Impact of a Municipal Policy Restricting Trans Fatty Acid Use in New York City Restaurants on Serum Trans Fatty Acid Levels in Adults

Abstract

Objectives. To estimate the impact of the 2006 policy restricting use of trans fatty acids (TFAs) in New York City restaurants on change in serum TFA concentrations in New York City adults.

Methods. Two cross-sectional population-based New York City Health and Nutrition Examination Surveys conducted in 2004 (n = 212) and 2013–2014 (n = 247) provided estimates of serum TFA exposure and average frequency of weekly restaurant meals. We estimated the geometric mean of the sum of serum TFAs by year and restaurant meal frequency by using linear regression.

Results. Among those who ate less than 1 restaurant meal per week, geometric mean of the sum of serum TFAs declined 51.1% (95% confidence interval [CI] = 42.7, 58.3)—from 44.6 (95% CI = 39.7, 50.1) to 21.8 (95% CI = 19.3, 24.5) micromoles per liter. The decline in the geometric mean was greater (P for interaction = .04) among those who ate 4 or more restaurant meals per week: 61.6% (95% CI = 55.8, 66.7) or from 54.6 (95% CI = 49.3, 60.5) to 21.0 (95% CI = 18.9, 23.3) micromoles per liter.

Conclusions. New York City adult serum TFA concentrations declined between 2004 and 2014. The indication of greater decline in serum TFAs among those eating restaurant meals more frequently suggests that the municipal restriction on TFA use was effective in reducing TFA exposure.

Public Health Implications. Local policies focused on restaurants can promote nutritional improvements.

Heart disease is the leading cause of death in the United States and New York City.¹ Trans fatty acid (TFA) intake adversely alters the cholesterol profile (increasing low-density lipoprotein while lowering high-density lipoprotein) and is associated with increased risk of coronary heart disease.² In 2003, the US Food and Drug Administration (FDA) mandated that the TFA content be included on nutrition labels of packaged foods by 2006.³ However, this regulation did not apply to restaurant foods, leaving restaurant diners at a disadvantage. In 2006, the New York City Board of Health restricted TFA use in all New York City restaurants.⁴

We estimated the change in serum TFA levels in the New York City adult population after policy implementation, overall and by weekly restaurant meal frequency (WRMF). We hypothesized that there would be population-wide declines in mean serum TFA levels consistent with national trends, but larger declines would be observed among frequent restaurant diners compared with those who hardly dined out.

METHODS

The New York City Health and Nutrition Examination Survey (HANES) is a population-based, cross-sectional study of New York City resident adults aged 20 years and older conducted in 2004⁵ and 2013–2014.⁶ Participants responded to questions about health status, behavior, and sociodemographics. Blood specimens were provided by 1811 out of 1999 participants in 2004, and 1201 out of 1527 participants in 2013–2014. A random selection of 250 specimens was drawn from fasting participants in each survey year. A subset of these (3 from 2014 and 38 from 2004) could not be analyzed as they failed quality control.

Participants were asked: “On average, how many times per week do you eat meals that were prepared in a restaurant? Please include eat-in restaurants, carryout restaurants and restaurants that deliver food to your house. ‘Meals’ mean more than a beverage or snack food like candy bars or bag of chips.”

In 2006, New York City restricted the use of TFAs in all New York City restaurants. Foods stored, used, or served could contain no more than 0.5 grams of TFAs per serving after July 2008 if partially hydrogenated vegetable oil was present.⁴

Serum TFA levels represent dietary TFAs in the medium term (i.e., weeks to months). The main outcome was the sum of TFAs: trans-vaccenic acid, elaidic acid, palmitelaidic acid, and linoelaidic acid.

We measured TFAs (micromoles per liter) in aliquots of stored sera by using an isotope dilution–gas chromatography–negative chemical ionization–mass spectrometry method with a 7890/5975C GC/MSD Agilent Technologies (Santa Clara, CA) instrument. Separation was achieved with a 200 meter by 250 micrometer by 0.25 micrometer Select-FAME column (Agilent Technologies, Santa Clara, CA). This method effectively separates TFAs from other fatty acids with limits of detection for TFAs ranging from 0.02 to 0.43 micromoles per liter.⁷

We natural log-transformed serum TFA concentrations to improve normality. We exponentiated the arithmetic means of logged values and 95% confidence intervals (CIs) and presented them as geometric means (GMs). We estimated the mean sum of serum TFAs by survey year. We compared geometric means before and after the TFA policy with the t test and reported differences as percentages.

We estimated mean TFA across categories of less than 1, 1, 2 to 3, and 4 or more WRMF. We chose categories to smooth random variation in responses. Exploratory data analysis suggested that mean TFA changed with WRMF in a linear manner, warranting a continuous specification of WRMF (scoring each category 0 through 3). We modeled the association of continuous WRMF with sum of serum TFAs by using a linear regression model that included year, WRMF, and the interaction between survey year and WRMF. The coefficient of the interaction term estimates how the association of WRMF with the sum of serum TFAs varied across surveys. The exponentiated coefficient is the change in this association expressed as a ratio, reported here as a percentage.

We assessed potential confounding by age, gender, education, income, neighborhood poverty, and having received any doctor’s orders to control weight, exercise, reduce dietary sodium, or decrease fat intake. We used a χ² test of the change in the distribution of these variables across WRMF by year (using threshold P < .20). Education was the only predictor that changed over time. However, we also included age (as age + age-squared) and gender in the model on the basis of prior knowledge of their associations with sum of serum TFAs and dining out.

We weighted all results to the New York City adult population. We conducted analyses with SAS Enterprise Guide version 7.1 (SAS Institute, Cary, NC) and SUDAAN version 11.0.1 (Research Triangle Institute, Research Triangle Park, NC) to apply sample weights.

RESULTS

The proportion of men (47% vs 46%) and the average age (46 vs 45 years) were similar in 2004 and 2013–2014 study populations (not shown). The WRMF was similar across years: the 2004 survey prevalence of less than 1, 1, 2 to 3, and 4 or more WRMF was 31%, 22%, 21%, and 27%, respectively, and these proportions were 28%, 25%, 27%, and 21%, respectively, in 2014 (χ² P = .40). The adult geometric mean of the sum of serum TFAs declined 56.6% (95% CI = 51.9, 60.8; Table A, available as a supplement to the online version of this article at http://www.ajph.org). Declines in the geometric means of individual TFA values ranged from 38.0% (95% CI = 32.5, 43.1) in linoelaidic acid to 61.7% (95% CI = 57.3, 65.7) in elaidic acid.

The linear regression coefficient for year, −0.716 (95% CI = –0.875, −0.557), suggests a decrease in the geometric mean of the sum of serum TFAs over time (Table B, available as a supplement to the online version of this article at http://www.ajph.org). Before the policy, sum of serum TFAs was estimated to increase with higher WRMF, coefficient = 0.068 (95% CI = 0.013, 0.122). The coefficient for the interaction between WRMF and survey year, −0.080 (95% CI = –0.157, 0.004), suggests that the positive association between WRMF and geometric mean of the sum of serum TFAs was attenuated after policy implementation (P = .04).

The model-generated estimates of the geometric mean of the sum of serum TFAs by WRMF and year (Figure 1) show WRMF was positively associated with the geometric mean in 2004, but not in 2013–2014. Among adults who ate 4 or more restaurant meals per week, the model estimated a 61.6% (95% CI = 55.8, 66.7) change in geometric mean of the sum of serum TFAs. This exceeded the estimated change in the geometric mean among adults who ate an average of less than 1 WRMF, 51.1% (95% CI = 42.7, 58.3).

FIGURE 1—

Geometric Mean of the Sum of Serum Trans Fatty Acids by Weekly Restaurant Meal Frequency Before (2004) and After (2013–2014) New York City Trans Fatty Acid Policy

Note. TFA = trans fatty acid; ΣTFA = geometric mean sum of elaidic, trans-vaccenic, palmitelaidic, and linoelaidic serum trans fatty acids; NYC = New York City. Data are based on the covariate-adjusted linear regression of ΣTFA on survey year, weekly restaurant meal frequency, and their interaction (Table B, available as a supplement to the online version of this article at http://www.ajph.org). Error bars are 95% confidence intervals.

DISCUSSION

We observed population-wide declines in New York City adult sum of serum TFAs, with greater reductions among frequent restaurant diners (61.6%) relative to those who hardly dined out (51.1%). This finding is consistent with reductions in TFA content of New York City restaurant foods measured before and after the policy.^8,9 The additional decline is expected to have clinical impact as just a 2% increase in calories from TFA has been associated with a 23% increase in coronary heart disease risk.² Further evidence for a clinical impact comes from ecologic studies comparing New York state jurisdictions with and without such TFA policies; the former have lower rates of myocardial infarction- and stroke-related hospitalizations¹⁰ and cardiovascular disease mortality.¹¹ Given the proliferation of policies to restrict restaurant use of TFAs across the United States, these findings suggest opportunity for broad impact. We expect that the FDA-mandated removal of partially hydrogenated oils from the US food supply, which went into effect on June 18, 2018, will result in a further decrease in TFA exposure nationwide.

The overall 57% decline in New York City adult serum TFAs between 2004 and 2014 is comparable in magnitude to the 54% decline observed nationally between 1999–2000 and 2009–2010.¹² The 2006 FDA requirement of TFA nutrition labeling, industry reformulation to reduce TFA in packaged food, and increased population avoidance of TFAs have likely contributed to the overall declines observed nationally and locally in New York City. Nevertheless, the differential decline we reported among high restaurant frequenters suggests an independent impact of the TFA restriction.

This study is limited by its small sample size, which reduced certainty around estimates and precluded explorations of subgroup differences. In addition, self-reported WRMF is unreliable. The definition of a “meal” is subjective and may represent a range of potential TFA exposure. The reported policy impact is specific to New York City patterns of frequenting restaurants and the healthfulness of local menu choices in 2004; the impact of a similar policy might be greater in jurisdictions with a greater number of fast-food chains where individuals with less healthy lifestyles dine out more frequently.

Study strengths include the use of a population-based HANES, objectively measured biomarkers, and an efficient statistical approach.

PUBLIC HEALTH IMPLICATIONS

Policy aimed at sustainable improvements in the nutritional content of restaurant foods can be a means to measurably improve public health.