Persistent decline in FEV1 exceeding 20% from baseline: observed decline in FEV1 exceeding 20% from baseline, 3 months after a confirmed decline (2 consecutive declines within 1 month) in FEV1 exceeding 20% from baseline. Second pulmonary function test (PFT) that confirmed decline was to occur within 14-42 days of the decline. Persistence: PFT that established persistence was to occur within 60-120 days of the confirming (2nd) decline. Index Visit: date subject had final Scheduled spirometry; was to occur within 2 months of Institutional Review Board/Ethics approval of April 2008 amendment.

Full Analysis Set: all randomized subjects.

Change from Baseline: mean of value of observed forced expiratory volume in the first second of forced exhalation [FEV1] in liters [L] at observation minus Baseline value. Index Visit: date subject had final scheduled spirometry; was to occur within 2 months of Institutional Review Board/Ethics approval of April 2008 amendment.

FAS; (n) = number of subjects with analyzable data at observation for Exubera® and Non-Exubera®, respectively.

Endpoint committee adjudicated the endpoint based on review of medical and hospital records, and results were classified using standard criteria. Definite: definite pneumonia, definite COPD, or definite asthma; possible: possible pneumonia, possible COPD, possible asthma, probable obstructive lung disease not otherwise specified or probable acute bronchitis; definite or possible: either definite or possible; insufficient: insufficient data.

FAS

Elapsed time, in days, from the start of a subject's participation in the study to the date of the first report of an event subsequently confirmed (according to protocol definition) as meeting the criteria for pulmonary SAE composite. Censoring time: elapsed time, in days, from the start of a subject's participation in the study to the latest contact with the subject for the particular analysis set of interest. Cox proportional hazards model to estimate treatment effect.

FAS. Due to early study termination, originally planned inferential analysis for time to event was not done.

Endpoint committee adjudicated the endpoint based on review of medical and hospital records, and results were classified using standard criteria (confirmation of deaths by blinded adjudicator(s) through medical records or death certificates). Patients meeting the endpoint All Cause Mortality after adjudication by the endpoint committee.

FAS

Time to all-cause mortality: elapsed time, in days, from the start of a subject's participation in the study to the date of the event subsequently confirmed (according to protocol definition) as meeting the criteria for all-cause mortality. Censoring time: elapsed time, in days, from the start of a subject's participation in the study to the latest contact with the subject for the particular analysis set of interest. Cox proportional hazards model to estimate treatment effect.

FAS. Due to early study termination, originally planned inferential analysis for time to event was not done.

Confirmed FEV1 decline: any two consecutive declines that are >= 14 days apart. The pulmonary function test that established persistence occured >= 60 days after the initial decline. A confirmed decline: any two consecutive declines ≥ 14 days apart. The third PFT that established persistence was to occur ≥ 60 days after the initial decline. Index Visit: date the subject had his/her final scheduled spirometry was to occur within 2 months of Institutional Review Board/Ethics approval of April 2008 amendment.

FAS

Endpoint committee adjudicated based on review of medical/hospital records; results classified using standard criteria. Definite: definite MI or stroke; Possible: possible MI or stroke; Other (non-MI, non-stroke): other cardiovascular event (non-MI, non-stroke); Definite or possible: either definite or possible or both; Insufficient: insufficient data; Death from cardiovascular or cerebrovascular: cardiovascular or cerebrovascular event; Definite or possible or death from cardiovascular or cerebrovascular: either definite or possible or both or cardiovascular or cerebrovascular event.

FAS

Elapsed time, in days, from the start of a subject's participation in the study to the date of the first event subsequently confirmed (according to protocol definition) as meeting the criteria for cardiovascular SAE composite. Censoring time: elapsed time, in days, from the start of a subject's participation in the study to the latest contact with the subject for the particular analysis set of interest. Cox proportional hazards model to estimate treatment effect.

FAS. Due to early study termination, originally planned inferential analysis for time to event was not done.

Endpoint committee adjudicated the endpoint based on review of medical and hospital records, and results were classified using standard criteria. Definite or possible: anaphylaxis, angioedema/urticaria, bronchospasm or possible allergic reaction not otherwise specified (NOS); Insufficient: insufficient data.

FAS

Elapsed time, in days, from the start of a subject's participation in the study to the date of the first event subsequently confirmed (according to protocol definition) as meeting the criteria for allergic response. Censoring time: elapsed time, in days, from the start of a subject's participation in the study to the latest contact with the subject for the particular analysis set of interest. Cox proportional hazards model to estimate treatment effect.

FAS. Due to early study termination, originally planned inferential analysis for time to event was not done.

Baseline HbA1c: the latest determination prior to beginning study participation. Change from Baseline: HbA1c at observation (falling within the time window associated with a given analysis set) minus the baseline value.

FAS; (n) = number of subjects with analyzable data at observation for Exubera® and Non-Exubera®, respectively.

Baseline HbA1c taken as the latest determination prior to beginning study participation. Change = on-study value (for measurements falling within the time window associated with a given analysis set) minus the baseline value. Linear model with terms for treatment, baseline HbA1c, time on study, and subject within treatment.

FAS. Due to early study termination, the originally planned inferential analysis (linear model) for change from baseline was not done.

Elapsed time, in days, from the start of subject's participation in the study to the first reading of FEV1 that is: 20% or more below the subject's latest pre-study measurement, subsequently confirmed as a >20% decline [(baseline observed value minus visit observed value)/by baseline observed value *100], and assessed as persistent as defined by protocol process. Censoring time: elapsed time, in days, from the start of a subject's participation in the study to latest valid FEV1 measurement for the particular analysis set of interest. Cox proportional hazards model to estimate treatment effect.

FAS. Due to early study termination, originally planned inferential analysis for time to event was not done.