Skip to main content

Table 2 Results of the best fitting model for each outcome variable (accounting for seasonality and autocorrelation)

From: The impact of celebrity influence and national media coverage on users of an alcohol reduction app: a natural experiment

 

B (95% CI)

p

Age, cubic model

 Trend

0.03 (0.01, 0.05)

.003

 Change in slope

−1.53 (−2.48, −0.58)

.002

 Change in slope2

0.12 (0.03, 0.21)

.011

 Change in slope3

−0.003 (−0.01, −0.001)

.019

 Level

8.17 (5.47, 10.87)

<.001

Sex (% female), cubic model

 Trend

−0.04 (−0.07, −0.01)

.015

 Change in slope

7.54 (3.30, 11.77)

.001

 Change in slope2

−0.69 (−1.09, −0.28)

.001

 Change in slope3

0.02 (0.01, 0.03)

.002

 Level

−27.71 (−36.54, −12.89)

<.001

Employment type (% non-manual), linear model

 Trend

−0.01 (−0.06, 0.04)

.712

 Change in slope

−0.11 (−0.34, 0.13)

.387

 Level

3.14 (−0.54, 6.82)

.098

AUDIT scorea, quadratic model

 Trend

−0.02 (−0.02, −0.01)

.006

 Change in slope

0.21 (0.01, 0.41)

.045

 Change in slope2

−0.06 (−0.01, 0.002)

.154

 Level

−1.43 (−2.70, −0.16)

.031

Percentage at-risk drinkersa, cubic model

 Trend

0.003 (−0.04, 0.04)

.864

 Change in slope

0.63 (−1.53, 2.78)

.571

 Change in slope2

−0.11 (−0.31, 0.10)

.308

 Change in slope3

0.004 (−0.002, 0.01)

.169

 Level

−2.54 (−9.55, 4.47)

.480

Number of days used the appa, quadratic model

 Trend

0.01 (−0.01, 0.02)

.445

 Change in slope

−0.42 (−0.67, −0.17)

.002

 Change in slope2

0.02 (0.01, 0.03)

.002

 Level

2.30 (0.73, 3.88)

.005

Number of sessions on the appa, quadratic model

 Trend

0.004 (−0.03, 0.03)

.801

 Change in slope

−0.67 (−1.21, −0.13)

.017

 Change in slope2

0.03 (0.01, 0.05)

.016

 Level

3.45 (0.25, 6.65)

.038

Percentage of screens vieweda, cubic model

 Trend

−0.02 (−0.04, −0.01)

.004

 Change in slope

0.31 (−0.57, 1.18)

.495

 Change in slope2

−0.05 (−0.13, 0.03)

.233

 Change in slope3

0.002 (−0.0004, 0.004)

.124

 Level

−0.68 (−3.68, 2.31)

.657

Time on appa, linear model

 Trend

0.004 (−0.04, 0.05)

.849

 Change in slope

−0.03 (−0.35, 0.29)

.833

 Level

−4.14 (−9.15, 0.88)

.110

Percentage follow-up responsea, linear model

 Trend

0.02 (−0.03, 0.06)

.449

 Change in slope

−0.10 (−0.32, 0.11)

.356

 Level

−1.57 (−5.34, 2.21)

.418

Reduction in weekly alcohol consumption (ITT)b, linear model

 Trend

0.004 (−0.003, 0.01)

.288

 Change in slope

−0.02 (−0.06, 0.02)

.374

 Level

0.23 (−0.43, 0.89)

.489

  1. aadjusted for socio-demographic characteristics
  2. badjusted for baseline AUDIT scores
  3. Change in slope2: quadratic trend
  4. Change in slope3: cubic trend