Between 2000 and 2002, 5,994 men aged 65 and older were enrolled from population-based listings at six clinical sites (Birmingham, AL; Minneapolis, MN; Pittsburgh, PA; Palo Alto, CA; Portland, OR; and San Diego, CA). Eligibility criteria included the ability to walk unaided and without bilateral hip replacements. A cross sectional study was completed from May 4, 2007, to November 12, 2008, involving participants recruited as a subset of MrOS. This ancillary BPb study was conducted in 445 non-Hispanic Caucasian participants enrolled in MrOS at the University of Pittsburgh clinic (Monongahela Valley, near Pittsburgh). Complete data on BMD and physical performance at MrOS Visit 3 (2007–2009) were available for these participants. The institutional review board approved the study protocol, and written informed consent was obtained from all participants. Details of the MrOS design, recruitment, and baseline characteristics have been described.14
Data Collection and Assessment Procedures Blood Lead Measurements For lead measurement, a 0.5-mL blood sample was taken by venipuncture according to published standardized study protocols.14 Lead level was determined using atomic absorption spectrophotometry conducted15 by the Special Chemistry Laboratory, an affiliate of the University of Pittsburgh Medical Center. This laboratory is certified for the analysis of lead in blood by the College of American Pathologists, Pennsylvania Department of Health, and Centers for Disease Control and Prevention. Lead concentrations were assayed on an inductively coupled plasma mass spectrometer (ELAN DRCe; PerkinElmer, Waltham, MA). The analytical limit of detection (LoD) by the laboratory was 1 lg/dL. A value of 0.71 lg/dL was substituted for statistical analysis for 10 participants with BPb concentrations below LoD, following usual practice (LoD divided by the square root of 2).16
Measurement of BMD Bone mineral density (g/cm2) was measured at the lumbar spine and the total hip and its subregion femoral neck using dual-energy X-ray absorptiometry (DXA; QDR-4500 W scanner, Hologic, Inc., Bedford, MA). Anteroposterior lumbar spine BMD was measured in all men, and mean BMD was computed from the first through fourth lumbar vertebrae. Hip BMD was measured on the right side. If a right hip replacement or metal objects in the right leg were reported, the left hip was measured. The quality control of DXA measurements included certification of DXA
JANUARY 2014–VOL. 62, NO. 1
JAGS
technicians and implementation of a standardized scanning procedure. At each study visit, a hip phantom was circulated and scanned at the six clinical sites. No linear differences were detected between scanners in crosscalibration studies. The interscanner coefficient of variation (CV) was 0.9%. For spine BMD, maximum difference between scanners was 1.4%.12
Other Measurements Self-administered questionnaires were used to ascertain participant demographic and lifestyle information. Trained study staff assessed anthropometric measures and neuromuscular function during interviews or examinations. Information on education (years of education completed) was obtained at baseline (2000–2002).14 Greater education is associated with lower osteoporosis risk.17 Lifestyle factors included self-report of smoking (current, past, never), and alcohol intake (drinks/wk). Smoking and alcohol intake are known risk factors for osteoporosis in men.18 A calibrated balance-beam scale was used to measure participant weight in kilograms in indoor clothing without shoes. Height was measured in meters using a Harpenden stadiometer (Holtain Ltd., Dyfed, UK). Body mass index (BMI) was calculated as weight in kilograms divided by height in meters.2 Greater body weight is protective against osteoporosis.18
Physical Performance Physical performance was measured using five tests: grip strength, leg extension power, walking speed, narrow-walk pace, and chair stands using arms: yes or no. Grip strength was measured using a JAMAR dynamometer (Sammons Preston, Bolingbrook, IL). The average value (kilograms) from two trials from both hands was analyzed. Leg extension power was measured using the Nottingham power rig.19 Leg extension power measured using this device correlates well with functional measures such as chair-rising speed, stair-climbing speed and power, and walking speed in elderly adults.19 Five trials for each leg were conducted, and the maximum value (W) from either leg was analyzed. Walking speed was assessed on a standard 6-m walking course.20 Participants were instructed to walk at their normal pace for this examination. The fastest speed (m/s) was analyzed. The narrow walk was used as an indirect measure of dynamic balance.19 Participants were asked to walk a 6-m course while keeping each foot within a 20-cm-wide lane. A trial was considered successful if the participant had no more than two deviations from the lane. Participants were asked to do two trials. If one or both trials were unsuccessful, they were asked to do a third trial. The fastest pace (m/s) from any successfully completed trial was analyzed.19 Ability to stand from a chair without using the arms (yes/no) was also measured.19
Cognitive Measures The modified Mini-Mental State Examination (3MS) was used to assess cognition,20 with higher scores (0–100) representing better performance. It is a global measurement of cognitive function, with components for orientation, concentration, language, praxis, and immediate and delayed
JAGS
JANUARY 2014–VOL. 62, NO. 1
memory. The Trail-Making Test Part B is a timed test that measures attention, sequencing, visual scanning, and executive function.21 A faster time for completion (0–300 s) represents better cognitive performance.
Statistical analysis Distribution and normality of outcome variables, covariates, and BPb was ascertained. Based on skewed distribution, BPb was log transformed (mean 2.25 lg/dL, median 2 lg/dL, range 1–10 lg/dL). To explore the possible nonlinear association between outcome variables and BPb, three-knot spline analysis was performed on logtransformed BPb (knots were placed at 25th percentile = 0.0, 50th percentile = 0.69, 75th percentile = 1.09). The Wald test was used to assess whether all four segments simultaneously had 0 slope (P < .001 for all three dependent variables). There was a significant nonlinear relationship between BPb and BMD at the total hip and femoral neck and grip strength and a linear relationship between BPb and BMD at the lumbar spine, cognitive function, leg extension power, walking speed, and narrow-walk pace. Participant characteristics were compared by categorizing them in BPb quartiles (Quartile 1, ≤1.0 lg/dL, n = 126; Quartile 2, 1.1–2.0 lg/dL, n = 175; Quartile 3, 2.1–3.0 lg/dL, n = 84; Quartile 4, >3.0 lg/dL, n = 60) using SAS Proc GLM (SAS Institute, Inc., Cary, NC) for continuous variables and chi-square tests for categorical variables. Further age- and multivariate-adjusted analysis for total hip and femoral neck BMD and grip strength was completed using Proc GLM across BPb quartiles (untransformed BPb). As described above, in spline analysis, there was a linear association between BPb and BMD at the lumbar spine, cognitive function, leg extension power, walking speed, and narrow-walk pace. Based on this linear association with BPb, these variables were analyzed in separate age- and multivariate-adjusted models as continuous outcomes. Association between BPb and chair stands (use of arms, yes/no) was analyzed using logistic regression; odds ratios and 95% confidence intervals are reported. For each outcome, model building began by regressing BPb on that function; sequentially adding age; and finally adding education, smoking, alcohol consumption, and BMI simultaneously. Selection of these covariates was decided a priori based on the literature and previous work.8 Analysis was performed using SAS 9.2 (SAS Institute, Inc.). STATA version 9 (StataCorp, College Station, TX) was used for spline analysis. Two-tailed P-values were used for all tests, at 5% statistical significance.
RESULTS Characteristics of the Participants Mean BPb standard deviation was 2.25 1.20 lg/dL (median 2 lg/dL, range 1–10 lg/dL), and mean age was 79.5 5 (Table 1). Average age was significantly higher in men with higher BPb (P-trend = .004). BMI, education, and smoking pattern did not differ between BPb quartiles. Alcohol consumption was highest in the second quartile and decreased across the rest of BPb strata (P = .01). In unadjusted models (Table 1), men in BPb quartiles 3 and 4
BLOOD LEAD AND PHYSCIAL FUNCTION IN ELDERLY MEN
143
had lower mean BMD at the femoral neck than those in quartile 2 (P-trend < .001) and total hip (P-trend = .004). Although lumbar spine BMD decreased from BPb quartile 2 to 4, it was not statistically significant. A significant trend of lower grip strength (P-trend = .006) from BPb quartile 2 to 4 was seen. Although decreases in all physical and cognitive functions from BPb quartile 2 to 4 were observed, they were not statistically significant.
Age- and Multivariate-Adjusted Models In age-adjusted models, men in BPb quartiles 3 and 4 had lower BMD at femoral neck and at total hip than those in quartile 2 (P-trend < .001 for both) (Table 2). Upon further adjustment with covariates, the significant difference persisted at the femoral neck (P-trend < .001) and total hip (P < .004). In age-adjusted models, grip strength decreased from BPb quartile 2 to 4 (P-trend < .001). Although the pattern of decreasing grip strength from BPb quartile 2 to 4 remained in multivariate-adjusted models, it was not statistically significant (Table 2). No significant association between BPB, lumbar spine BMD, cognitive function, and physical function was noted in age- and multivariate-adjusted models (Table 3).
DISCUSSION The results of this study suggest that a circulating average lead concentration of 2 lg/dL is inversely associated with BMD at the femoral neck and total hip in older men, although no effect was seen on grip strength, leg power, walking speed, narrow walk pace, chair stands, cognitive function, or lumbar spine BMD. The study provides further evidence that, despite the significant decline in mean BPb in the U.S. population that has followed the lowering of the concentration of lead in the environment since leaded gasoline use was banned in the United States in 1976, older adults with higher lifetime environmental lead exposure may have some risk of adverse bone changes. To the knowledge of the authors, this is the first study to report an inverse association between BPb and BMD in older men even at current low environmental exposure to lead. Several lines of evidence suggest that BPb may have a causative role in osteoporosis. Lead is contained in the mineral matrix of bone in close proximity with calcium, mimicking many physical and chemical properties. Lead competes with calcium in intestinal absorption, transport, and bone deposition.13 Similar to calcium, parathyroid hormone and vitamin D regulate lead absorption from the intestine.13 A trend toward higher BPb with age was noted in this study, possibly related to age-associated osteoporosis. Mean BPb in men aged 80 and older was 2.5 lg/dL, compared with 2.1 lg/dL in those aged 70 to 75 (P = .004, data not shown). Although BMD was tested at three bone sites, a significant inverse association with BPb was seen only at the total hip and femoral neck, which is consistent with existing research.23 There was no association between lead exposure and lumbar spine BMD. In the MrOS cohort (n = 5,995), hip BMD declined with age, whereas spine BMD increased.25 A significant inverse correlation was noted between BMD and BPb at the total hip (correlation coefficient (r) = 0.14, P = .003) and femoral neck (r = 0.14,
144
KHALIL ET AL.
JANUARY 2014–VOL. 62, NO. 1
JAGS
Table 1. Characteristics of Osteoporotic Fractures in Men Study Lead Study Participants According to Blood Lead Concentration Quartile Blood Lead Concentration Quartile (lg/dL)
Characteristic
Age, mean SD (n = 445) Education, college or more, n (%) (n = 445) Body mass index, kg/m2, mean SD (n = 445) Current smoker, n (%) (n = 445) Number of alcoholic drinks/wk, median (interquartile range) (n = 444) Bone mineral density, g/cm2, mean SD Femoral neck (n = 438) Total hip (n = 438) Lumbar spine (n = 434) Physical function, mean SD Grip strength, kg (n = 432) Leg extension power, W (n = 329) Walk speed, m/s (n = 435) Narrow walk, m/s (n = 362) Use arms to stand up, n (%) (n = 445) Cognitive function, mean SD Trail-Making Test Part B score, seconds (range 0–300) (n = 414)b Modified Mini-Mental State Examination score (range 0–100) (n = 442)
Overall, N = 445
1 (≤1.0), n = 126
2 (1.1–2.0), n = 175
3 (2.1–3.0), n = 84
4 (>3.0), n = 60
P-Value
79.5 4.9 214 (48)
78.9 4.6 66 (52)
79.1 4.9 86 (49)
79.5 4.8 39 (46)
81.5 5.1 23 (38)
.004 .34
28.2 4.2
28.6 4.5
28.0 3.8
28.0 4.9
27.9 3.6
.66
11 (2.5) 0.23 (4.3)
0 (0.0) 0.00 (2.0)
7 (4.0) 0.69 (5.0)
3 (3.6) 0.23 (7.0)
1 (1.7) 0.23 (7.0)
.14 .01a
0.8 0.1 0.9 0.2 1.3 0.3
0.8 0.1 1.0 0.2 1.3 0.3
0.8 0.1 1.0 0.2 1.3 0.4
0.8 0.2 0.9 0.2 1.3 0.3
0.7 0.1 0.9 0.1 1.3 0.3
.001 .004 .18
34.8 8.2 164.9 53
35.4 7.4 166.4 49.8
35.9 7.6 169.7 55.3
33.9 9.9 159.7 53.8
31.8 8.2 154.0 46.8
.006 .31
1.0 0.2 1.1 0.2 59 (13.3)
1.0 0.2 1.1 0.3 19 (15.1)
1.0 0.2 1.1 0.2 17 (9.7)
1.0 0.2 1.1 0.3 13 (15.5)
1.0 0.2 1.0 0.3 10 (16.7)
.09 .26 .36
123.2 57.0
120.7 56.4
120.8 57.2
122.8 51.2
136.6 64.8
.34
92.3 6.6
92.5 7.1
92.7 6.1
92.3 5.8
90.7 7.7
.25
SD = Standard Deviation. a Wilcoxon nonparametric test. b Higher scores indicate poorer cognitive function.
Table 2. Age- and Multivariate-Adjusted Bone Mineral Density and Grip Strength According to Blood Lead Concentration Quartile Blood Lead Concentration Quartile, lg/dL Characteristic
1
Bone mineral density Femoral neck Age adjusted 0.78 (0.74–0.79) 0.77 (0.75–0.79) Multivariate adjusteda Total hip Age adjusted 0.96 (0.93–0.98) 0.95 (0.93–0.98) Multivariate adjusteda Grip strength, average of right and left, kg Age adjusted 34.9 (33.6–36.3) 34.9 (33.6–36.3) Multivariate adjusteda
2
3
4
P-Trend
0.80 (0.78–0.82) 0.80 (0.78–0.82)
0.76 (0.73–0.79) 0.76 (0.73–0.79)
0.74 (0.71–0.77) 0.73 (0.70–0.76)
OBJECTIVES: To study the association between blood lead concentration (BPb) and bone mineral density (BMD), physical function, and cognitive function in noninstitutionalized community-dwelling older men. DESIGN: Cross-sectional study. SETTING: University of Pittsburgh clinic, Pittsburgh, Pennsylvania. PARTICIPANTS: Non-Hispanic Caucasian men aged 65 and older (N = 445) recruited as a subset of a prospective cohort for the Osteoporotic Fractures in Men Study. MEASUREMENTS: BPb was measured in 2007/08. From 2007 to 2009, BMD (g/cm2) was measured using dualenergy X-ray absorptiometry. At the same time, physical performance was measured using five tests: grip strength, leg extension power, walking speed, narrow-walk pace, and chair stands. Cognitive performance was assessed using the modified Mini-Mental State Examination and the Trail-Making Test Part B. Participants were categorized into quartiles of BPb. Multivariate regression analysis was used to evaluate the independent relationship between BPb, BMD, and cognitive and physical function. RESULTS: Mean BPb standard deviation was 2.25 1.20 lg/dL (median 2 lg/dL, range 1–10 lg/dL). In multivariate-adjusted models, men in higher BPb quartiles had lower BMD at femoral neck and total hip (P-trend < .001 for both). Men with higher BPb had lower age-adjusted score for grip strength (P-trend < .001), although this association was not significant in multivariate-adjusted models (P-trend < .15). BPb was not associated with lumbar spine BMD, cognition, leg extension power, walking speed, narrow-walk pace, or chair stands. From the *Center for Global Health, Department of Community Health, Boonshoft School of Medicine, Wright State University, Dayton, Ohio; † Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, and ‡Division of Geriatric Medicine, Department of Medicine, Medical School, University of Pittsburgh, Pittsburgh, Pennsylvania. Address correspondence to Dr. Naila Khalil, Center for Global Health, Department of Community Health, Boonshoft School of Medicine, Wright State University, 3123 Research Blvd, Suite #200, Dayton, OH 45430. E-mail: [email protected] DOI: 10.1111/jgs.12603
JAGS 62:141–146, 2014 © 2013, Copyright the Authors Journal compilation © 2013, The American Geriatrics Society
CONCLUSION: Environmental lead exposure may adversely affect bone health in older men. These findings support consideration of environmental exposure in ageassociated bone fragility. J Am Geriatr Soc 62:141–146, 2014.
Key words: lead; elderly; men; bone; grip strength; cognition; physical function
A
ge-related osteoporosis and decline in physical function are significant public health concerns because they are associated with falls,1,2 fractures,2,3 frailty,2 and morbidity.4,5 In the U.S. National Health and Nutrition Examination Survey (NHANES) data completed in 2005 to 2008, 9% of adults aged 50 and older had osteoporosis at the femoral neck or lumbar spine, and half had low bone mass at one of these two sites.6 It is estimated that one-third of all osteoporotic fractures and 25% of osteoporosis-related costs are observed in men,6 with higher associated mortality than in women.6 The healthcare burden will increase with increasing longevity in the population,6 so it is important to identify determinants of bone strength and physical function in older men, including environmental exposures such as lead. Lead is a toxic element highly prevalent in the environment. Lead exposure has been associated with low bone mineral density (BMD),7,8 poor physical function,9 and deterioration in fine motor skills10 in communityexposed adults and occupational cohorts,11 but concurrent associations between blood lead concentration (BPb), BMD, and physical and cognitive function in environmentally exposed men remain unexplored. The NHANES survey from 2007/08 showed that older adults had a higher BPb (1.4 lg/dL) than the population mean of 1.3 lg/dL and that men had a higher BPb than women.12 The skeleton stores 95% of the lead in the body, with a half life of up to 27 years.13 During rapid bone turnover related to aging, lead can be released into the circulation, where it becomes bioavailable to exert toxic effects on target organs. The current study examined the relationship
0002-8614/14/$15.00
142
KHALIL ET AL.
between BPb, BMD, and physical and cognitive performance in a cohort of older men.
METHODS Study Population The Osteoporotic Fractures in Men Study (MrOS) is a prospective cohort study of community-dwelling noninstitutionalized men in the United States. Between 2000 and 2002, 5,994 men aged 65 and older were enrolled from population-based listings at six clinical sites (Birmingham, AL; Minneapolis, MN; Pittsburgh, PA; Palo Alto, CA; Portland, OR; and San Diego, CA). Eligibility criteria included the ability to walk unaided and without bilateral hip replacements. A cross sectional study was completed from May 4, 2007, to November 12, 2008, involving participants recruited as a subset of MrOS. This ancillary BPb study was conducted in 445 non-Hispanic Caucasian participants enrolled in MrOS at the University of Pittsburgh clinic (Monongahela Valley, near Pittsburgh). Complete data on BMD and physical performance at MrOS Visit 3 (2007–2009) were available for these participants. The institutional review board approved the study protocol, and written informed consent was obtained from all participants. Details of the MrOS design, recruitment, and baseline characteristics have been described.14
Data Collection and Assessment Procedures Blood Lead Measurements For lead measurement, a 0.5-mL blood sample was taken by venipuncture according to published standardized study protocols.14 Lead level was determined using atomic absorption spectrophotometry conducted15 by the Special Chemistry Laboratory, an affiliate of the University of Pittsburgh Medical Center. This laboratory is certified for the analysis of lead in blood by the College of American Pathologists, Pennsylvania Department of Health, and Centers for Disease Control and Prevention. Lead concentrations were assayed on an inductively coupled plasma mass spectrometer (ELAN DRCe; PerkinElmer, Waltham, MA). The analytical limit of detection (LoD) by the laboratory was 1 lg/dL. A value of 0.71 lg/dL was substituted for statistical analysis for 10 participants with BPb concentrations below LoD, following usual practice (LoD divided by the square root of 2).16
Measurement of BMD Bone mineral density (g/cm2) was measured at the lumbar spine and the total hip and its subregion femoral neck using dual-energy X-ray absorptiometry (DXA; QDR-4500 W scanner, Hologic, Inc., Bedford, MA). Anteroposterior lumbar spine BMD was measured in all men, and mean BMD was computed from the first through fourth lumbar vertebrae. Hip BMD was measured on the right side. If a right hip replacement or metal objects in the right leg were reported, the left hip was measured. The quality control of DXA measurements included certification of DXA
JANUARY 2014–VOL. 62, NO. 1
JAGS
technicians and implementation of a standardized scanning procedure. At each study visit, a hip phantom was circulated and scanned at the six clinical sites. No linear differences were detected between scanners in crosscalibration studies. The interscanner coefficient of variation (CV) was 0.9%. For spine BMD, maximum difference between scanners was 1.4%.12
Other Measurements Self-administered questionnaires were used to ascertain participant demographic and lifestyle information. Trained study staff assessed anthropometric measures and neuromuscular function during interviews or examinations. Information on education (years of education completed) was obtained at baseline (2000–2002).14 Greater education is associated with lower osteoporosis risk.17 Lifestyle factors included self-report of smoking (current, past, never), and alcohol intake (drinks/wk). Smoking and alcohol intake are known risk factors for osteoporosis in men.18 A calibrated balance-beam scale was used to measure participant weight in kilograms in indoor clothing without shoes. Height was measured in meters using a Harpenden stadiometer (Holtain Ltd., Dyfed, UK). Body mass index (BMI) was calculated as weight in kilograms divided by height in meters.2 Greater body weight is protective against osteoporosis.18
Physical Performance Physical performance was measured using five tests: grip strength, leg extension power, walking speed, narrow-walk pace, and chair stands using arms: yes or no. Grip strength was measured using a JAMAR dynamometer (Sammons Preston, Bolingbrook, IL). The average value (kilograms) from two trials from both hands was analyzed. Leg extension power was measured using the Nottingham power rig.19 Leg extension power measured using this device correlates well with functional measures such as chair-rising speed, stair-climbing speed and power, and walking speed in elderly adults.19 Five trials for each leg were conducted, and the maximum value (W) from either leg was analyzed. Walking speed was assessed on a standard 6-m walking course.20 Participants were instructed to walk at their normal pace for this examination. The fastest speed (m/s) was analyzed. The narrow walk was used as an indirect measure of dynamic balance.19 Participants were asked to walk a 6-m course while keeping each foot within a 20-cm-wide lane. A trial was considered successful if the participant had no more than two deviations from the lane. Participants were asked to do two trials. If one or both trials were unsuccessful, they were asked to do a third trial. The fastest pace (m/s) from any successfully completed trial was analyzed.19 Ability to stand from a chair without using the arms (yes/no) was also measured.19
Cognitive Measures The modified Mini-Mental State Examination (3MS) was used to assess cognition,20 with higher scores (0–100) representing better performance. It is a global measurement of cognitive function, with components for orientation, concentration, language, praxis, and immediate and delayed
JAGS
JANUARY 2014–VOL. 62, NO. 1
memory. The Trail-Making Test Part B is a timed test that measures attention, sequencing, visual scanning, and executive function.21 A faster time for completion (0–300 s) represents better cognitive performance.
Statistical analysis Distribution and normality of outcome variables, covariates, and BPb was ascertained. Based on skewed distribution, BPb was log transformed (mean 2.25 lg/dL, median 2 lg/dL, range 1–10 lg/dL). To explore the possible nonlinear association between outcome variables and BPb, three-knot spline analysis was performed on logtransformed BPb (knots were placed at 25th percentile = 0.0, 50th percentile = 0.69, 75th percentile = 1.09). The Wald test was used to assess whether all four segments simultaneously had 0 slope (P < .001 for all three dependent variables). There was a significant nonlinear relationship between BPb and BMD at the total hip and femoral neck and grip strength and a linear relationship between BPb and BMD at the lumbar spine, cognitive function, leg extension power, walking speed, and narrow-walk pace. Participant characteristics were compared by categorizing them in BPb quartiles (Quartile 1, ≤1.0 lg/dL, n = 126; Quartile 2, 1.1–2.0 lg/dL, n = 175; Quartile 3, 2.1–3.0 lg/dL, n = 84; Quartile 4, >3.0 lg/dL, n = 60) using SAS Proc GLM (SAS Institute, Inc., Cary, NC) for continuous variables and chi-square tests for categorical variables. Further age- and multivariate-adjusted analysis for total hip and femoral neck BMD and grip strength was completed using Proc GLM across BPb quartiles (untransformed BPb). As described above, in spline analysis, there was a linear association between BPb and BMD at the lumbar spine, cognitive function, leg extension power, walking speed, and narrow-walk pace. Based on this linear association with BPb, these variables were analyzed in separate age- and multivariate-adjusted models as continuous outcomes. Association between BPb and chair stands (use of arms, yes/no) was analyzed using logistic regression; odds ratios and 95% confidence intervals are reported. For each outcome, model building began by regressing BPb on that function; sequentially adding age; and finally adding education, smoking, alcohol consumption, and BMI simultaneously. Selection of these covariates was decided a priori based on the literature and previous work.8 Analysis was performed using SAS 9.2 (SAS Institute, Inc.). STATA version 9 (StataCorp, College Station, TX) was used for spline analysis. Two-tailed P-values were used for all tests, at 5% statistical significance.
RESULTS Characteristics of the Participants Mean BPb standard deviation was 2.25 1.20 lg/dL (median 2 lg/dL, range 1–10 lg/dL), and mean age was 79.5 5 (Table 1). Average age was significantly higher in men with higher BPb (P-trend = .004). BMI, education, and smoking pattern did not differ between BPb quartiles. Alcohol consumption was highest in the second quartile and decreased across the rest of BPb strata (P = .01). In unadjusted models (Table 1), men in BPb quartiles 3 and 4
BLOOD LEAD AND PHYSCIAL FUNCTION IN ELDERLY MEN
143
had lower mean BMD at the femoral neck than those in quartile 2 (P-trend < .001) and total hip (P-trend = .004). Although lumbar spine BMD decreased from BPb quartile 2 to 4, it was not statistically significant. A significant trend of lower grip strength (P-trend = .006) from BPb quartile 2 to 4 was seen. Although decreases in all physical and cognitive functions from BPb quartile 2 to 4 were observed, they were not statistically significant.
Age- and Multivariate-Adjusted Models In age-adjusted models, men in BPb quartiles 3 and 4 had lower BMD at femoral neck and at total hip than those in quartile 2 (P-trend < .001 for both) (Table 2). Upon further adjustment with covariates, the significant difference persisted at the femoral neck (P-trend < .001) and total hip (P < .004). In age-adjusted models, grip strength decreased from BPb quartile 2 to 4 (P-trend < .001). Although the pattern of decreasing grip strength from BPb quartile 2 to 4 remained in multivariate-adjusted models, it was not statistically significant (Table 2). No significant association between BPB, lumbar spine BMD, cognitive function, and physical function was noted in age- and multivariate-adjusted models (Table 3).
DISCUSSION The results of this study suggest that a circulating average lead concentration of 2 lg/dL is inversely associated with BMD at the femoral neck and total hip in older men, although no effect was seen on grip strength, leg power, walking speed, narrow walk pace, chair stands, cognitive function, or lumbar spine BMD. The study provides further evidence that, despite the significant decline in mean BPb in the U.S. population that has followed the lowering of the concentration of lead in the environment since leaded gasoline use was banned in the United States in 1976, older adults with higher lifetime environmental lead exposure may have some risk of adverse bone changes. To the knowledge of the authors, this is the first study to report an inverse association between BPb and BMD in older men even at current low environmental exposure to lead. Several lines of evidence suggest that BPb may have a causative role in osteoporosis. Lead is contained in the mineral matrix of bone in close proximity with calcium, mimicking many physical and chemical properties. Lead competes with calcium in intestinal absorption, transport, and bone deposition.13 Similar to calcium, parathyroid hormone and vitamin D regulate lead absorption from the intestine.13 A trend toward higher BPb with age was noted in this study, possibly related to age-associated osteoporosis. Mean BPb in men aged 80 and older was 2.5 lg/dL, compared with 2.1 lg/dL in those aged 70 to 75 (P = .004, data not shown). Although BMD was tested at three bone sites, a significant inverse association with BPb was seen only at the total hip and femoral neck, which is consistent with existing research.23 There was no association between lead exposure and lumbar spine BMD. In the MrOS cohort (n = 5,995), hip BMD declined with age, whereas spine BMD increased.25 A significant inverse correlation was noted between BMD and BPb at the total hip (correlation coefficient (r) = 0.14, P = .003) and femoral neck (r = 0.14,
144
KHALIL ET AL.
JANUARY 2014–VOL. 62, NO. 1
JAGS
Table 1. Characteristics of Osteoporotic Fractures in Men Study Lead Study Participants According to Blood Lead Concentration Quartile Blood Lead Concentration Quartile (lg/dL)
Characteristic
Age, mean SD (n = 445) Education, college or more, n (%) (n = 445) Body mass index, kg/m2, mean SD (n = 445) Current smoker, n (%) (n = 445) Number of alcoholic drinks/wk, median (interquartile range) (n = 444) Bone mineral density, g/cm2, mean SD Femoral neck (n = 438) Total hip (n = 438) Lumbar spine (n = 434) Physical function, mean SD Grip strength, kg (n = 432) Leg extension power, W (n = 329) Walk speed, m/s (n = 435) Narrow walk, m/s (n = 362) Use arms to stand up, n (%) (n = 445) Cognitive function, mean SD Trail-Making Test Part B score, seconds (range 0–300) (n = 414)b Modified Mini-Mental State Examination score (range 0–100) (n = 442)
Overall, N = 445
1 (≤1.0), n = 126
2 (1.1–2.0), n = 175
3 (2.1–3.0), n = 84
4 (>3.0), n = 60
P-Value
79.5 4.9 214 (48)
78.9 4.6 66 (52)
79.1 4.9 86 (49)
79.5 4.8 39 (46)
81.5 5.1 23 (38)
.004 .34
28.2 4.2
28.6 4.5
28.0 3.8
28.0 4.9
27.9 3.6
.66
11 (2.5) 0.23 (4.3)
0 (0.0) 0.00 (2.0)
7 (4.0) 0.69 (5.0)
3 (3.6) 0.23 (7.0)
1 (1.7) 0.23 (7.0)
.14 .01a
0.8 0.1 0.9 0.2 1.3 0.3
0.8 0.1 1.0 0.2 1.3 0.3
0.8 0.1 1.0 0.2 1.3 0.4
0.8 0.2 0.9 0.2 1.3 0.3
0.7 0.1 0.9 0.1 1.3 0.3
.001 .004 .18
34.8 8.2 164.9 53
35.4 7.4 166.4 49.8
35.9 7.6 169.7 55.3
33.9 9.9 159.7 53.8
31.8 8.2 154.0 46.8
.006 .31
1.0 0.2 1.1 0.2 59 (13.3)
1.0 0.2 1.1 0.3 19 (15.1)
1.0 0.2 1.1 0.2 17 (9.7)
1.0 0.2 1.1 0.3 13 (15.5)
1.0 0.2 1.0 0.3 10 (16.7)
.09 .26 .36
123.2 57.0
120.7 56.4
120.8 57.2
122.8 51.2
136.6 64.8
.34
92.3 6.6
92.5 7.1
92.7 6.1
92.3 5.8
90.7 7.7
.25
SD = Standard Deviation. a Wilcoxon nonparametric test. b Higher scores indicate poorer cognitive function.
Table 2. Age- and Multivariate-Adjusted Bone Mineral Density and Grip Strength According to Blood Lead Concentration Quartile Blood Lead Concentration Quartile, lg/dL Characteristic
1
Bone mineral density Femoral neck Age adjusted 0.78 (0.74–0.79) 0.77 (0.75–0.79) Multivariate adjusteda Total hip Age adjusted 0.96 (0.93–0.98) 0.95 (0.93–0.98) Multivariate adjusteda Grip strength, average of right and left, kg Age adjusted 34.9 (33.6–36.3) 34.9 (33.6–36.3) Multivariate adjusteda
2
3
4
P-Trend
0.80 (0.78–0.82) 0.80 (0.78–0.82)
0.76 (0.73–0.79) 0.76 (0.73–0.79)
0.74 (0.71–0.77) 0.73 (0.70–0.76)
