Association Between Chemosensory Dysfunction and Diet Quality in United States Adults

Abstract

Background

Evidence suggests chemosensory dysfunction (CSD) patients have altered diet, but population-level evidence assessing diet quality in CSD patients is lacking.

Objective

We examined the association between CSD and diet quality in a representative sample of United States adults.

Methods

This cross-sectional study included 2831 adults aged greater than 40 years from the 2013-2014 National Health and Nutrition Examination Survey who completed the taste/smell questionnaire and examination. Mean nutrient intake in subjects with self-reported olfactory/gustatory dysfunction (sOD/sGD) and measured olfactory/gustatory dysfunction (mOD/mGD) were compared to those without CSD using univariate Wilcoxon rank-sum tests. The Healthy Eating Index (HEI), a validated measure of diet quality, was calculated. The proportion of subjects with CSD with bottom-quartile HEI was compared to those without CSD using multivariate logistic regression, adjusting for demographic and socioeconomic covariates.

Results

The population-weighted prevalence of sOD, sGD, mOD, and mGD was 20.1%, 14.4%, 15.9% and 25.6%, respectively. Subjects with mOD had lower mean intake of total calories, total fat, protein, sodium, and potassium compared to normal subjects (1873.4 ± 49.6 vs 2010.2 ± 24.2 kcal, 72.3 ± 2.7 vs 78.6 ± 1.0 gm, 74.0 ± 2.5 vs 80.4 ± 0.6 gm, 3122 ± 97.2 vs 3353.2 ± 37.0 mg, 2509.8 ± 69.8 vs 2684.7 ± 26.1 mg, P < 0.05 respectively). When controlling for sociodemographic factors and comorbidities, subjects with sOD were more likely to have bottom-quartile HEI compared to normal subjects (OR 1.33, 95% CI 1.04–1.70).

Conclusions

This population-level study suggests an association between poor diet quality and variation in dietary intake in patients with CSD, which warrants further investigation and suggests the possible need for nutritional counseling for CSD patients.

Keywords

chemosensory dysfunction diet gustation NHANES nutrition olfaction smell taste

Introduction

Chemosensory dysfunction (CSD) refers to a broad range of alterations in the chemical senses of smell, taste and trigeminal function that are associated with quality of life disruption, and have been linked to vitamin D deficiency, diabetes and depression in population-level studies.^1–3 Though often overlooked, CSD is common, with population level prevalence of self-reported olfactory dysfunction (sOD) ranging from 10.6% to 23.0%, and self-reported gustatory dysfunction (sGD) ranging from 5.3% to 19.0%, both of which invariably increase with age.^2,4–6 Chemosensory loss or distortion has been shown to have widely variable effects on weight change, dietary intake, appetite, and food tolerability in small-scale studies, although results have been inconsistent.^7–12 Indeed, a recent study from Germany suggested heterogeneity in intake of components of a Western-style diet in a group of 222 subjects with various causes of olfactory dysfunction.¹³ Moreover, recent studies have suggested that CSD, particularly in the aging population, may be associated with decreased appetite, altered perception of foods, poorer nutritional intake and diet quality, and decreased immune system function.^14–16 Additionally, a previous population-based study assessing CSD and diet in patients with diabetes mellitus suggested an association between lower caloric intake and olfactory dysfunction.¹⁷

To date, there have been no population-based studies specifically assessing the association between CSD and diet quality in the general population of United States (US) adults. The National Health and Nutrition Examination Survey (NHANES) is an ongoing population-based survey conducted by the US Centers for Disease Control and Prevention, with the purpose of estimating disease prevalence. This publicly‐available series of questionnaires and examinations focusing on health and nutrition uses a complex, multistage, probability sampling design to produce a nationally representative sample of about 10,000 persons per 2‐year cycle.¹⁸ The 2013-2014 NHANES included a subjective and objective taste and smell assessment, as well as a dietary intake interview. The purpose of this study is to analyze the relationship of both self-reported and measured CSD with diet quality and specific dietary changes in a representative sample of US adults.

Materials and Methods

Data Source

The NHANES uses a complex stratified multistage sampling design to select a nationally representative sample of non-institutionalized US civilians to participate in a series of comprehensive health-related interviews and examinations every 2 years. Recruitment and testing are performed across the US, with 15 different counties visited yearly. NHANES purposefully oversamples larger numbers of certain subgroups of particular public health interest including Mexican Americans, African Americans and persons 60 years of age or older, in order to increase the reliability and precision of the health-related information gathered about these demographic groups. The NHANES protocols were approved by the National Center for Health Statistics research ethics review board and informed consent was obtained from all participants.

Demographics

Participants’ age, gender, race, education, income/poverty ratio, and smoking status were collected in a demographic questionnaire. NHANES categorizes race and ethnicity as follows: Mexican American, Other Hispanic, Non-Hispanic Black, Non-Hispanic White, and Non-Hispanic Asian or Other Race including Multi-Racial. Participants’ body mass indices (BMI) were obtained from the body measurements examination. NHANES categorizes education level as follows: less than 9th grade, 9-11th grade or 12th grade with no diploma, high school graduate/GED or equivalent, some college or associates in arts (AA) degree, and college graduate or above. A full list of variable names and codes is available in Supplementary Table 1.

Subjective Chemosensory Dysfunction

Participants aged 40 years or older were administered a Taste & Smell Questionnaire (CSQ) which contains a series of “yes/no” questions related to chemosensory function. Questions included problems with ability to smell in the past year, smelling an unpleasant/bad/burning odor when nothing is there, problems tasting sweet, sour, salty or bitter foods and drinks in the past year, and having a taste or other sensation in the mouth that does not go away. Additional questions regarding changes to smell since age 25 included the following: change in ability to smell (better/worse/no change), change in ability to taste salty/sweet/sour/bitter (better/worse/no change), diminished ability to taste food flavors such as chocolate, vanilla, or strawberry (yes/no).¹⁹ Self-reported olfactory dysfunction (sOD) was defined as a positive answer to any of the following questions: 1) “Have you had problem with smell in the past 12 months?”; 2) “Have you had a change in the ability to smell since age 25?” and 3) “Do you have phantom smells?”⁶ Self-reported gustatory dysfunction (sGD) was defined as a positive answer to any of the following questions: 1) “Have you had problem with taste in the past 12 months?”; 2) “Is your ability to taste food flavors such as chocolate, vanilla or strawberry as good as when you were 25 years old?”; and 3) “Over the last 12 months, have you had a taste or other sensation in your mouth that does not go away?

Objective Chemosensory Dysfunction

The 2013-2014 NHANES contained objective measures of CSD that were adapted from the National Institutes of Health (NIH) Toolbox project.^20,21 Olfaction was objectively assessed with the 8-item Pocket Smell Test (PST, Sensonics, Inc., Haddon Heights, NJ), which is based on the 40-item University of Pennsylvania Smell Identification Test (UPSIT),²² and consists of 4 food related odors (chocolate, strawberry, grape, onion), 2 warning odors (smoke, natural gas), and 2 common household odors (leather, soap).^5,23 For each item, microencapsulated odorants were placed on test strips, technicians scratched the test strips with a plastic stylus in a standardized fashion, and participants were asked to identify each odorant from a set of 4 choices. Participants were required to make a choice even if no odor was perceived.^5,23 Participants were scored on a 0 to 5 scale and considered to have measured olfactory dysfunction (mOD) if they identified 3 or more out of 8 odorants incorrectly.^5,23

Gustation was objectively assessed with a previously validated general Labeled Magnitude Scale (gLMS)²⁴ to measure the intensity of 1 mM quinine and 1 M NaCl placed on the anterior tongue with a cotton-tipped applicator.¹⁹ After regional taste testing, participants sampled 3 tastants (1 M NaCl, 1 mM quinine, 0.32 M NaCl) by filling their mouth with 10 ml of tastant solution, swishing for 5 seconds, then expectorating into a sink.¹⁹ Between tastants, participants rinsed their mouths with deionized water to remove any residual stimulus. Additional prototype tastants were also administered in similar fashion (1 M sucrose for sweet, 32 mM citrate for sour, and 3.2 mM propylthiouracil for bitter).¹⁹ For this analysis, measured gustatory dysfunction (mGD) was defined as inability to correctly identify 1 M or 0.32 M NaCl as salty or 1 mM quinine as bitter. Objective gustatory and olfactory testing in NHANES were previously noted to have moderate-to-good test-retest reliability.¹⁹

Dietary Intake Interview and Healthy Eating Index

The 2013-2014 NHANES contained a dietary intake interview which assessed dietary intake on three non-consecutive 24-hour periods, and guidance was provided by a registered dietitian regarding keeping a detailed food record.²⁵ In this study, the “Total Nutrient Intakes” files from NHANES 2013-2014 were used. Nutrient data and dietary data-specific sample weights for each dietary interview day were averaged for all survey-adjusted statistical analyses. The Healthy Eating Index (HEI), a previously validated measure of conformance to federal dietary guidance and diet quality developed by the US Department of Agriculture (USDA), was calculated.²⁶ The 2015 HEI contains 9 adequacy components (total fruit, whole fruit, total vegetables, beans/greens, whole grains, dairy, total proteins, seafood/plant proteins, fatty acids) and 4 moderation components (refined grains, sodium, added sugars, saturated fats). For adequacy components, higher scores reflect higher intakes, since these dietary elements are considered desirable. For moderation components, higher scores reflect lower intakes, since there are recommended limits to consumption of these dietary elements. Overall, higher HEI aligns better with current dietary recommendations.²⁷ For this analysis, the HEI was categorized into 4 quartiles, with the lowest quartile representing the most nutritionally poor diet. Investigations of diet quality were therefore conducted to assess for association with the lowest quartile HEI.

Covariates

Potential confounders in this analysis included demographic and anthropometric factors including age, gender, race, smoking status and BMI; socioeconomic factors including education level and income/poverty ratio; and comorbid conditions such as diabetes, depression, persistent dry mouth, persistent cold/flu, frequent nasal congestion, history of 2 or more sinus infections, history of broken nose/serious injury to the face/skull, and history of head injury/loss of consciousness. These sociodemographic factors and comorbidities were included as covariates in the multivariable logistic regression. A full list of variable names and codes is available in Supplementary Table 1.

Study Population

The study population for this analysis included subjects from the 2013-2014 NHANES cycle aged 40 years and older at the time of NHANES participation who participated in the CSQ, olfactory and gustatory examinations, and dietary interview. The taste and smell questionnaire was only offered to participants 40 years and older, so there was insufficient data to include younger participants in this analysis.

Data Analysis

Categorical variables were presented as frequency and weighted percentages, and continuous variables were presented as weighted means and standard deviations (SD). A survey-adjusted Shapiro-Wilk Test was used to assess continuous variables for normality. Categorical covariates were compared between participant groups using survey-adjusted Pearson’s chi-squared test with Yates’ continuity correction when needed, and continuous variables were compared using a survey-adjusted Wilcoxon rank-sum test. Survey-adjusted logistic regressions were performed to compare dietary intake and dietary quality among subjects from each CSD subgroup (sOD, mOD, sGD, mGD) with a healthy control group, defined as having no CSD. The Bonferroni correction was not applied as there was no need to adjust for multiple comparisons. Further, although the Bonferroni correction could possibly reduce the chance of type I error, it does so at the risk of increasing type II error.^28,29 Multivariable logistic regression models were adjusted for age, gender, BMI, race, education, smoking, income/poverty ratio, and history of comorbidities and head trauma. Results are presented as odds ratios (OR) and 95% confidence intervals (CI). A P value of < 0.05 was considered as statistically significant. All statistical analysis was performed using R software and data packages (version 3.6.1; R Foundation, Vienna, Austria). The ‘survey’ package was used to account for complex survey design weights in all calculations. The ‘mice’ package was used for random forest imputation of non-index variables missing between 8 and 12% of data.

Results

The 2013-2014 NHANES yielded 2,831 subjects aged 40 years and older who completed the CSQ, measured olfactory and gustatory examinations, and the dietary interview. There were 570 (20.1%) subjects with sOD and 409 (14.4%) subjects with sGD. There were 451 (15.9%) subjects with mOD and 726 (25.6%) subjects with mGD. Of those subjects with sOD, 124 (21.8%) also had mOD. Of those subjects with sGD, 104 (25.4%) also had mGD. Among the 451 subjects with mOD, 124 (27.4%) also had sOD. Among the 726 subjects with mGD, 104 (14.3%) also had sGD. This distribution is show in Figure 1. The population-weighted mean age was 58.1 years, 53.2% of the population was female, and 73.7% of the population was non-Hispanic white. Overall population characteristics and population characteristics of the sOD, sGD, mOD, and mGD subgroups are summarized in Tables 1 and 2.

Table 1.

Cohort Characteristics by Self-Reported Chemosensory Dysfunction.

	All Subjects	Self-reported GD		Self-reported OD
Characteristic	N = 2,831	N = 409	P	N = 570	P
Age	58.1 ± 0.3	57.6 ± 0.7	0.492	58.2 ± 0.6	0.866
Body mass index (BMI) (kg/m²)	29.6 ± 0.2	30.9 ± 0.6	0.012	30.7 ± 0.4	0.005
Gender
Male	1326 (46.8)	157 (41.2)		251 (45.9)
Female	1505 (53.2)	252 (58.8)	0.072	319 (54.1)	0.644
Race
Mexican American	356 (6.3)	73 (8.8)		70 (6.3)
Other Hispanic	249 (4.2)	34 (3.7)		47 (3.9)
Non-Hispanic White	1348 (73.7)	192 (72.5)		310 (77.0)
Non-Hispanic Black	580 (10.1)	71 (9.0)		94 (8.1)
Non-Hispanic Asian	239 (3.6)	24 (2.7)		29 (2.1)
Other race—including multi-racial	59 (2.1)	15 (3.2)	0.053	20 (2.6)	0.128
Education level
Less than 9th grade	209 (4.1)	44 (5.5)		34 (3.1)
9–11th grade or 12 grade with no diploma	362 (9.6)	68 (13.7)		71 (8.5)
High school graduate/GED or equivalent	640 (21.4)	80 (21.5)		136 (22.5)
Some college or AA degree	854 (31.7)	138 (33.5)		199 (38.7)
College graduate or above	766 (33.2)	79 (25.8)	0.072	130 (27.2)	0.008
Ratio of family income to poverty	3.2 ± 0.1	2.9 ± 0.2	0.033	3.0 ± 0.2	0.037
Current smoker	482 (16.1)	89 (21.8)	<0.001	123 (21.2)	0.001
Persistent dry mouth in past 12 months	430 (13.7)	158 (34.5)	<0.001	161 (25.1)	<0.001
Persistent cold/flu in past 12 months	185 (6.4)	69 (17.9)	<0.001	69 (12.4)	<0.001
Frequent nasal congestion in past 12 months	875 (31.1)	183 (44.7)	0.005	246 (42.5)	0.002
Ever broke nose/serious injury to the face/skull	421 (17.7)	86 (23.7)	0.040	120 (23.3)	0.005
Ever had head injury/loss of consciousness	407 (16.7)	78 (20.3)	0.085	126 (23.9)	<0.001
Ever had two or more sinus infections	1032 (44.7)	183 (53.6)	0.023	271 (54.4)	0.003

*Data are presented as n (weighted % of participants) for categorical variables and weighted mean ± SE for continuous variables.

Survey-weighted Wilcoxon rank sum test used for continuous variables.

Survey-weighted Chi square test used for categorical variables.

Table 2.

Cohort Characteristics by Measured Chemosensory Dysfunction.

	All Subjects	Measured GD*		Measured OD**
Characteristic	N = 2,831	N = 726	P	N = 451	P
Age	58.1 ± 0.3	57.5 ± 0.7	0.377	64.5 ± 0.8	<0.001
Body mass index (BMI) (kg/m²)	29.6 ± 0.2	29.3 ± 0.5	0.422	30.0 ± 0.5	0.612
Gender
Male	1326 (46.8)	347 (47.3)		261 (56.3)
Female	1505 (53.2)	379 (52.7)	0.814	190 (43.7)	0.009
Race
Mexican American	356 (6.3)	83 (5.5)		55 (7.3)
Other Hispanic	249 (4.2)	56 (3.4)		45 (5.1)
Non-Hispanic White	1348 (73.7)	359 (75.3)		194 (68.2)
Non-Hispanic Black	580 (10.1)	167 (11.7)		106 (12.9)
Non-Hispanic Asian	239 (3.6)	44 (2.5)		45 (5.3)
Other Race—Including Multi-Racial	59 (2.1)	17 (1.7)	0.113	6 (1.3)	0.140
Education level
Less than 9th grade	209 (4.1)	41 (3.4)		61 (7.8)
9–11th grade or 12 grade with no diploma	362 (9.6)	98 (9.8)		76 (15.8)
High school graduate/GED or equivalent	640 (21.4)	182 (23.5)		109 (22.5)
Some college or AA degree	854 (31.7)	212 (30.2)		107 (28.1)
College graduate or above	766 (33.2)	193 (33.1)	0.583	98 (25.8)	<0.001
Ratio of family income to poverty	3.2 ± 0.1	3.2 ± 0.1	0.561	2.8 ± 0.1	<0.001
Current smoker	482 (16.1)	142 (17.8)	0.213	70 (14.0)	0.391
Persistent dry mouth in past 12 months	430 (13.7)	103 (13.3)	0.746	89 (21.0)	0.001
Persistent cold/flu in past 12 months	185 (6.4)	49 (6.1)	0.616	30 (5.2)	0.306
Frequent nasal congestion in past 12 months	875 (31.1)	239 (30.1)	0.509	138 (29.5)	0.518
Ever broke nose/serious injury to the face/skull	421 (17.7)	107 (18.0)	0.800	73 (19.7)	0.353
Ever had head injury/loss of consciousness	407 (16.7)	116 (18.8)	0.255	61 (16.2)	0.747
Ever had two or more sinus infections	1032 (44.7)	259 (41.0)	0.081	124 (34.6)	0.003

Data are presented as n (weighted % of participants) for categorical variables and weighted mean ± SE for continuous variables. Survey-weighted Wilcoxon rank sum test used for continuous variables. Survey-weighted Chi square test used for categorical variables.

*Measured GD defined as failure to correctly identify 0.32 or 1 M NaCl solutions as salty, or 1 mM quinine solution as bitter.

**Measured OD defined as identification of 3 or more out of 8 odorants incorrectly on 8-item Pocket Smell Test.

Several differences were noted among study participants with self-reported CSD. Compared to all subjects, subjects with sOD were less likely to have a college degree or above (27.2% vs 33.2%, P = 0.008), had a lower income-to-poverty ratio (3.0 ± 0.2 vs 3.2 ± 0.1, P = 0.037), and were more likely to be current smokers (21.2% vs 16.1%, P = 0.001), have persistent cold/flu symptoms in the past 12 months (12.4% vs 6.4%, P < 0.001), have frequent nasal congestion in the past 12 months (42.5% vs 31.1%, P = 0.002), have a history of 2 or more sinus infections (54.4% vs 44.7%, P = 0.003), and have a history of head injury/loss of consciousness (23.9% vs 16.7%, P < 0.001). Compared to all subjects, subjects with sGD also had a lower income-to-poverty ratio (2.9 ± 0.2 vs 3.2 ± 0.1, P = 0.033) and were more likely to be current smokers (21.8% vs 16.1%, P < 0.001), have persistent cold/flu symptoms in the past 12 months (17.9% vs 6.4%, P < 0.001), have frequent nasal congestion in the past 12 months (44.7% vs 31.1%, P = 0.005), have a history of 2 or more sinus infections (53.6% vs 44.7%, P = 0.023), and have a history of broken nose or serious injury to the face/skull (23.7% vs 17.7%, P = 0.040) (Table 1).

Likewise, there were also several differences noted in the study population with measured CSD. Compared to all subjects, subjects with mOD had a greater mean age (64.5 ± 0.8 vs 58.1 ± 0.3, P < 0.001), were less likely to be female (43.7% vs 53.2%, P = 0.009), were less likely to have a college degree or above (25.8% vs 33.2%, P < 0.001), had a lower income-to-poverty ratio (2.8 ± 0.1 vs 3.2 ± 0.1, P < 0.001), and were less likely to have a history of 2 or more sinus infections (34.6% vs 44.7%, P = 0.003). There were no statistically significant differences between subjects with mGD compared to those without CSD (Table 2).

When comparing mean HEI, subjects with sGD had significantly lower mean scores compared to all subjects (54.7 ± 1.03 vs 56.4 ± 0.47, P = 0.035). There were also trends toward lower mean HEI scores in subjects with sOD and mGD, but an association between mOD and lower mean HEI was not detected (Figure 2, Table 3).

Table 3.

HEI by Chemosensory Dysfunction: NHANES 2013–2014.

	All Subjects	Self-reported GD		Self-reported OD		Measured GD*		Measured OD**
Characteristic	N = 2,831	N = 409	P	N = 570	P	N = 726	P	N = 451	P
Healthy eating index	56.4 ± 0.47	54.7 ± 1.03	0.035	55.1 ± 0.86	0.090	55.2 ± 1.00	0.093	55.1 ± 1.07	0.805
Q1	702 (24.5)	123 (28.0)		169 (29.4)		208 (27.8)		103 (24.6)
Q2	704 (25.3)	110 (31.5)		142 (26.2)		175 (25.9)		121 (27.1)
Q3	718 (24.8)	97 (18.7)		132 (21.2)		180 (23.0)		114 (22.4)
Q4	707 (25.4)	79 (21.7)	0.066	127 (23.1)	0.120	163 (23.3)	0.266	113 (25.9)	0.780

Data are presented as n (weighted % of participants) for categorical variables and weighted mean ± SE for continuous variables. Survey-weighted Wilcoxon rank-sum test used for continuous variables.

* Measured GD defined as failure to correctly identify 0.32 or 1 M NaCl solutions as salty, or 1 mM quinine solution as bitter.

** Measured OD defined as identification of 3 or more out of 8 odorants incorrectly on 8-item Pocket Smell Test.

Upon further analysis of specific dietary constituents, subjects with sOD had higher mean intake of total saturated fats (26.1 ± 0.6g vs 25.3 ± 0.3g, P = 0.044). There were trends toward less dietary intake of all dietary components in the mOD group, with significantly lower mean intake of total calories (1872.4 ± 49.6 kcal vs 2010.2 ± 24.2 kcal, P = 0.022), total fat (72.3 ± 2.7g vs 78.6 ± 1.0g, P < 0.001), total saturated fats (23.5 ± 0.8g vs 25.3 ± 0.3g, P = 0.005), total monounsaturated fatty acids (25.3 ± 1.0g vs 27.5 ± 0.4g, P < 0.001), total polyunsaturated fatty acids (17.0 ± 0.9g vs 18.4 ± 0.4g, P = 0.006), sodium (3122.2 ± 97.2 mg vs 3353.2 ± 37.0 mg, P = 0.008), potassium (2509.8 ± 69.8 mg vs 2684.7 ± 26.1 mg, P = 0.005), and caffeine (132.9 ± 7.2 mg vs 177.0 ± 6.8 mg, P < 0.001). There were no statistically significant differences in dietary intake in the sGD or mGD groups (Table 4).

Table 4.

Dietary Constituents as Potential Moderators of Relationship Between Chemosensory Impairment and HEI—2-Day Average.

	All Subjects	Self-reported GD		Self-reported OD		Measured GD*		Measured OD**
Characteristic	N = 2,831	N = 409	P	N = 570	P	N = 726	P	N = 451	P
Energy (kcal)	2010.2 ± 24.2	1999.1 ± 60.9	0.647	2029.0 ± 44.7	0.251	2027.1 ± 59.3	0.59	1873.4 ± 49.6	0.022
Total sugars (gm)	102.0 ± 2.0	103.7 ± 3.8	0.647	104.5 ± 3.8	0.284	105.5 ± 4.5	0.484	99.4 ± 2.5	0.657
Total fat (gm)	78.6 ± 1.0	79.1 ± 3.1	0.543	80.1 ± 1.8	0.074	79.2 ± 2.5	0.893	72.3 ± 2.7	<0.001
Protein (gm)	80.4 ± 0.6	78.5 ± 2.4	0.839	79.2 ± 1.5	0.529	79.4 ± 1.7	0.43	74.0 ± 2.5	0.005
Carbohydrate (gm)	236.7 ± 3.4	235.2 ± 5.9	0.887	239.0 ± 4.9	0.201	240.0 ± 9.0	0.7	225.1 ± 5.5	0.243
Total saturated fatty acids (gm)	25.3 ± 0.3	25.9 ± 1.2	0.988	26.1 ± 0.6	0.044	25.8 ± 0.8	0.532	23.5 ± 0.8	0.005
Total monounsaturated fatty acids (gm)	27.5 ± 0.4	27.1 ± 1.1	0.226	26.1 ± 0.6	0.307	27.7 ± 1.0	0.999	25.3 ± 1.0	<0.001
Total polyunsaturated fatty acids (gm)	18.4 ± 0.4	18.6 ± 0.8	0.256	18.7 ± 0.6	0.355	18.2 ± 0.6	0.628	17.0 ± 0.9	0.006
Cholesterol (mg)	284.0 ± 4.5	283.8 ± 14.4	0.8	285.2 ± 11.9	0.376	270.6 ± 5.2	0.077	260.7 ± 11.3	0.08
Sodium (mg)	3353.2 ± 37.0	3299.7 ± 106.3	0.798	3317.7 ± 79.3	0.934	3312.6 ± 89.3	0.525	3122.2 ± 97.2	0.008
Potassium (mg)	2684.7 ± 26.1	2588.3 ± 78.3	0.982	2667.6 ± 29.8	0.753	2660.5 ± 47.3	0.681	2509.8 ± 69.8	0.005
Caffeine (mg)	177.0 ± 6.8	180.4 ± 15.2	0.344	199.5 ± 11.6	0.114	179.5 ± 12.0	0.67	132.9 ± 7.2	<0.001
Alcohol (gm)	8.9 ± 0.7	8.7 ± 1.7	0.908	9.1 ± 1.2	0.606	9.2 ± 0.7	0.806	7.7 ± 1.3	0.08

Data are presented as weighted mean ± SE for continuous variables.

*Measured GD defined as failure to correctly identify 0.32 or 1 M NaCl solutions as salty, or 1 mM quinine solution as bitter.

**Measured OD defined as identification of 3 or more out of 8 odorants incorrectly on 8-item Pocket Smell Test.

Survey-weighted Wilcoxon rank sum test used for continuous variables.

On multivariable logistic regression adjusting for age, gender, BMI, race, education, smoking, income/poverty ratio, diabetes, depression, history of sinonasal morbidity and history of head trauma, subjects with sOD had 33% higher odds of lowest quartile HEI (OR 1.33, CI: 1.04-1.70) compared to subjects without CSD. There were also trends toward lowest quartile HEI in subjects with phantom smells, sGD, and mGD that did not reach statistical significance (Figure 3, Table 5).

Table 5:

Multivariable Logistic Regression Assessing the Association Between Healthy Eating Index and Chemosensory Dysfunction.

	HEI Q1	HEI Q2	HEI Q3	HEI Q4†
Variable	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)
Measured olfactory dysfunction	1.05 (0.75, 1.48)	1.14 (0.86, 1.51)	0.83 (0.57, 1.23)	1.00
Self-reported olfactory dysfunction	1.33 (1.04, 1.70)*	0.92 (0.65, 1.29)	0.87 (0.65, 1.17)	1.00
Problem with smell in past 12 months	1.09 (0.79, 1.52)	0.77 (0.43, 1.37)	0.82 (0.51, 1.32)	1.00
Had change in ability to smell since age 25	1.04 (0.82, 1.31)	0.83 (0.56, 1.23)	0.83 (0.58, 1.19)	1.00
Phantom smells	1.71 (1.00, 2.91)	0.83 (0.42, 1.65)	0.90 (0.49, 1.65)	1.00
Measured gustatory dysfunction	1.33 (0.94, 1.89)	1.11 (0.85, 1.43)	1.03 (0.82, 1.30)	1.00
Self-reported gustatory dysfunction	1.31 (0.87, 1.97)	1.18 (0.72, 1.93)	0.92 (0.60, 1.39)	1.00
Had problem with taste past 12 months	1.52 (0.83, 2.76)	1.51 (0.71, 3.20)	1.14 (0.67, 1.95)	1.00
Change in ability to taste food flavors since age 25	1.09 (0.61, 1.96)	1.12 (0.63, 2.01)	0.95 (0.53, 1.73)	1.00
Persistent taste in mouth past 12 months	1.31 (0.69, 2.48)	1.29 (0.74, 2.26)	0.92 (0.44, 1.89)	1.00

Multivariate logistic regression model adjusted for age, body mass index, gender, race, education, income/poverty ratio, diabetes, depression, ever had ≥ 2 sinus infections, ever had head injury/loss of consciousness, ever broke nose/serious injury to face/skull, frequent nasal congestion in past 12 months, persistent cold/flu in past 12 months, and persistent dry mouth in past 12 months.

*P ≤ 0.05.

† Reference value.

Discussion

This NHANES analysis explores the relationship between olfactory and gustatory dysfunction and diet on a national level. When assessing intake of specific dietary components, there were differences in dietary intake amongst subjects with sOD and mOD compared to those without CSD, with subjects in the mOD group noted to have significantly lower intake of almost all macronutrients. Moreover, subjects with sOD were more likely to have bottom-quartile HEI compared to those without CSD when controlling for covariates, suggesting that Americans with perceived olfactory dysfunction may have poorer overall diet quality.

Previous literature assessing the effects of CSD on diet are largely limited to small cohort studies, and to our knowledge, this is the first study to demonstrate an association between CSD and dietary changes on a population level. Interestingly, subjects with either sGD or mGD had no difference in mean intake of specific dietary components compared to subjects without CSD; this may reflect the multifactorial nature of food selection, including physiologic perception of taste, as well as psychologic and neurologic processes of “liking”, “wanting”, and food reward.^30,31

Subjects with mOD had decreased levels of overall dietary component intake, which adds to a growing body of evidence suggesting that olfaction and olfactory memory may be significant factors in food pleasure, food selection, and overall consumption.^32–34 In addition, studies have shown that patients with eating disorders have alterations in olfactory capacity,³⁵ and community-dwelling elderly persons with lower BMIs were more likely to have olfactory deficits,³⁶ which further supports the hypothesis that there may be a relationship between olfactory function and food selection. In this analysis, the mOD group had a lower intake of most dietary components, but did not reduce their intake of total carbohydrates, total sugars, cholesterol, and alcohol. Unexpectedly, the sOD group did not show this trend. This finding underscores the complex interplay between olfactory dysfunction and dietary choices, and warrants further investigation.

Despite nutrient intake variation among different CSD subgroups, only subjects with sOD were more likely to have lowest quartile HEI-measured diet quality, even when controlling for sociodemographic factors and medical comorbidities. These findings suggest that the perceptual experience of food may drive overall diet quality, and that this perceptual experience may be linked to subjective olfaction. Indeed, a previous study found that in the elderly population, adding odors to food may increase intake, suggesting that decreased perception of food odors may be linked to food enjoyment.³⁷ Moreover, these findings are consistent with prior studies suggesting that those with olfactory deficits may have poorer diet quality. Namely, a longitudinal study of 557 adults aged 60 years or older demonstrated that women with poor olfaction were less likely to adhere to national dietary guidelines.¹⁵ The mean age of subjects in the aforementioned study was higher than that of subjects in this NHANES analysis (70.7 vs 58.1 years), and this current analysis demonstrated a relationship between CSD and poorer diet quality while controlling for age and gender.

There are several limitations to the current study. The NHANES is a cross-sectional database without longitudinal follow-up, so it is not possible to draw conclusions about the direction of causation or the mechanism of the hypothesized association between CSD, dietary changes and poor diet quality. Therefore, while one may hypothesize that chemosensory dysfunction is driving dietary changes in this cohort, it is not possible to reject the hypothesis that dietary changes are a causal factor in chemosensory dysfunction. Furthermore, there was low concordance seen in this sample between subjective and measured chemosensory outcomes for both olfaction and gustation and therefore for this analysis these four outcomes variables should be considered independently. Further studies should evaluate the interplay between patients’ perceived versus psychophysical chemosensory function and how this may affect dietary choices. Additionally, while the NHANES oversamples racial groups such as Blacks and Hispanics, these results may not be generalizable to individuals belonging to other ethnic groups such as Asians and other minorities that were not oversampled in NHANES. Moreover, assessment of diet quality in this study was determined by HEI. While HEI has been developed by the USDA and previously validated,²⁶ it is only one measure of diet quality and is specific only to recommended dietary standards in the US. Further, HEI was divided into quartiles for the purposes of this analysis and while the results suggest a strong association between CSD (sOD) and lowest quartile HEI, the clinical significance of having lowest quartile HEI is yet to be determined. Finally, while these results are generalizable to the US population, they cannot be extrapolated to populations outside of the US.

Despite these limitations, these data suggest that at a population level, subjective olfactory dysfunction is associated with poor diet quality even when adjusting for sociodemographic and medical confounders. These results may provide the basis for future large scale and prospective studies assessing specific dietary changes in patients with poor olfaction, along with potential strategies to improve dietary intake and quality in these patients.

Conclusion

This is the first study to utilize representative population-level data from US adults to analyze the association between chemosensory dysfunction and diet. The results of this study suggest that olfactory dysfunction is associated with dietary changes and subjective olfactory dysfunction is associated with a decrease in overall diet quality, even when controlling for other possible sociodemographic confounders. While future studies are warranted to further characterize specific dietary changes associated with chemosensory dysfunction, the findings of this study suggest that patients with these deficits may have altered dietary intake, and could be at higher risk for poorer diet quality.

Supplemental Material

sj-pdf-1-ajr-10.1177_19458924211016611 - Supplemental material for Association Between Chemosensory Dysfunction and Diet Quality in United States Adults

Supplemental material, sj-pdf-1-ajr-10.1177_19458924211016611 for Association Between Chemosensory Dysfunction and Diet Quality in United States Adults by Christopher R. Roxbury MD Isaac A. Bernstein BS Sandra Y. Lin MD Nicholas R. Rowan MD in American Journal of Rhinology & Allergy

Footnotes

Authors’ Note

This work was presented at the American Rhinologic Society at the American Academy of Otolaryngology-Head and Neck Surgery, September.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.Financial Support

ORCID iDs

Christopher R. Roxbury

Nicholas R. Rowan

Supplemental Material

Supplementary material for this article is available online.

References

Bigman

Age-related smell and taste impairments and vitamin D associations in the U.S. Adults. National health and nutrition examination survey. Nutrients. 2020; 12(4):984.

Chan JYK, Garcia-Esquinas E, Ko OH, Tong M, Lin SY. The association between diabetes and olfactory function in adults. Chem Senses. 2018; 43(1):59–64.

Hur K, Choi JS, Zheng M, Shen J, Wrobel B. Association of alterations in smell and taste with depression in older adults. Laryngosc Investig Otolaryngol. 2018; 3(2):94–99.

Bhattacharyya

Kepnes

LJ.

Contemporary assessment of the prevalence of smell and taste problems in adults. Laryngoscope. 2015; 125(5):1102–1106.

Gallo S, Byham-Gray L, Duffy VB, Hoffman HJ, Hayes JE, Rawal S. Associations of olfactory dysfunction with anthropometric and cardiometabolic measures: findings from the 2013-2014 national health and nutrition examination survey (NHANES). Physiol Behav. 2020; 215:112702. 1

Rawal S, Hoffman HJ, Bainbridge KE, Huedo-Medina TB, Duffy VB. Prevalence and risk factors of Self-Reported smell and taste alterations: results from the 2011–2012 US national health and nutrition examination survey (NHANES). CHEMSE. 2016; 41(1):69–76.

Mattes RD, Cowart BJ, Schiavo MA, Arnold C, Garrison B, Kare MR, Lowry LD. Dietary evaluation of patients with smell and/or taste disorders. Clin Nutr. 1990; 51(2):233–240.

Mattes

Cowart

BJ.

Dietary assessment of patients with chemosensotyr disorders. J Am Diet Assoc. 1994; 94(11):50–56.

Mattes-Kulig

Henkin

RI.

Energy and nutrient consumption of patients with dysgeusia. J Am Diet Assoc. 1985; 85(7):822–826.

10.

Aschenbrenner K, Hummel C, Teszmer K, Krone F, Ishimaru T, Seo HS, Hummel T. The influence of olfactory loss on dietary behaviors. Laryngoscope. 2008; 118(1):135–144.

11.

Griep MI, Verleye G, Franck AH, Collys K, Mets TF, Massart DL. Variation in nutrient intake with dental status, age and odour perception. Eur J Clin Nutr. 1996; 50(12):816–825.

12.

Duffy

Backstrand

Ferris

AM.

Olfactory dysfunction and related nutritional risk in free-living, elderly women. J Am Diet Assoc. 1995; 95(8):879–884.

13.

Stevenson RJ, Mahmut MK, Horstmann A, Hummel T. The aetiology of olfactory dysfunction and its relationship to diet quality. Brain Sci. 2020; 10(11):11.

14.

Schiffman

Graham

BG.

Taste and smell perception affect appetite and immunity in the elderly. Eur J Clin Nutr. 2000; 54(S3):S54–S63.

15.

Gopinath B, Russell J, Sue CM, Flood VM, Burlutsky G, Mitchell P. Olfactory impairment in older adults is associated with poorer diet quality over 5 years. Eur J Nutr. 2016; 55(3):1081–1087.

16.

Zang Y, Han P, Burghardt S, Knaapila A, Schriever V, Hummel T. Influence of olfactory dysfunction on the perception of food. Eur Arch Otorhinolaryngol. 2019; 276(10):2811–2817.

17.

Rasmussen VF, Vestergaard ET, Hejlesen O, Andersson CUN, Cichosz SL. Prevalence of taste and smell impairment in adults with diabetes: a cross-sectional analysis of data from the national health and nutrition examination survey (NHANES). Prim Care Diabetes. 2018; 12(5):453–459.

18.

NHANES Questionnaires, Datasets, and Related Documentation. https://wwwn.cdc.gov/nchs/nhanes/Default.aspx. Accessed April 27, 2021.

19.

Rawal S, Hoffman HJ, Honda M, Huedo-Medin TB, Duffy VB. The Taste and Smell Protocol in the 2011-2014 US National Health and Nutrition Examination Survey (NHANES): test-retest reliability and validity testing. Chemosens Percept. 2015; 8(3):138–148.

20.

Dalton P, Doty RL, Murphy C, Frank R, Hoffman HJ, Maute C, Kallen MA, et al. Olfactory assessment using the NIH toolbox. Neurology. 2013; 80(11 Suppl 3):S32–S36.

21.

Coldwell SE, Mennella JA, Duffy VB, Pelchat ML, Griffith JW, Smutzer G, Cowart BJ, et al. Gustation assessment using the NIH toolbox. Neurology. 2013; 80(11 Suppl 3):S20–S24.

22.

Doty

Shaman

Dann

Development of the university of Pennsylvania smell identification test: a standardized microencapsulated test of olfactory function. Physiol Behav. 1984; 32(3):489–502.

23.

Hoffman

Rawal

, et al. New chemosensory component in the U.S. National health and nutrition examination survey (NHANES): first-year results for measured olfactory dysfunction. Rev Endocr Metab Disord. 2016; 17(2):221–240.

24.

Bartoshuk LM, Duffy VB, Green BG, Hoffman HJ, Ko CW, Lucchina LA, Marks LE, et al. Valid across-group comparisons with labeled scales: the gLMS versus magnitude matching. Physiol Behav. 2004; 82(1):109–114.

25.

Sharafi M, Rawal S, Fernandez ML, Huedo-Medina TB, Duffy VB. Taste phenotype associates with cardiovascular disease risk factors via diet quality in multivariate modeling. Physiol Behav. 2018; 194:103–112.

26.

Guenther PM, Kirkpatrick SI, Reedy J, Krebs-Smith SM, Buckman DW, Dodd KW, Casavale KO, Carroll RJ. The healthy eating index-2010 is a valid and reliable measure of diet quality according to the 2010 dietary guidelines for americans. J Nutr. 2014; 144(3):399–407.

27.

How the HEI Is Scored | USDA-FNS. https://www.fns.usda.gov/how-hei-scored. Accessed April 27, 2021.

28.

Rothman

KJ.

No adjustments are needed for multiple comparisons. Epidemiology. 1990; 1(1):43–46.

29.

Armstrong

RA.

When to use the bonferroni correction. Ophthalmic Physiol Opt. 2014; 34(5):502–508.

30.

Berridge

KC.

Food reward: brain substrates of wanting and liking.

Neurosci Biobehav Rev. 1996;20(1):1–25.

31.

Frank GK, Reynolds JR, Shott ME, Jappe L, Yang TT, Tregellas JR, O'Reilly RC. Anorexia nervosa and obesity are associated with opposite brain reward response. Neuropsychopharmacol. 2012; 37(9):2031–2046.

32.

Ferris

Duffy

VB.

Effect of olfactory deficits on nutritional status does age predict persons at risk?

Ann NY Acad Sci. 1989; 561(1):113–123.

33.

Mccrickerd

Forde

CG.

Etiology and pathophysiology sensory influences on food intake control: moving beyond palatability. Obes Rev. 2016; 17(1):18–29.

34.

Fine

Riera

CE.

Sense of smell as the central driver of pavlovian appetite behavior in mammals. Front Physiol. 2019; 10:1151.

35.

Platte

Gallant

Schweiger

, et al. Olfaction in eating disorders and abnormal eating behavior: a systematic review. Front Psychol. 20156; 6:1431.

36.

Fluitman KS, Nadar HJ, Roos DS, Berendse HW, Keijser BJF, Nieuwdorp M, Ijzerman RG, et al. The association of olfactory function with BMI, appetite, and prospective weight change in Dutch Community-Dwelling older adults. J Nutr Health Aging. 2019; 23(8):746–752.

37.

Rolls

BJ.

Do chemosensory changes influence food intake in the elderly?

Physiol Behav. 1999; 66(2):193–197.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB