For smell testing, the two four-item versions (A and B) of the NHANES Pocket Smell Test (Sensonics International, Haddon Heights, New Jersey, USA), developed in conjunction with the NIH, were sequentially administered, resulting an eight-item ‘scratch and sniff’ test.23 (link) The eight odorants (chocolate, strawberry, smoke, leather, soap, grape, onion and natural gas) were presented in a fixed order. A participant was required, in a forced-choice situation, to identify each odorant from four alternative names. Smell impairment was defined as not being able to correctly identify six or more of the eight odours, each from a list of four possible responses.8 (link) A recent validation study demonstrated moderate-to-good test–retest reliability of the NHANES smell protocol (intraclass correlations were 0.82 and 0.69 for 2-week and 6-month intervals, respectively).23 (link) Of note, the eight odorants used in NHANES test are components of the 40-item University of Pennsylvania Smell Identification Test (UPSIT).24 (link) Our definition of smell impairment approximately corresponds to the definition of being unable to correctly identify 29 or more of the 40 odours using the UPSIT test (see online supplementary table S1 ).
The taste tests employed in this NHANES survey included a tongue tip taste test and a whole-mouth taste test. In the tongue tip test, the taste stimuli (0.32 mg/mL quinine (bitter) and 58.5 mg/mL NaCl (salty) in 10 mL solution) were presented on a cotton swab that was gently moved across the tip of the tongue in a standardised manner.23 (link) Participants were asked to identify the taste (salty, bitter, sour, something else, no taste) and rate the perceived intensity on the Generalised Labelled Magnitude Scale.25 (link) A 30 s interval was interspersed between stimulus presentations, during which time participants rinsed their mouths with water. In the whole-mouth taste test, participants swished 10 mL of each tastant solution (19.5 mg/mL NaCl, 58.5 mg/mL NaCl or 0.32 mg/mL quinine) for 3 s, expectorated and rinsed their mouths with water. The participants then were asked to identify the taste quality and rate the solution's intensity on a standardised scale. As a replication test, another whole-mouth taste test for salt was performed at the end of the chemosensory test. The participants were randomised to receive either a 0.32 M NaCl or a 1 M NaCl salt solution.
A recent test–retest reliability and validity examination of NHANES taste test protocol demonstrated a reasonable correlation between quinine whole-mouth measurement (0.32 mg/mL) and other taste measurements, including tongue tip tests of NaCl (r=0.53) and quinine (r=0.44), and whole-mouth tests of NaCl (r=0.60 for 19.5 mg/mL NaCl and 0.77 for 58.5 mg/mL NaCl), sucrose, citric acid and propylthiouracil, suggesting that the whole-mouth quinine assessment was a reasonable assessment for overall taste functioning.23 (link) Thus, in our study, instead of intensity ratings, failing to correctly identify quinine in the whole-mouth test was used to define taste dysfunction. In sensitivity analyses, we defined taste impairment as failing to correctly identify quinine (tongue tip and whole-mouth test) or NaCl (tongue tip and whole-mouth test).
The taste tests employed in this NHANES survey included a tongue tip taste test and a whole-mouth taste test. In the tongue tip test, the taste stimuli (0.32 mg/mL quinine (bitter) and 58.5 mg/mL NaCl (salty) in 10 mL solution) were presented on a cotton swab that was gently moved across the tip of the tongue in a standardised manner.23 (link) Participants were asked to identify the taste (salty, bitter, sour, something else, no taste) and rate the perceived intensity on the Generalised Labelled Magnitude Scale.25 (link) A 30 s interval was interspersed between stimulus presentations, during which time participants rinsed their mouths with water. In the whole-mouth taste test, participants swished 10 mL of each tastant solution (19.5 mg/mL NaCl, 58.5 mg/mL NaCl or 0.32 mg/mL quinine) for 3 s, expectorated and rinsed their mouths with water. The participants then were asked to identify the taste quality and rate the solution's intensity on a standardised scale. As a replication test, another whole-mouth taste test for salt was performed at the end of the chemosensory test. The participants were randomised to receive either a 0.32 M NaCl or a 1 M NaCl salt solution.
A recent test–retest reliability and validity examination of NHANES taste test protocol demonstrated a reasonable correlation between quinine whole-mouth measurement (0.32 mg/mL) and other taste measurements, including tongue tip tests of NaCl (r=0.53) and quinine (r=0.44), and whole-mouth tests of NaCl (r=0.60 for 19.5 mg/mL NaCl and 0.77 for 58.5 mg/mL NaCl), sucrose, citric acid and propylthiouracil, suggesting that the whole-mouth quinine assessment was a reasonable assessment for overall taste functioning.23 (link) Thus, in our study, instead of intensity ratings, failing to correctly identify quinine in the whole-mouth test was used to define taste dysfunction. In sensitivity analyses, we defined taste impairment as failing to correctly identify quinine (tongue tip and whole-mouth test) or NaCl (tongue tip and whole-mouth test).