Abstract
Introduction:
Lipoprotein-associated phospholipase A2 (Lp-PLA2) has been consistently associated with cardiovascular disease (CVD) risk factors and predictive of CVD outcomes; furthermore, it is consistently higher among type 2 diabetics than nondiabetics. However, the relationships of circulating Lp-PLA2 mass and activity with incident type 2 diabetes mellitus have not been examined. Therefore, the purpose of this study was to determine the association of Lp-PLA2 mass and activity with type 2 diabetes among older adults.
Methods:
We conducted analyses of Lp-PLA2 and prevalent and incident diabetes among 5474 men and women from the Cardiovascular Health Study (1989–2007). Lp-PLA2 mass and activity were measured in baseline plasma. Diabetes status was ascertained annually with medication inventories and repeated blood glucose measurements. Generalized linear and Cox proportional hazards models were used to adjust for confounding factors including body mass index and inflammation.
Results:
At baseline, the top two quintiles of Lp-PLA2 activity were significantly associated with prevalent type 2 diabetes with a multivariable relative risk = 1.35 [95% confidence interval (CI) = 1.11–1.63] for quintile 4, and relative risk = 1.33 (95% CI = 1.07–1.66) for quintile 5. Among participants free of diabetes at baseline, we found a significant positive association with both the homeostatic model assessment for insulin resistance and β-cell function per sd increase in Lp-PLA2 activity (P values for both <0.01). In prospective analyses, the risk of incident type 2 diabetes was significantly higher among those in the highest quintile of Lp-PLA2 activity [multivariable hazard ratio = 1.45 (95% CI = 1.01–2.07)] compared with the lowest quintile. Lp-PLA2 mass was not significantly associated with incident type 2 diabetes.
Discussion:
Lp-PLA2 activity is positively associated with insulin resistance and predicts incident type 2 diabetes among older adults independent of multiple factors associated with diabetes pathogenesis.
Lipoprotein-associated phospholipase A2 (Lp-PLA2) is a proinflammatory enzyme that has been independently associated with cardiovascular disease (CVD) (1–3). Measured in the plasma, it is a marker for inflammation in the vessel wall (3). Lp-PLA2 is primarily secreted by macrophages, and when released, it binds to the apolipoprotein B moiety on low-density lipoprotein (LDL) particles and remains latent until LDL is oxidized. Oxidized phosphatidylcholine contained in the oxidized LDL acts as a substrate for Lp-PLA2. Lp-PLA2 cleaves the oxidized phosphatidylcholine into two bioactive compounds lysophosphatidylcholine and oxidized nonesterified free fatty acids (4), both of which are considered proinflammatory (3).
There are several plausible mechanisms linking Lp-PLA2 with diabetes, including an increase in inflammatory activity associated with Lp-PLA2's hydrolysis of oxidized phospholipids that could induce insulin resistance (5, 6), thereby increasing the risk for type 2 diabetes. Alternatively, increased Lp-PLA2 could be a marker of increased adipose tissue inflammation, other metabolic processes such as increased fatty acid flux, and/or adipokine perturbations that underlie the development of insulin resistance and pancreatic β-cell failure (7).
In cross-sectional studies, both Lp-PLA2 activity and/or mass were higher among individuals with diabetes than among those without diabetes (8–11). We are aware of only one study that evaluated the prospective association between Lp-PLA2 and incident type 2 diabetes (12). This study, including 889 middle-aged males, reported no association of Lp-PLA2 mass with risk of type 2 diabetes (measurements of Lp-PLA2 activity were not obtained).
The purpose of this study is to determine whether baseline Lp-PLA2 activity and mass are associated with risk of incident type 2 diabetes among older adults, independent of other diabetes risk factors.
Subjects and Methods
Study population
Participants were members of the Cardiovascular Health Study (CHS), a population-based cohort study of risk factors for CVD in older adults. The design and recruitment are described in detail in previous publications (13, 14). A total of 5888 adults aged 65 yr and older were recruited from a random sample of Medicare-eligible residents of four U.S. communities including Forsyth County, NC; Sacramento County, CA; Washington County, MD; and Pittsburgh, PA. Enrollment of 5201 participants was completed in 1989/1990, and an additional 687 African-American subjects were enrolled in 1992/1993. Participants were excluded if they were institutionalized, wheelchair-bound, or currently receiving treatment for cancer. All subjects gave written informed consent for participation in the study, and all procedures were conducted under institutionally approved protocols at each center. Mean age at enrollment was 73 yr (range 65–100); 58% were female, and 16% were African-American. Baseline evaluations included standardized clinic examination, laboratory evaluation, and questionnaires on health status, medical history, and risk factors.
We excluded 55 participants missing information on baseline diabetes status and 359 missing either Lp-PLA2 mass or activity measures, leaving 5474 participants overall. For the primary analysis of incident diabetes, we excluded participants with prevalent diabetes at baseline and those with no diabetes ascertainment beyond baseline (n = 863 and n = 103, respectively) leaving 4508 participants for this analysis.
At the baseline clinic visit, participants answered standardized questionnaires assessing demographics, CVD risk factors (e.g. tobacco use, alcohol consumption, and physical activity), symptoms, medical history, and medications. Blood tests were done on 8- to 12-h fasting samples taken at the baseline examination for each cohort (1989/1990 for the original cohort and 1992/1993 for the African-American cohort). Multiple aliquots of plasma and serum were collected and shipped to the CHS Central Laboratory at the University of Vermont for analyses of lipids, creatinine, C-reactive protein (CRP), IL-6, and fibrinogen. White blood cells were measured in local laboratories (15). All participants had standardized measures of seated blood pressure and a 12-lead electrocardiogram. We defined hypertension as an average baseline seated blood pressure of at least 140 mm Hg (systolic) or at least 90 mm Hg (diastolic) or a combination of self-reported hypertension and use of antihypertensive medication. Weight and height were measured according to standardized protocols, and body mass index (BMI) was calculated as measured weight in kilograms divided by standing height in meters squared.
Ascertainment of diabetes status
Glucose was measured in fasting blood samples collected during clinic examinations in 1989–1990, 1992–1993, 1996–1997, 1998–1999, and 2005–2006 and nonfasting blood samples in 1994–1995. Medication use was assessed using an inventory (16) at baseline and annually thereafter, through 2007. Participants were categorized as having diabetes mellitus if they reported use of insulin or oral hypoglycemic medications or had a fasting glucose level of at least 7 mmol/liter (126 mg/dl) or a nonfasting glucose level of at least 11.1 mmol/liter (200 mg/dl). Diabetes status was ascertained through 2007.
Lp-PLA2 measurements
GlaxoSmithKline measured plasma concentrations of Lp-PLA2 activity (nanomoles per minute per milliliter) using a tritium-labeled form of platelet-activating factor as substrate in a 96-well microplate, as described in detail previously (2). Samples were tested in duplicate, and the analytical coefficient of variation was 7.5%.
Plasma Lp-PLA2 mass (nanograms per milliliter) was measured at the CHS Central Laboratory at the University of Vermont using an ELISA kit (second-generation PLAC test; diaDexus Inc., South San Francisco, CA). All samples were analyzed in duplicate, and the analytical coefficient of variation was 6.3%.
Homeostatic model assessment of insulin resistance (HOMA-IR) and homeostatic model assessment of β-cell function (HOMA-B)
The HOMA-IR index was calculated as follows: [basal glucose (millimoles per liter) × basal insulin (milliunits per liter)]/22.5. HOMA-B index was computed as follows: [20 × basal insulin (milliunits per liter)]/[basal glucose concentrations (millimoles per liter) − 3.5] (17).
Statistical analysis
We categorized participants by quintiles of Lp-PLA2 activity and Lp-PLA2 mass, based on the distribution among participants free of diabetes at baseline. To assess the association between Lp-PLA2 activity and mass and prevalent diabetes at baseline, we used generalized linear models with log link and Gaussian error structure and used a robust estimator of variance (18). Among participants free of diabetes at baseline and excluding two participants with fasting insulin values more than 25 sd above the mean, we used linear regression to estimate the mean change in baseline HOMA-IR and HOMA-B associated with Lp-PLA2 activity and mass (n = 4506). For the latter analysis, we rescaled each HOMA measure by its sd to facilitate comparisons between the two measures. We used Cox proportional hazards regression stratified by enrollment wave to estimate the relative risk of incident diabetes associated with the Lp-PLA2 categories. Time at risk was calculated as the interval in days from the date of the baseline visit to the earliest of 1) date of the first follow-up study examination at which diabetes was ascertained, 2) date of last study examination at which diabetes could be ascertained, or 3) date of 2006–2007 study examination (end of follow-up for diabetes ascertainment). Multivariable models were adjusted for the following covariates assessed at baseline: age, sex, race (African-American, non-African-American), smoking status (never, former, or current), pack-years of smoking, physical activity (kilocalories per week), alcohol consumption (zero, fewer than, or seven or more drinks/wk), educational attainment (high school or less or at least some college), use of antihypertensive medications, systolic blood pressure, BMI, serum creatinine, CRP, fibrinogen, white blood cell count, triglycerides, LDL-cholesterol, and high-density lipoprotein (HDL)-cholesterol. We performed sensitivity analysis with additional adjustment for HOMA-IR and fasting insulin and fasting glucose, based on our hypothesis that these factors may mediate the effects of Lp-PLA2 on incident diabetes status. We also evaluated for effect modification by BMI and gender using cross-product terms. The proportion of missing data was very low (<3% for any single variable) and missing covariate data were imputed using a single imputation (19). We evaluated the validity of the proportional hazards assumption using Schoenfeld residuals and found no evidence of nonproportionality.
Results
Of the 4508 participants free of diabetes at baseline and with at least one follow-up exam, 40% were men and 86% were Caucasians. The mean age was 72.6 (±5.5) years. Mean concentrations of Lp-PLA2 were 344.9 ± 118.1 ng/ml (range, 57.5–944.3) for mass and 39.3 ± 13.0 nmol/min · ml (range, 8.7–146.7) for activity. Lp-PLA2 mass and activity were modestly correlated (r = 0.50; P < 0.001). Table 1 shows the demographic, clinical, and lifestyle characteristics of CHS participants free of type 2 diabetes at baseline, by quintile of Lp-PLA2 activity and mass.
Table 1.
Demographic, clinical, and lifestyle characteristics of CHS participants free of diabetes at baseline, by quintile of Lp-PLA2 mass and Lp-PLA2 activity
| Characteristic | Lp-PLA2 mass (ng/ml) (n = 4508) |
Lp-PLA2 activity (nmol/min · ml) (n = 4508) |
||||||
|---|---|---|---|---|---|---|---|---|
| ≤244.9 | >299.2–356.6 | >427.7 | Ptrend | ≤28.8 | >34.6–40.5 | >48.9 | Ptrend | |
| Age [mean (sd)] (yr) | 72.0 (5.2) | 72.5 (5.4) | 73.2 (5.8) | <0.001 | 72.5 (5.5) | 72.7 (5.5) | 72.5 (5.2) | 0.69 |
| Male sex [n (%)] | 299 (33.1) | 369 (41.0) | 412 (45.7) | <0.001 | 216 (23.9) | 340 (37.7) | 498 (55.3) | <0.001 |
| African-American [n (%)] | 218 (24.2) | 104 (11.5) | 58 (6.4) | <0.001 | 241 (26.7) | 100 (11.1) | 39 (4.3) | <0.001 |
| High school education or less [n (%)] | 477 (52.9) | 480 (53.3) | 560 (62.2) | <0.001 | 495 (54.9) | 478 (53.1) | 550 (61.0) | <0.001 |
| Current smoking [n (%)] | 99 (11.0) | 109 (12.1) | 136 (15.1) | 0.001 | 113 (12.5) | 104 (11.5) | 121 (13.4) | 0.02 |
| No. alcoholic drinks/wk [n (%)] | ||||||||
| None | 404 (44.8) | 405 (45.0) | 453 (50.3) | 422 (46.8) | 395 (43.8) | 458 (50.8) | ||
| <7 | 347 (38.5) | 349 (38.7) | 341 (37.8) | 336 (37.3) | 371 (41.2) | 325 (36.1) | ||
| ≥7 | 151 (16.7) | 147 (16.3) | 107 (11.9) | 0.001 | 144 (16.0) | 135 (15.0) | 118 (13.1) | 0.005 |
| BMI [mean (sd)] (kg/m2) | 26.4 (4.8) | 26.3 (4.4) | 25.9 (4.3) | 0.04 | 26.5 (5.3) | 26.2 (4.4) | 26.2 (4.0) | 0.42 |
| Waist circumference [mean (sd)] (cm) | 92.7 (13.4) | 93.0 (12.6) | 92.9 (12.7) | 0.27 | 92.1 (14.6) | 93.0 (12.4) | 94.2 (11.8) | <0.001 |
| Baseline fasting glucose [mean (sd)] (mg/dl) | 98.5 (9.9) | 99.3 (9.9) | 99.4 (9.6) | 0.02 | 97.7 (9.7) | 98.6 (9.4) | 101.0 (10.0) | <0.001 |
| Fasting insulina [mean (sd)] (μIU/ml) | 13.1 (7.3) | 13.8 (8.1) | 13.7 (7.1) | 0.006 | 12.5 (7.1) | 13.4 (6.8) | 14.7 (8.2) | <0.001 |
| HOMA-IRa [median (IQR)] | 2.7 (2.0–3.8) | 2.9 (2.2–4.1) | 3.0 (2.2–4.1) | <0.001 | 2.6 (1.9–3.6) | 2.8 (2.1–3.9) | 3.1 (2.3–4.4) | <0.001 |
| HOMA-Ba [median (IQR)] | 121.2 (91.0–165.0) | 127.8 (94.4–168.5) | 125.5 (97.1–166.5) | 0.02 | 117.3 (90.9–158.8) | 124.8 (96.0–166.5) | 127.2 (94.5–167.2) | <0.001 |
| Systolic BP [mean (sd)] (mm Hg) | 133.4 (21.1) | 136.1 (21.7) | 135.7 (22.2) | 0.07 | 135.6 (22.8) | 135.7 (22.6) | 134.7 (20.5) | 0.91 |
| Hypertension [n (%)] | 529 (58.6) | 548 (60.8) | 577 (64.0) | 0.02 | 545 (60.4) | 557 (61.8) | 558 (61.9) | 0.06 |
| HDL [mean (sd)] (mg/dl) | 58.4 (17.0) | 55.9 (15.9) | 53.1 (14.5) | <0.001 | 64.1 (18.1) | 55.5 (14.8) | 47.9 (12.1) | <0.001 |
| LDL [mean (sd)] (mg/dl) | 117.6 (33.4) | 131.6 (34.0) | 142.2 (37.2) | <0.001 | 108.7 (29.9) | 134.2 (33.6) | 148.0 (38.1) | <0.001 |
| Triglycerides [median (IQR)] (mg/dl) | 107.8 (84.8–151.4) | 121.5 (90.9–157.6) | 114.4 (93.7–154.5) | <0.001 | 100.5 (79.8–133.0) | 114.4 (90.0–156.0) | 138.4 (107.8–181.3) | <0.001 |
| Physical activity [mean (sd)] (kcal/wk) | 1779.1 (2080.6) | 1739.1 (2116.6) | 1799.1 (2096.6) | 0.96 | 1647.0 (1964.9) | 1724.1 (1950.5) | 1971.1 (2153.5) | <0.001 |
| Serum creatinine [mean (sd)] (mg/dl) | 1.01 (0.28) | 1.05 (0.29) | 1.09 (0.33) | <0.001 | 0.99 (0.29) | 1.04 (0.30) | 1.12 (0.31) | <0.001 |
| White blood count [mean (sd)] (×1000 mm3) | 5.9 (1.6) | 6.2 (2.4) | 6.4 (1.8) | <0.001 | 6.1 (2.2) | 6.2 (2.0) | 6.3 (1.8) | 0.002 |
| C-reactive protein [median (IQR)] (mg/liter) | 1.6 (0.8–3.2) | 1.6 (0.9–2.8) | 1.9 (1.0–3.4) | 0.002 | 1.9 (0.9–3.8) | 1.6 (0.9–3.0) | 1.7 (1.0–2.9) | 0.05 |
| Fibrinogen [mean (sd)] (mg/dl) | 316.8 (65.7) | 318.1 (61.1) | 322.5 (66.5) | 0.004 | 327.2 (68.9) | 318.2 (61.6) | 316.9 (61.8) | 0.002 |
| Lipid-lowering medication use [n (%)] | 65 (7.2) | 44 (4.9) | 31 (3.4) | <0.001 | 44 (4.9) | 51 (5.7) | 51 (5.7) | 0.47 |
| Hypertension medication use [n (%)] | 396 (43.9) | 369 (41.0) | 402 (44.6) | 0.83 | 388 (43.0) | 387 (43.0) | 390 (43.3) | 0.47 |
We show only quintiles 1, 3, and 5. BP, Blood pressure; IQR, interquartile rage.
Excludes two fasting insulin outliers.
We first evaluated the cross-sectional associations of Lp-PLA2 measures with prevalent diabetes and measures of insulin resistance and secretion. Lp-PLA2 mass was not associated with prevalent type 2 diabetes, but the top two quintiles of Lp-PLA2 activity were associated with an increased prevalence of type 2 diabetes with multivariable relative risk = 1.35 [95% confidence interval (CI) = 1.11–1.63] for quintile 4, and relative risk = 1.33 (95% CI = 1.07–1.66) for quintile 5 (results not shown). Both baseline HOMA measures were positively associated with Lp-PLA2 mass and activity in those without diabetes (Table 2).
Table 2.
Mean adjusted difference in baseline HOMA associated with Lp-PLA2 mass and activity among participants free of diabetes at baseline
| Log(HOMA-IR) β (P value) |
Log(HOMA-B) β (P value) |
|||
|---|---|---|---|---|
| Age-, sex-, and race-adjusted | Multivariable adjustment | Age-, sex-, and race-adjusted | Multivariable adjustment | |
| Mean difference in log(HOMA) per sd (118 ng/ml) of Lp-PLA2 mass | 0.022 (0.003) | 0.015 (0.02) | 0.014 (0.04) | 0.013 (0.04) |
| Mean difference in log(HOMA) per sd (13.0 nmol/min · ml) of Lp-PLA2 activity | 0.059 (<0.001) | 0.034 (<0.001) | 0.030 (<0.001) | 0.018 (0.01) |
The multivariable model includes adjustment for age, race (Black or non-Black), sex, BMI, smoking status (never, former, or current), smoking history (lifetime pack-years), alcohol use (none, fewer than seven drinks/wk, seven or more drinks/wk), systolic blood pressure, triglycerides, LDL, use of lipid-lowering medication, use of antihypertensive medication, C-reactive protein, education (high school or less or more than high school), physical activity (kilocalories), and creatinine.
Over a median follow-up of 11.7 yr, 419 cases of incident diabetes were ascertained. The risk of incident diabetes was significantly higher among those in the highest quintile of Lp-PLA2 activity in all three multivariate models tested; however, the P for trend was not significant in the third model (P = 0.09) (Table 3). Further adjustment for BMI at age 50 or hormone replacement therapy use in women did not appreciably influence the estimates. However, the association between Lp-PLA2 activity and incident diabetes appeared to be stronger among participants with a higher BMI (P for interaction = 0.004). For example, the hazard ratio (HR) of diabetes per sd increase in Lp-PLA2 activity was 1.07 (95% CI = 0.95–1.21) for individuals with BMI below 30 kg/m2 compared with 1.25 (95% CI = 1.07–1.47) for those with BMI of 30 kg/m2 or higher. Relative risk estimates of incident diabetes associated with Lp-PLA2 activity were similar among women and men.
Table 3.
HR of incident diabetes associated with quintiles of baseline Lp-PLA2 mass and Lp-PLA2 activity
| Lp-PLA2 mass (ng/ml) (n = 4508) |
Lp-PLA2 activity (nmol/min · ml) (n = 4508) |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ≤244.9 | >244.9–299.2 | >299.2–356.6 | >356.6–427.7 | >427.7 | Ptrend | ≤28.8 | >28.8–34.6 | >34.6–40.5 | >40.5–48.9 | >48.9 | Ptrend | |
| No. cases of diabetes | 87 | 79 | 88 | 88 | 77 | 74 | 86 | 73 | 82 | 104 | ||
| Person-years | 10,640 | 10,059 | 9,962 | 9,826 | 9,346 | 10,215 | 10,204 | 10,004 | 9,660 | 9,752 | ||
| Model 1 HR (95% CI)a | 1.0 | 1.03 (0.76–1.40) | 1.15 (0.86–1.56) | 1.19 (0.88–1.61) | 1.13 (0.82–1.54) | 0.27 | 1.0 | 1.22 (0.89–1.67) | 1.10 (0.79–1.53) | 1.27 (0.92–1.76) | 1.62 (1.18–2.22) | 0.005 |
| Model 2 HR (95% CI)b | 1.0 | 1.01 (0.74–1.37) | 1.16 (0.85–1.57) | 1.12 (0.82–1.52) | 1.11 (0.80–1.54) | 0.39 | 1.0 | 1.29 (0.93–1.77) | 1.10 (0.78–1.55) | 1.23 (0.87–1.73) | 1.53 (1.07–2.17) | 0.046 |
| Model 3 HR (95% CI)c | 1.0 | 1.01 (0.74–1.37) | 1.16 (0.85–1.58) | 1.12 (0.82–1.53) | 1.11 (0.80–1.53) | 0.40 | 1.0 | 1.25 (0.90–1.72) | 1.06 (0.75–1.50) | 1.17 (0.83–1.66) | 1.45 (1.01–2.07) | 0.09 |
Incident diabetes is defined as fasting glucose of at least 126 mg/dl or use of diabetes medication.
Model 1 is adjusted for age, race, sex, and enrollment wave.
Model 2 is adjusted for model 1, and smoking, systolic blood pressure, physical activity, alcohol consumption, education, antihypertensive medications, BMI, serum creatinine, CRP, Fibrinogen, white blood cells, triglycerides, and LDL.
Model 3 is adjusted for model 1 and model 2 covariates and HDL.
As expected, further adjustment for HOMA-IR materially weakened the observed association [HR = 1.27 (95% CI = 0.89–1.82); P for trend = 0.30] as did adjustment for fasting insulin and fasting glucose [HR = 1.25 (95% CI = 0.88–1.79); P for trend = 0.37].
When we repeated the above analyses using only those with treated type 2 diabetes (as the incident event), the results did not change appreciably (data not shown).
Discussion
In this study of older adults, Lp-PLA2 activity was associated with increased risk of incident type 2 diabetes, apparently related in part to greater insulin resistance in those with higher Lp-PLA2 levels. In contrast, Lp-PLA2 mass was not associated with risk of incident type 2 diabetes.
Several cross-sectional analyses, including in the CHS cohort, have shown that individuals with diabetes have higher Lp-PLA2 mass/activity than do individuals without diabetes (9–11, 20). Although we are not aware of other published prospective studies that have considered the association of Lp-PLA2 with incident type 2 diabetes, one published abstract is available among healthy males aged 45–64 yr in Augsburg, Germany (MONICA/KORA survey). After approximately 13.5 yr of follow-up, there was no significant association of Lp-PLA2 mass with incident type 2 diabetes (activity was not measured) (12). Similarly, in the present study, Lp-PLA2 mass was not associated with incident type 2 diabetes.
It is important to note that the measurement of Lp-PLA2 mass and activity represent, in part, potentially different physiologically aspects of Lp-PLA2 biology, which is highlighted by the modest correlation between these factors (r = 0.50). The mass measure represents Lp-PLA2's mass in intact lipoproteins, whereas the activity assay represents phospholipase activity. Although Lp-PLA2 mass and activity are associated with CVD risk, our findings suggest that total Lp-PLA2 activity in plasma is the more relevant measure as related to diabetes.
The mechanisms that may link Lp-PLA2 activity with incident type 2 diabetes may be through inflammatory activity associated with Lp-PLA2's hydrolysis of oxidized phospholipids that could potentially increase insulin resistance (5, 6). Because inflamed adipose tissue is a forerunner of insulin resistance (21), it is possible that monocyte/macrophage infiltration of the adipose compartment could lead to increased production of Lp-PLA2 locally, wherein oxidative byproducts could foster adipocyte-specific insulin resistance. Alternatively, increased Lp-PLA2 could be a marker of increased adipose tissue inflammation, other metabolic processes such as increased fatty acid flux, and/or adipokine perturbations that underlie the development of insulin resistance and pancreatic β-cell failure (7). Interestingly, there are only modest or weak associations of Lp-PLA2 with nonspecific markers of inflammation (e.g. C-reactive protein) (20, 22, 23). Furthermore, in the present study, associations of Lp-PLA2 activity with incident diabetes remained significant after controlling for white blood cells, fibrinogen and CRP, and BMI. This suggests that Lp-PLA2 reflects diabetogenic pathways independently of these other inflammatory markers or that the latter makers may only be weakly related to the more specific inflammatory pathways by which lysophosphatidylcholine and oxidized nonesterified free fatty acids may act (3, 5).
When we considered HDL in model 3 (Table 3), the risk estimate was attenuated and the P for trend became nonsignificant, as it was when we further considered HOMA-IR as well as fasting insulin and fasting glucose in a sensitivity analysis, suggesting that, along with Lp-PLA2, both may serve as overlapping markers of the pathophysiological process leading to diabetes or, alternatively, that both could act as mediators to the association of Lp-PLA2 with type 2 diabetes (perhaps through the mechanisms discussed above). Lp-PLA2 may be a byproduct of, but also directly contribute to, adipose tissue inflammation. The latter renders fat cells resistant to insulin, with less inhibition of hormone-sensitive lipase, increased triglyceride hydrolysis, and greater catabolism of HDL (7, 24); furthermore, HDL is also a carrier of Lp-PLA2 in plasma and, in contrast to LDL-bound Lp-PLA2, has been postulated to have beneficial (antiinflammatory) effects, which could potentially protect against diabetes (3). With regard to HOMA-IR, our results suggest that insulin resistance may mediate the association between Lp-PLA2 activity and type 2 diabetes; furthermore, both HOMA measures were cross-sectionally associated with Lp-PLA2 activity in individuals without diabetes, supporting the biological plausibility of our findings. Clearly, more work is needed in understanding the potential pathways by which Lp-PLA2 may be associated with insulin resistance and type 2 diabetes.
Limitations of the present study include having only a single measurement of Lp-PLA2, which may have led to an underestimation of the true strength of the observed relationships. Furthermore, we cannot definitively determine from the prospective analysis whether insulin resistance, estimated by HOMA, preceded higher Lp-PLA2 levels or whether it was a consequence. Finally, this is an older population, and these findings may not be generalized to younger populations. The strengths of this study include a large sample size, measures of both Lp-PLA2 mass and activity, extensive covariate availability, and prospective data on incident diabetes with a large number of cases.
Overall, our results show that higher Lp-PLA2 activity is associated with risk of incident type 2 diabetes in an elderly cohort, independent of factors associated with diabetes pathogenesis, including inflammation. Although an association of this magnitude may not lead to changes in diagnostic testing, it does suggest novel and potentially modifiable pathways that may be important in the etiology of type 2 diabetes.
Acknowledgments
The research reported in this article was supported by contracts N01-HC-85239, N01-HC-85079 through N01-HC-85086, N01-HC-35129, N01 HC-15103, N01 HC-55222, N01-HC-75150, N01-HC-45133, and grant HL080295 from the National Heart, Lung, and Blood Institute (NHLBI), with additional contribution from the National Institute of Neurological Disorders and Stroke (NINDS). Additional support was provided through AG-023629, AG-15928, AG-20098, and AG-027058 from the National Institute on Aging (NIA). A full list of principal CHS investigators and institutions can be found at http://www.chs-nhlbi.org/pi.htm. Measurements of Lp-PLA2 were funded by an investigator-initiated grant from Glaxo Smith Kline.
Author contributions were as follows: study concept and design, T.L.N., M.C., K.J.M., and M.L.B.; analysis and interpretation of data, M.L.B., T.L.N., K.J.M., M.C., J.R.K., J.E.H., and C.D.F.; drafting of manuscript, T.L.N. and M.L.B.; critical revision of manuscript for important intellectual content, T.L.N., M.L.B., K.J.M., M.C., J.R.K., J.E.H., and C.D.F.; and statistical analysis, M.L.B.
Disclosure Summary: T.L.N., M.L.B., J.E.H., C.D.F., and K.J.M. have nothing to declare. J.R.K. has received grant support from diaDexus, Inc., to investigate the association of Lp-PLA2 with cardiovascular disease in the STRONG HEART Study, a separate cohort. M.C. has received funding from Glaxo Smith Kline for studies related to Lp-PLA2 epidemiology.
Footnotes
- BMI
- Body mass index
- CHS
- Cardiovascular Health Study
- CI
- confidence interval
- CRP
- C-reactive protein
- CVD
- cardiovascular disease
- HDL
- high-density lipoprotein
- HOMA-B
- homeostatic model assessment of β-cell function
- HOMA-IR
- homeostatic model assessment of insulin resistance
- HR
- hazard ratio
- LDL
- low-density lipoprotein
- Lp-PLA2
- lipoprotein-associated phospholipase A2.
References
- 1. Jenny NS, Solomon C, Cushman M, Tracy RP, Nelson JJ, Psaty BM, Furberg CD. 2010. Lipoprotein-associated phospholipase A2 (Lp-PLA2) and risk of cardiovascular disease in older adults: results from the Cardiovascular Health Study. Atherosclerosis 209:528–532 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Oei HH, van der Meer IM, Hofman A, Koudstaal PJ, Stijnen T, Breteler MM, Witteman JC. 2005. Lipoprotein-associated phospholipase A2 activity is associated with risk of coronary heart disease and ischemic stroke: the Rotterdam Study. Circulation 111:570–575 [DOI] [PubMed] [Google Scholar]
- 3. Zalewski A, Macphee C, Nelson JJ. 2005. Lipoprotein-associated phospholipase A2: a potential therapeutic target for atherosclerosis. Curr Drug Targets Cardiovasc Haematol Disord 5:527–532 [DOI] [PubMed] [Google Scholar]
- 4. Macphee CH, Nelson J, Zalewski A. 2006. Role of lipoprotein-associated phospholipase A2 in atherosclerosis and its potential as a therapeutic target. Curr Opin Pharmacol 6:154–161 [DOI] [PubMed] [Google Scholar]
- 5. Iwase M, Sonoki K, Sasaki N, Ohdo S, Higuchi S, Hattori H, Iida M. 2008. Lysophosphatidylcholine contents in plasma LDL in patients with type 2 diabetes mellitus: relation with lipoprotein-associated phospholipase A2 and effects of simvastatin treatment. Atherosclerosis 196:931–936 [DOI] [PubMed] [Google Scholar]
- 6. Noto H, Chitkara P, Raskin P. 2006. The role of lipoprotein-associated phospholipase A2 in the metabolic syndrome and diabetes. J Diabetes Complications 20:343–348 [DOI] [PubMed] [Google Scholar]
- 7. Rabe K, Lehrke M, Parhofer KG, Broedl UC. 2008. Adipokines and insulin resistance. Mol Med 14:741–751 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Nelson TL, Kamineni A, Psaty B, Cushman M, Jenny NS, Hokanson J, Furberg C, Mukamal KJ. 2011. Lipoprotein-associated phospholipase A2 and future risk of subclinical disease and cardiovascular events in individuals with type 2 diabetes: the Cardiovascular Health Study. Diabetologia 54:329–333 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Serban M, Tanaseanu C, Kosaka T, Vidulescu C, Stoian I, Marta DS, Tanaseanu S, Moldoveanu E. 2002. Significance of platelet-activating factor acetylhydrolase in patients with non-insulin-dependent (type 2) diabetes mellitus. J Cell Mol Med 6:643–647 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kudolo GB, Bressler P, DeFronzo RA. 1997. Plasma PAF acetylhydrolase in non-insulin dependent diabetes mellitus and obesity: effect of hyperinsulinemia and lovastatin treatment. J Lipid Mediat Cell Signal 17:97–113 [DOI] [PubMed] [Google Scholar]
- 11. Basu A, Jensen MD, McCann F, Nandy D, Mukhopadhyay D, McConnell JP, Rizza RA. 2007. Lack of an effect of pioglitazone or glipizide on lipoprotein-associated phopholipase A2 in type 2 diabetes. Endocr Pract 13:147–152 [DOI] [PubMed] [Google Scholar]
- 12. Khuseyinova N, Baumert J, Trischler G, Meisinger C, Loewel H, Koenig W. 2006. Lipoprotein-associated phospholipase A2 does not predict risk of incident type 2 diabetes mellitus in apparently healthy middle-aged men: a more specific marker for vascular inflammation? Circulation 114:869–870 (Abstract) [Google Scholar]
- 13. Fried LP, Borhani NO, Enright P, Furberg CD, Gardin JM, Kronmal RA, Kuller LH, Manolio TA, Mittelmark MB, Newman A, et al. 1991. The Cardiovascular Health Study: design and rationale. Ann Epidemiol 1:263–276 [DOI] [PubMed] [Google Scholar]
- 14. Tell GS, Fried LP, Hermanson B, Manolio TA, Newman AB, Borhani NO. 1993. Recruitment of adults 65 years and older as participants in the cardiovascular health study. Ann Epidemiol 3:358–366 [DOI] [PubMed] [Google Scholar]
- 15. Cushman M, Cornell ES, Howard PR, Bovill EG, Tracy RP. 1995. Laboratory methods and quality assurance in the Cardiovascular Health Study. Clin Chem 41:264–270 [PubMed] [Google Scholar]
- 16. Psaty BM, Lee M, Savage PJ, Rutan GH, German PS, Lyles M. 1992. Assessing the use of medications in the elderly: method and intial results in the Cardiovascular Health Study. J Clin Epidemiol 45:683–692 [DOI] [PubMed] [Google Scholar]
- 17. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. 1985. Homeostasis model assessment: insulin resistance and β-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28:412–419 [DOI] [PubMed] [Google Scholar]
- 18. Lumley T, Kronmal R, Ma S. 2006. Relative risk regression in medical research: models, contrasts, estimators and algorithms. Seattle, WA: University of Washington [Google Scholar]
- 19. Arnold AM, Kronmal RA. 2003. Multiple imputation of baseline data in the cardiovascular health study. Am J Epidemiol 157:74–84 [DOI] [PubMed] [Google Scholar]
- 20. Furberg CD, Nelson JJ, Solomon C, Cushman M, Jenny NS, Psaty BM. 2008. Distribution and correlates of lipoportein-associated phospholipase A2 in an elderly cohort: the cardiovascular health study. J Am Geriatr Soc 56:792–799 [DOI] [PubMed] [Google Scholar]
- 21. Gustafson B, Hammarstedt A, Andersson CX, Smith U. 2007. Inflamed adipose tissue: a culprit underlying the metabolic syndrome and atherosclerosis. Arterioscler Thromb Vasc Biol 27:2276–2283 [DOI] [PubMed] [Google Scholar]
- 22. Winkler K, Winkelmann BR, Scharnagl H, Hoffmann MM, Grawitz AB, Nauck M, Böhm BO, März W. 2005. Platelet-activating factor acetylhydrolase activity indicates angiographic coronary artery disease independently of systemic inflammation and other risk factors: the Ludwigshafen Risk and Cardiovascular Health Study. Circulation 111:980–987 [DOI] [PubMed] [Google Scholar]
- 23. Ballantyne CM, Hoogeveen RC, Bang H, Coresh J, Folsom AR, Chambless LE, Myerson M, Wu KK, Sharrett AR, Boerwinkle E. 2005. Lipoprotein-associated phospholipase A2, high-sensitivity C-reactive protein, and risk for incident ischemic stroke in middle-aged men and women in the Atherosclerosis Risk in Communities (ARIC) study. Arch Intern Med 165:2479–2484 [DOI] [PubMed] [Google Scholar]
- 24. Gauther MS, Ruderman NB. 2010. Adipose tissue inflammation and insulin resistnace: all obese humans are not created equal. Biochem J 430:e1–e4 [DOI] [PubMed] [Google Scholar]
