Tea and Coffee Consumption and MRSA Nasal Carriage

Study Design and Participants

We analyzed data from the 2003–2004 National Health and Nutrition Examination Survey (NHANES) sample,¹⁶ a nationally representative sample of the non-institutionalized population of the United States.¹⁷ The NHANES uses a complex survey design that allows for the determination of population estimates. To enable these estimates, the survey uses sampling weights that are calculated by taking into account the unequal probabilities of selection resulting from sample design, nonresponse, and planned oversampling of minorities.

The NHANES sampling procedure consists of 4 stages. Stage 1 entails selection of primary sampling units (PSUs), which are mostly single counties or, in a few cases, groups of contiguous counties with probability proportional to a measure of size. In stage 2, the PSUs are divided into segments (generally city blocks or their equivalent). In stage 3, households within each segment are listed and a sample is randomly drawn. In geographic areas where the proportion of age, ethnic, or income groups selected for oversampling is high, the probability of selection for those groups is greater than in other areas. In stage 4, individuals are chosen to participate in NHANES from a list of all persons residing in selected households. The individuals are drawn at random within designated domains of age, sex, and race/ethnicity. Each individual is assigned a sample weight, a measure of the number of people in the population represented by that individual in NHANES, reflecting the unequal probability of selection, adjustment for nonresponse, and adjustment to individual population controls. The 30 PSUs included in the 2003–2004 NHANES were selected from a frame of all US counties, using the 2000 census data.

More information on the methods of the 2003–2004 NHANES survey, including laboratory assessments, can be found elsewhere.¹⁶ Nasal carriage of S aureus was assessed in all participants aged 1 year or older, but many of the other variables were not included for individuals younger than age 2. Consequently, we included in our study only participants aged 2 years old or older.

Assessment of MRSA Nasal Carriage

Nasal cultures were collected and plated on mannitol salt agar (MSA). The MSA plates were incubated at 37°C for 48 hours. Mannitol-fermenting colonies were selected and subcultured to plates with trypticase soy agar containing 5% sheep blood, that is, blood agar plates (BAPs), and incubated at 37°C overnight. MSA plates with little or no growth were reincubated at 37°C overnight, and plates with non–mannitol-fermenting growth were held at room temperature. The plates were reexamined the next day, and any yellow or gold colonies were subcultured to BAPs.

Overnight cultures on BAPs were first screened using Staphaurex, a rapid latex kit for the identification of S aureus (Remel Laboratories, Lenexa, Kansas). A tube coagulase test using rabbit plasma with EDTA was then performed both on Staphaurex-negative isolates from BAPs with morphology consistent with S aureus and on Staphaurex-positive isolates with morphology inconsistent with S aureus. Staphaurex-positive isolates and Staphaurex-negative, tube coagulase–positive isolates were identified as S aureus and saved for further testing.

Determination of Methicillin Resistance

S aureus isolates were screened for methicillin resistance by the disk diffusion method. Overnight cultures from BAPs were plated on Mueller-Hinton agar, and a 1-μg oxacillin disk was placed on the inoculated plate. Zone diameters were measured and recorded after a 24-hour incubation at 37°C; on the basis of the results, isolates were classified as sensitive (≥13 mm), intermediate (11–12 mm), or resistant (≤10 mm). For the purposes of this study, S aureus isolates with a zone diameter of 10 mm or less were considered MRSA and all other Staphylococcus isolates were considered methicillin sensitive (MSSA).

Beverage Consumption

Beverages evaluated were hot tea, iced tea, coffee, soda consumed during the summer, and soda consumed during the remainder of the year as ascertained with 6 questions. The first question was, “How many cups of hot tea, caffeinated or decaffeinated, did you drink (over the past 12 months)?” The second question was, “How many glasses of iced tea, caffeinated or decaffeinated, did you drink (over the past 12 months)?” The third question was “How many cups of coffee, caffeinated or decaffeinated, did you drink (over the past 12 months)?” Responses to these 3 questions were categorized as (1) none or (2) less than 1 cup a month or more for coffee, tea, and iced tea. Participants were also asked, “Over the past 12 months, did you drink soft drinks, soda, or pop?” Those who answered yes were then asked, “How often did you drink soft drinks, soda, or pop in the summer?” and subsequently asked, “How often did you drink soft drinks, soda, or pop during the rest of the year?” Because of the very small number of individuals who did not consume any soft drinks, answers for soft drink consumers were classified as (1) never to 2 to 3 times per month or (2) 1 to 6 times per week or more.

Control Variables

We evaluated numerous variables previously associated with MRSA nasal carriage as potential confounding variables in the relationship between MRSA carriage and beverage consumption. They included age, race, sex, poverty status, health status, recent hospital stay, and recent antibiotic use.

Age was categorized into groups (2–6, 7–19, 20–64, and ≥65 years old). Race/ethnicity was classified as non-Hispanic white, non-Hispanic black, or Hispanic. We did not include individuals categorized as Other Race or Multiracial in our study because there were too few with MRSA (unweighted n = 4). Poverty status was determined using the poverty-income ratio. This ratio is calculated by dividing a family’s income by a poverty threshold income that is based on family size.¹⁸ The poverty threshold is determined each year by the US Office of Management and Budget. Poverty-income ratios ranged up to 5, with values of less than 1 indicating poverty. Current health status was based on self-report and was classified as either (1) excellent, very good, or good, or (2) fair or poor. Recent hospitalization was determined by participants’ response to the question, “During the past 12 months were you a patient in a hospital overnight? Do not include an overnight stay in the emergency room.”

Recent antibiotic use was evaluated in the NHANES by examining the individual’s prescription medications. Participants were asked to bring the containers for all the prescription medications taken during the past month and to report any medications taken for which the container was not available. These medications were then matched by trained survey interviewers to an annually updated comprehensive database of all prescription drugs in the US market.

Missing Data

For the purposes of this analysis, we excluded individuals who were missing data on any variable. A total of 3,172 individuals were missing data. Of these individuals, the majority were missing data on coffee consumption (2,654), tea consumption (2,658) or poverty-income ratio (449).

Analysis

We used SUDAAN software (Research Triangle Institute, Research Triangle Park, North Carolina) for all analyses to account for the complex sample design used in NHANES. The analyses incorporated both the stratification and clustering aspects of the sampling design. The proper weighting procedures include adjustments for basic probability of selection and nonresponse. Because minorities were oversampled, and the sampling design was complex, we used sampling weights provided by the NHANES to compute population estimates based on weighted parameter estimates and standard errors.

We computed descriptive statistics for the study sample. Because there were relatively small numbers of MRSA carriers within each beverage consumption group, it was necessary to use a binary categorization (drinker or nondrinker) for the beverage groups to gain a standard error of less than 30%, conforming to recommendations of the National Center for Health Statistics (which conducts NHANES) for small sample sizes. We performed a χ² analysis to determine that none of the standard errors exceeded 30% on population estimates, as recommended by the Center. We computed unadjusted and adjusted logistic regression models to determine the odds ratios (ORs) and 95% confidence intervals (CIs) for MRSA nasal carriage according to beverage consumption. In addition to hot tea and coffee, iced tea and soft drink consumption were examined to determine if the antimicrobial properties of hot tea and coffee differed from that of other types of beverages. The adjusted models controlled for age, race/ethnicity, sex, poverty-income ratio, current health status, hospitalization in the past year, and antibiotic use in the past month. A statistical level of P <.05 was considered significant for all analyses.